Energy Convergence: The Beginning of the Multi-Commodity Market (Wiley Finance)

Energy

convergence The Beginning of the Multi-Commodity Market

PETER C. FUSARO

John Wiley & Sons, Inc.

Energy

convergence

John Wiley & Sons Founded in 1807, John Wiley & Sons is the oldest independent publishing company in the United States. With offices in North America, Europe, Australia, and Asia, Wiley is globally committed to developing and marketing print and electronic products and services for our customers’ professional and personal knowledge and understanding. The Wiley Finance series contains books written specifically for finance and investment professionals as well as sophisticated individual investors and their financial advisors. Book topics range from portfolio management to e-commerce, risk management, financial engineering, valuation, and financial instrument analysis, as well as much more. For a list of available titles, please visit our web site at www.Wiley Finance.com.

Energy

convergence The Beginning of the Multi-Commodity Market

PETER C. FUSARO

John Wiley & Sons, Inc.

Copyright © 2002 by Peter C. Fusaro. All rights reserved. Published by John Wiley & Sons, Inc. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted under Sections 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4744. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 605 Third Avenue, New York, NY 10158-0012, (212) 850-6011, fax (212) 850-6008, E-Mail: [email protected]. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold with the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional person should be sought. This title is also available in print as ISBN 0-471-21946-0. Some content that appears in the print version of this book may not be available in this electronic edition. For more information about Wiley products, visit our web site at www.Wiley.com.

This book is dedicated to the people of New York City and their enduring spirit.

contents

CHAPTER 1 The New Millennium in Energy Trading

1

Peter C. Fusaro, President, Global Change Associates Inc.

CHAPTER 2 The Importance of Market Indexes in Emerging Industries

5

Dr. Antoine Eustache, Global Index Manager, Dow Jones Newswires

CHAPTER 3 Weather Derivatives and Reinsurance

15

Nick Ward, Spectron Energy Group, London

CHAPTER 4 Looking Forward: The Development of Bandwidth Market Liquidity

33

J.P. Crametz, RateXchange Labs

CHAPTER 5 New Techniques in Energy Options

51

Robert Brooks, Ph.D., CFA, President, Financial Risk Management LLC

CHAPTER 6 The New Accounting Rules for Derivatives: FAS 133 and Its Impact on Energy Convergence

89

Dr. Nedia Miller, Options Principal, MILLER CTA, member of NYMEX

vii

viii

CONTENTS

CHAPTER 7 The Central and Eastern European Energy Sector Reforms: Convergence versus Divergence

101

Dr. Markus Reichel, President, EconTrade Deutschland GmbH

CHAPTER 8 An Italian Road Map to the New Energy Markets

131

Alessandro Mauro, Risk Analyst, Energia SpA

CHAPTER 9 Freight Trading: The Emerging Commodity Market

143

Kirk H. Vann, CEO, Freight Advanntage and Advannce Energy

CHAPTER 10 Market Risk in Electric Generation Finance

153

William A. Klun, Vice President, DZ Bank, AG

CHAPTER 11 Convergent Systems

177

P. Kumar, President, and Shiva Gowrinathan, Vice President Client Services, Nirvanasoft Inc.

CHAPTER 12 Energy Risk Management in the Merger Context

191

Howard L. Margulis, Partner, Squire, Sanders & Dempsey LLP

CHAPTER 13 Managing Energy Risk for Industrial and Large Commercial Entities

201

Kelly Douvlis

CHAPTER 14 Green Finance: The Emerging Financial Markets for Protecting the Environment Peter C. Fusaro, President, Global Change Associates Inc.

213

Contents

CHAPTER 15 Energy Convergence: What’s on the Horizon?

ix

223

Peter C. Fusaro, President, Global Change Associates Inc.

GLOSSARY OF ENERGY RISK MANAGEMENT TERMS

233

ENDNOTES

239

INDEX

243

CHAPTER

1

The New Millennium in Energy Trading Peter C. Fusaro President, Global Change Associates Inc.

he premise of this book is that commodity trading is evolving in many areas of the energy complex and extending into emerging commodity markets, such as emissions, telecommunications broadband, and weather trading. However, the thread that brings all these markets into focus is that all are emerging and converging markets. No book currently published has tackled this subject for commodity trading, and what have been published are sporadic magazine articles. In this book, the best emerging market energy experts have been chosen to disclose the latest developments in the new field of energy convergence and multicommodity trading. This book is also meant as a companion piece to the author’s other two books on energy risk management entitled Energy Risk Management (McGraw-Hill, 1998), and Energy Derivatives: Trading Emerging Markets (Energy Publishing Enterprises, 2000). This chapter provides the overview of developments of the new emerging markets catalyzed by the Internet electronic trading developments. The chapter also lays out the broad themes of the book and argues why the time is now for the development of these multicommodity markets and why they failed in the past. Running through the individual chapters, we see that Dr. Antoine Eustache, Global Index Manager at Dow Jones Newswires of Princeton, New Jersey, explains the intricacies of index construction in the age of the Internet. Dr. Eustache has created the first electric power indexes in

T

1

2

THE NEW MILLENNIUM IN ENERGY TRADING

the United States and Europe for trading electricity, and the first bandwidth indexes during 2001. He is looking at creating metals and emissions trading indexes for the future. He is the world’s expert in creating emerging market indexes for trading and provides insights into future market developments for other commodities. Chapter 3 examines the weather reinsurance markets and was written by Nick Ward, an experienced broker at Spectron Energy Group in London. The weather trading markets have generated much press over the past five years but actually little liquidity despite all the hype. Nick Ward will show why this is occurring and why this is leading us back to the reinsurance markets and their financial instruments. The status of the telecommunications bandwidth trading is examined in Chapter 4 by J.P. Crametz of RateXchange Labs, Palo Alto, California. RateXchange is one of the first exchanges trading telecommunications bandwidth. Mr. Crametz is a leading researcher at Stanford University’s Business Laboratories. While bandwidth trading will never be as large as the energy market, it is a growing and robust market ripe for commoditization. It is also a converging market with electric power. Mr. Crametz is working on the ground floor of one of the fledgling bandwith trading exchanges. The importance of pricing energy options is shown by Dr. Robert Brooks in Chapter 5. Dr. Robert Brooks, President, Financial Risk Management LLC and finance professor, University of Alabama, Birmingham, Alabama, is an options expert and modeler. Energy options are unique to themselves and have implied volatility like no other commodity market. Dr. Brooks is a renowned energy options specialist who has worked with the Southern Company in trading strategies. He helps define the necessity of options trading in emerging markets and explains their value to the layperson. In Chapter 6, Nedia Miller, CTA of New York, New York. Dr. Miller is a leading expert of the new FAS 133 hedge accounting rules for energy. She is an ex-NYMEX floor trader and has given many seminars on FAS 133 to KPMG, Royal Bank of Scotland, and Bank of Montreal. This ruling is changing not only how energy companies hedge, but their trading strategies as evidenced by the Enron debacle. Dr. Markus Reichel, President of EconTrade in Dresden, Germany, in Chapter 7 explains the antecedents of trading in these emerging mar-

The New Millennium in Energy Trading

3

kets. Dr. Reichel focuses primarily on the emerging markets of Poland and Ukraine, which are just beginning to open up for energy trading. Dr. Reichel is a leading expert on Eastern Bloc energy liberalization and trading. In Chapter 8, Alessandro Mauro, Manager of Risk at Energia SpA in Milan, Italy, focuses on the Italian energy trading markets. He was formerly with PwC and ENI, the Italian oil company. Mr. Mauro will focus on the emerging gas and electric markets in Italy that have evolved from the EU Energy Directives. In Chapter 9, Kirk Vann, CEO, Freight Advanntage and Advannce Energy in Houston is a very experienced oil trader examining the changes underway in the tanker markets. Mr. Vann has more than 25 years trading oil for Enron and Glencore, and has established a model for the shipping industry on how they should hedge their freight rates, which are very variable and volatile. Tankers provide the lifeblood of global energy trading for oil and gas. In Chapter 10, William A. Klun, vice president of DZ Bank, AG, in New York investigates market risk for financing merchant power plants. These are plants that have no retail load. P. Kumar and Shiva Gowrinathan of the New York-based software company Nirvanasoft, Inc., examine and explain the importance of convergent software systems for energy trading, in Chapter 11. They examine the current software architecture and take a look at these new information technology developments for the energy industry. In Chapter 12, Howard Margulis, partner, Squire, Sanders & Dempsey LLP in New York investigates energy risk management in the merger context looking at the legal and tax issues as well. Mr. Margulis is a leading attorney in the area of energy projects finance and shows how the use of energy derivatives can reduce the cost structure of energy projects. Mr. Margulis was formerly a partner at Baker and McKenzie. In Chapter 13, Kelly Douvlis examines retail gas and electricity risk for industrial and commercial customers. In Chapter 14, the author, Peter Fusaro, proposes a radically new concept in emissions trading through the use of structured products. He presents the background for the Kyoto Protocols and the United States experience in emissions trading as the nexus for launching CO2 emissions

4

THE NEW MILLENNIUM IN ENERGY TRADING

trading. Green finance is using the project finance mechanism to jump-start the CO2 emissions trading market. Finally, in Chapter 15, the author investigates the phenomenon of energy convergence for financial, energy, and Internet markets. Besides summarizing the themes of the previous chapters of the book, Fusaro provides a forward spin on how these new financial and commodity markets will converge in the age of market liberalization, consolidation, and globalization.

CHAPTER

2

The Importance of Market Indexes in Emerging Industries Dr. Antoine Eustache Global Index Manager, Dow Jones Newswires

uring the past three decades of the twentieth century, financial derivatives grew from a little-known industry to what the Federal Reserve Board terms the most significant event in finance.1 By June 1998 global positions in over-the-counter (OTC) financial derivatives covering all categories of market risks stood at more than $70 trillion and since 1990 this industry has maintained a steady yearly growth of more than 20 percent.2 As phenomenal as this explosion may be, it would not have been possible without reliable market indexes. This is particularly true in newly emerging sectors, especially in the OTC markets where prices are not always validated through the clearing mechanisms typical in most other organized markets. This chapter offers a brief perspective on some of the key characteristics that have made market indexes so critical to the growth of these markets, with special emphasis on financial derivatives.

D

WHY ARE MARKET INDEXES SO IMPORTANT? At the center of every derivative contract is a price. Depending on the market, this price may be simply that of the underlying commodity or a basket of prices commonly referred to as a market index. Whether the outright price or a basket index is used to value the underlying asset, the quality of the data utilized in the valuation process is critical to the success of the contract. In the equities markets, where prices are determined

5

6

THE IMPORTANCE OF MARKET INDEXES IN EMERGING INDUSTRIES

through an exchange and validated through various clearing mechanisms, the quality issues are less a concern. In the commodities markets, however, the issues are not just centered on data quality. Since very few traders will trade on the prices generated by one company alone regardless of the quality of the data, reliable market indexes are essential to the growth of financial derivative. In more mature markets real-time price indexes in the form of reporting the last sale have existed for decades. With the advent of the computer age, Internet-based, real-time price indexes are becoming a trend in the commodities markets. However, in most emerging markets the lack of liquidity makes real-time indexes impossible during the initial phase of market development. Typically this phase may last several years until market participants grow comfortable first with new market practices and then with the tools and technologies crucial to the development of a real-time market. As a result most commodities indexes published during this embryonic phase of market development are based on voluntary reporting of end-of-day prices. Typically the reporting arrangements revolve around informal data exchanges between market participants and index publishers. In some small number of cases, however, the data reporting mechanism may be more elaborate. In an industry where fundamental factors lead to extreme price volatility, an index can serve as a powerful tool for stabilizing revenue and expenditure through various hedging mechanisms. However, unless market participants can agree on the reliability of this index, it is nearly impossible to create risk management products suitable for price risk mitigation. For instance, the concept of an option to buy a particular good or service is not completely foreign to the industry professional in newly emerging markets. Often, as was the case for the electric and natural gas industries prior to deregulation, companies deciding to produce or buy long-term supply would issue a request for a proposal, hoping someone would come along with an acceptable offer. What seemed to be the best offers were way beyond what these companies would pay, if they had access to a market index that not only gave a clearer indication of the fair value of the products but also the flexibility to reset these contracts as prices fell. The introduction of market indexes in these sectors removed this impediment, making it possible for buyers and sellers to acquire various options with ease and in a timely fashion. Price indexes can be extremely useful in improving business operations

Why Are Market Indexes So Important?

7

and analyzing markets. They can also reduce the cost of doing business where information is scarce. Markets need a certain level of price transparency to function. Without that transparency, traders need to rely on their own price gathering mechanism. Such activities are often expensive, time-consuming, and unreliable. Even the most elaborate price gathering systems cannot substitute for an independent source of information most traders can agree on. As a result many would rather use indexes published by a neutral party as a reference benchmark. These benchmarks are often purchased by traders and end-users at a fraction of the costs involved in gathering their own information. Market indexes can also serve as major triggers for new capital investment newly emerging industries. Often investors seeking to enter a market are reluctant to move in too early for fear of diverting valuable resources from profitable activities. Nor do they want to miss out on a great opportunity by waiting until it’s too late. As a result, many will time their entry around the publication of reliable market indexes. A typical example is the electricity market. During the early phases of commoditization, many U.S. companies wanted to enter the European power market but were reluctant to do so for fear of being too early. This fear was amplified by the fact that the few electric utilities that had bought assets in the United Kingdom had not fared very well. The publication of two major market indexes, the Swiss Electricity Price Index and the Central European Power Index, in 1997 and 1998 allayed those fears to the point that by 1999 several U.S. companies had set up shop there. By the end of 2001 almost every major U.S. power marketer has set up shop there. Suppliers often fear that market indexes will fuel competition, which in turn will invariably result in the collapse of market prices. Likewise, buyers often take the position that more transparency will result in lower prices. As intuitively appealing as this assumption seems at first blush, it does not always hold. In reality transparency is not necessarily interchangeable with competition. Neither are price indexes. Certainly, an index can foster the growth of a market prone to adopt competitive practices. But in isolation an index is simply a passive barometer that helps track fundamental changes in a given industry. In other words, in isolation, an index is merely a tracking device that simply validates the status quo. Take the case of the electricity and natural industries. Long before they became deregulated, consumers were well aware of how much they paid for gas and power thanks to stringent regulatory requirements

8

THE IMPORTANCE OF MARKET INDEXES IN EMERGING INDUSTRIES

for price disclosure. The knowledge and awareness of the price paid for electricity and natural gas did not mean that they could switch suppliers. More compelling is the way electricity and natural gas prices moved in the years that followed liberalization. In the years that followed the creation of the electricity market, wholesale electricity prices skyrocketed to levels not seen prior to deregulation. In 1995, for instance, average wholesale electricity prices were $15 per megawatt hour; in 2000 the yearly average was at $200. As a direct consequence of this upturn many utilities that once campaigned against market indexes are now their most avid supporters and consumers. At this time consumers with great market power believe their unique relationship with suppliers guarantees them the lowest possible price and that a market index could end that. Although counterintuitive at first blush, this argument may sometimes hold in a few cases. For instance, in markets like forest products where supply is plentiful and large buyers are positioned to set the prices, suppliers are often held hostage by the constant threat that customers will switch to the competitor with the lowest price unless they offer major discounts way below the market price. Often these customers contract for merchandise knowing full well they will renegotiate the deal, should prices move in their favor. And 9 times out of 10, producers will comply. This after the fact that negotiations tend to cause major inaccuracies in any index based on these prices, unless appropriate measures are taken to thwart them. In most emerging markets where competitive forces are at play, lack of transparency tends to keep buyers and sellers on the sidelines waiting for reliable price signals. Far from bringing a complete halt to market activity, this lack of transparency creates profit opportunities for a small number of arbitrageurs. With vast amounts of resources at their disposal these market agents often specialize in profiting from imbalances in the price of a good across time and space. Often this involves the purchase of a good or service at one price and the concurrent sale of that good at a higher price, resulting in a risk-free profit. Although legitimate, these activities have limited benefits to suppliers and consumers. While they may not eliminate these inefficiencies completely, market indexes can bolster confidence in the prevailing market price and enable buyers and sellers to agree on the true market value of their goods. Greater confidence in the true market value of product ultimately generates more liquidity, making it easier for sellers to find buyers willing to transact.

Some Methodological Considerations

9

SOME METHODOLOGICAL CONSIDERATIONS The elements included in an index ultimately determine how useful this index will be to the marketplace. For instance an index designed for trend analysis may be a powerful tool for econometric studies and time series analysis. However, this index may have limited applications in the trading community. The data inputted in the compilation of an index will also have a major bearing on the accuracy and reliability of this index. In most emerging markets buyers and sellers are often reluctant to disclose the price at which they buy or sell goods or services. As a result, most indexes published in these informal markets are typically based on market surveys. Further complicating the issue, the survey data is often based on list prices with no means for verification. Indexes based on such data are open to manipulations and are often not trusted. Due to these limitations the new generations of indexes designed for trading purposes are transactionbased price indexes. Often the need for an index emerges along with the need to create a liquid and vibrant market. In many such cases there are no existing trades to support that index. This is often the case in emerging markets where trading rules and practices are foreign to the corporate culture. Constructing an index for a market that is in its early development stages is extremely problematic. It is time-consuming, it requires a tremendous amount of intellectual capital, and often the returns outweigh the costs. As a result only a small number of specialized organizations are involved in supplying these indexes. These are for the most part organizations with vast intellectual assets. Often these organizations play the role of market facilitators that bring together market participants seeking to trade a particular product. They are often highly trusted. Their services range from assisting the market with product standardization to defining rules governing trading practices. In the process they often create a series of elaborate mechanisms for collecting trade data that can be traced back to the sources. In the absence of a formal clearing mechanism they play an auditing function that puts to rest any doubt of the validity of the data used to compile the indexes. The emergence of Internet trading will also have a bearing on how future market indexes are designed. For instance since the mid-1990s there has been a proliferation of Internet trading platforms covering a wide

10

THE IMPORTANCE OF MARKET INDEXES IN EMERGING INDUSTRIES

spectrum of commodities markets. These systems range from neutral trading platform to large proprietary systems owned by one or several market operators. Although some—well-positioned and well-financed—have captured a sizable portion of the traded market, few will ever provide the full flexibility of, or supplant, the over-the-counter market. Certainly systems with the built-in functionality that simplify the trading process will generate a great deal of traffic, perhaps enough to support very robust indexes. But the evidence from some of the most successful of the neutral platforms suggests the market share of any of these exchanges should not be expected to be in the double digits in large economies like that of the United States or the European Union. On most of these platforms, for each product that is heavily traded there will be many more that will be so illiquid that the underlying data may not be robust enough to support reliable indexes. To further complicate the issues, the existence of several competing exchanges with no clear differentiation in the products traded will split the liquidity to the point that the value of the data will diminish. In the end there will be lots of data points but few will ever be used as the underlying ones for derivative trading. It is, however, possible to create robust indexes from the combined output generated from the various platforms. To the extent these exchanges are open to exploring a workable solution, these indexes could be strong enough to serve as the underlying benchmarks for settling futures contracts and various other derivatives. This would prove to be a profitable solution, since in all likelihood the exchanges could have joint ownership rights to these indexes.

APPLICATIONS Beyond their use as benchmarks for pricing physical commodities, market indexes are widely used in pricing financial products, such as swaps and options. Using an index for pricing a physical commodity is very straightforward. For instance, depending on how accurately the index tracks the market, merchants may tie the price of a good directly to that index. Alternatively, they may price their products at a differential to that index. For instance, merchant powers are known to enter into multiyear agreements with large end users, whereby they agree to tie the price of electric energy directly to an index. Depending on the cases involved, it is also common to see contracts where electric energy is priced at, say, the Dow Jones California/Oregon border electricity price index (COB Index) plus or minus x dol-

Applications

11

lars per megawatt hour. This practice is also common in the natural gas and oil markets. The application of market indexes to pricing financial derivatives is slightly more complicated. These applications range from plain vanilla swaps and options covering a single commodity to the most exotic. This section provides some simple examples illustrating the use of indexes in the power sector. Plain Vanilla Options: Take the case of the electricity market. In 2000 and 2001 the Dow Jones electricity price index for California/Oregon border reached respective highs of $800 and $600 and respective lows of $14 and $16 per megawatt hour for nonfirm power. A buyer who has reasons to believe the recurrence of such price volatility is possible in 2002 may decide to hedge against the upside risks. The buyer may decide to purchase a call option. For a small premium of, say, $8 per megawatt hour, the buyer may enter into an agreement that gives the right, not the obligation, to purchase electric energy at $100 per megawatt hour. This $100 is called the exercise or the strike price, meaning it is the price at which the buyers will exercise their right to purchase power if they so choose. This strike price may be set at the COB Index averaged over an entire calendar month or any combination thereof. Should market prices remain below $100, the buyers will mostly likely not exercise this option. In that case their only loss will be the premium paid to purchase that option. Should prices, as feared, move up substantially this option to purchase at $100 could result in major savings to the buyers. Figure 2.1 illustrates this scenario. Note that the savings are unlimited while losses are capped to the $8 paid for purchasing the option. The example illustrated in Figure 2.1 could apply equally well to sellers seeking to protect revenue in the event of a major decline in prices. For instance a supplier selling power at $200 per megawatt hour may not want to settle for the low prices of $14 and $16 recorded in 2000 and 2001, respectively. They may decide they want more for their power in 2002. For instance, for a small premium of, say, $10, suppliers may decide to enter into an agreement that gives them the right to sell power at $100 per megawatt hour. This type of agreement is commonly known as a put option. It gives suppliers the right, not the obligation, to sell power at $100 per megawatt hour if they so choose regardless of how low prices fall. To simplify the deal the seller may choose to use a monthly average of the COB indexes. Figure 2.2 illustrates how much of the seller’s revenue is protected in the event prices fall below the $100 target. Should prices rise

12

THE IMPORTANCE OF MARKET INDEXES IN EMERGING INDUSTRIES

Savings—$/MWH

80 60 40 20 0 –20

0

50

100

150

200

Index Value—$/MWH

FIGURE 2.1 Consumer’s Savings Resulting from the Purchase of a Call Option

Revenue Protection— $/MWH

Using a Price Index

80 60 40 20 0

–20 0

50

100

150

200

Index Value—$/MWH

FIGURE 2.2 Revenue Protection from the Purchase of a Put Option Using a Price Index

above $100, the supplier merely forgoes the $10 premium paid to protect that revenue and sells their power at the going market rate. Simply put, it’s a win-win situation. Plain Vanilla Swaps: Options are only a few of the applications available to organizations seeking to use indexes as tools for price risk mitigation. For instance, in many instances a risk manager may choose to use various alternatives to options or a combination of options and some other applications. Financial swaps are one of these alternatives. Some applications range from the most basic financial swaps, known as plain vanilla swaps, to the most complex. Figure 2.3 offers a simple illustration of how price indexes are used in price mitigation using a plain vanilla swap. It is possible to observe from the examples illustrated in Figures 2.1 and 2.2 that the buyer and the seller have equal and opposite needs. On the one hand the

13

Summary

$800.00

Buyer receives floating price Buyer pays fixed price

$600.00 $400.00 $200.00 $0.00 09-Jan-01

28-Feb-01

19-Apr-01

08-Jun-01

28-Jul-01

16-Sep-01 05-Nov-01

FIGURE 2.3 Fixed Price Swap: Buyers Are Protected from Major Upswings Source: Dow Jones & Company—California/Oregon border price index

buyer, fearing a major price increase, wants to cap the total energy expenditure to no more than $100 per megawatt hour. On the other hand the seller, fearing a collapse in energy prices, fills the need to limit the downside risk. Suppose an intermediary could match the buyer and the seller. This intermediary could guarantee the buyer the cash needed to purchase power in the open market, in return for a fixed payment over a predetermined time. In return for this guarantee of stability the buyer, mostly concerned with meeting the budget, would be expected to forgo the benefits associated with any further decline in prices. The intermediary could in turn guarantee the seller a fixed payment in line with that received from the buyer, in return for a cash swap equal to some value of the index plus a differential to that index.

SUMMARY Price indexes are powerful instruments that can play a central role in facilitating the growth of emerging markets. Not only do they provide a tool that helps simplify the exchange of goods and services, they also offer great protections against revenue shortfalls that result from volatile market prices. Although they can be viewed as major instruments of change without the proper market mechanisms in place, they can be viewed only as passive barometers that track fundamental changes in a given industry.

CHAPTER

3

Weather Derivatives and Reinsurance Nick Ward Spectron Energy Group, London

INTRODUCTION Weather has become a hot new market over the past few years with more than 5,000 deals representing a notional value of more than $7.5 billion have been consummated. Estimates of total quantifiable economic exposure to weather vary and will always be subject to some uncertainty. However, it is clear that the development of a widely available tool that isolates the specific elements of weather risk that companies face will have a wideranging impact on business at large. In this chapter, Nicholas Ward, head of New Markets at Spectron, one of Europe’s leading independent OTC energy brokers, maps out the development of weather derivatives and their interrelationship with the reinsurance sector.

HISTORICAL WEATHER RISK MANAGEMENT Weather exposure has, of course, always been a fact of life. Indeed, human development and geographical progress around the globe could be seen as largely a process of experimentation with various techniques for adaptation to weather, (or more correctly, climate); positive and negative, extreme and prevailing conditions. Clothing, housing, crop techniques, and nomadic lifestyles have all demonstrated human awareness of the influence of weather variables, and the distinction between normal conditions and (possible but

15

16

WEATHER DERIVATIVES AND REINSURANCE

rare) extreme conditions. In fact, it is as much the probabilistic analysis as the determination of mean conditions that has led to long-run success in any settlement: even if you anticipate living only another 60 years, you should make contingency plans for that 100-year flood, since it may come tomorrow. On a broader scale, the development of trade could be seen as a way of maximizing the returns from weather and climate diversity, while ensuring survival of the species through the spreading of risk, a form of selfinsurance against weather (and numerous other risks, of course). In fact, the risk transfer can be seen both as a form of classic derivativestyle comparative advantage swap between equally risk-averse counterparties, and alternatively as an assumption of risk, at a price, by ultimate underwriters of weather risk. An orange growing region and an elk farming (or vodka producing or salt mining) region may have struggled to exist in isolation, but through trade they can source the full range of other products needed for continued existence, and thus maximize returns under their local weather inputs, rather than attempting the inefficient local production of staple goods, such as wheat or rice. By generating a store of wealth from trade, they can survive adverse extreme events. Likewise, a beneficent system of government might tax such trade in good years, in order to be able to provide disaster relief in extreme bad years.

REINSURANCE BROADLY DEFINED In essence, all forms of insurance are risk transfer businesses, the distinction being that whereas insurance is the process of a licensed party providing risk coverage to an unlicensed policy buyer, reinsurance involves the further sale of risk between two licensed parties. This sale is referred to as a cession of risk by the insurer to the reinsurer or underwriter. In practice, while insurance may be seen as a many-to-one operation (many individuals, for example, will purchase virtually identical policies from the same household contents insurer), reinsurance may often be a one-tomany operation, where the risk that is being ceded by the insurer, in any one cession, is so large as to require multiple reinsurance underwriters (See Figure 3.1). In order for a reinsurance contract to apply, the following must generally hold.

Facultative and Treaty Reinsurance

17

FIGURE 3.1 Reinsurance ■ There must be an insurable interest (i.e., the parties must be able to demonstrate that real assets are involved). ■ There must be risk. ■ The contract must transfer some, or all, of the risk from the buyer to the underwriter. Note that coverage should reflect the actual exposure (loss), not the physical damage. For example, if a hotel is severely damaged by a hurricane, the coverage is not limited to the cost of repairs (or demolition and rebuilding), but also to the economic loss caused by loss of revenue (present and future sales).

FACULTATIVE AND TREATY REINSURANCE To manage this risk transfer, there are two principal forms of reinsurance: facultative and treaty. Facultative reinsurance deals with specific risks, analyzing large individual cases on a stand-alone basis, whereas treaty reinsurance considers portfolios of (generally related) risks. Under treaty reinsurance, the underwriter has little or no opportunity to cherry pick the risks that are most attractive or that best fit the reinsurance company’s risk appetite. The classic sourcing method for reinsurance business has been the cession by an insurer of part of the risk incurred by policies written (or about to be written) by the insurer. Again, the reinsurance contract will be between the reinsurer and the insurer, not between the reinsurer and the ultimate policyholders. The insurer will normally retain some portion of the risk, but for reasons of capacity constraints, concentration of event risk, or cash flow and balance sheet management, may need to offload specific risks or groups of risk. Details of individual risks may not be made available to any substantial

18

WEATHER DERIVATIVES AND REINSURANCE

degree under treaty reinsurance, the principle being that a reasonably homogeneous portfolio of risks that fit within certain parameters is being presented, and the underwriter is expected to accept the portfolio as a “job lot.” For facultative reinsurance, though, the underwriting process requires a more detail-oriented valuation process. The underwriter will become intimate with the full details of the underlying policy and the business surrounding it. Facultative reinsurers are the providers of risk insurance for those emerging areas of business activity for which tools like actuarial tables have yet to be established: satellite launches; new economies; new industries.

RISK ALLOCATION: PROPORTIONAL AND EXCESS Among the methods of risk distribution between insurer and reinsurer for treaty reinsurance, two of the principal techniques are proportional allocation and excess of loss allocation. Proportional allocation involves the insurer passing a given percentage of all premia received to the reinsurer, against the reinsurance firm reimbursing the same percentage of all payouts or losses. Proportional allocation is a simple risk-sharing arrangement, where the reinsurer and insurer are effectively facing the same risk profile. However, for excess of loss allocation the risk profiles differ significantly. Just as an individual might face a minimum claim threshold on a car or property insurance policy, the insurer will retain liability for all payouts up to a certain threshold amount. This threshold is known as the attachment point. Payouts in excess of the attachment point will be covered by the reinsurer, subject to an absolute maximum payout. The risk profile for this coverage is therefore one of less frequent (but larger) payouts (See Figure 3.2). The graph illustrates the payout profile for a fictitious portfolio of policies, which incurred payouts every year for 20 years. The graph shows the total paid by the insurance carrier to the policyholders, and the reinsurer’s contribution to that payout under both proportional and excess of loss treaty participation. While both participation methods involved total payouts of roughly 25 percent of the full portfolio payout, the distribution of payments is much more uneven for excess of loss. Although large payouts may in practice be spread over several years, the reinsurer must take into account the capital cost of provisioning for large payouts, particularly if this exposure represents a significant portion of the total risks covered in the reinsurer’s book (see Figures 3.3 and 3.4).

19

Risk Allocation: Proportional and Excess

$600,000

full payout

Annual Payout

$500,000

proportional payout excess of loss payout

$400,000 $300,000 $200,000

excess paid over $130k per annum, capped at $200k payout

$100,000 $0 1980

1985

1990

1995

2000

Year

FIGURE 3.2 Historical Comparison of Payouts: Proportional versus Excess

$250,000

Amount

$200,000 $150,000 $100,000 $50,000 0k k–1 9 180

0k k–1 7 160

k–1 5

0k

0k 140

120

k–1 3

0k k–1 1

k 100

80k –90

k 60k –70

k 40k –50

0–1

0k 20k –30 k

$0

Bracket

FIGURE 3.3 Payout Distribution—Proportional From these figures it can be seen that risk assessment for facultative and concentrated (especially excess of risk) treaty reinsurance must take proper account of extreme events. If an underwriter’s book is well-diversified, the impact of any one extreme event will be a small proportion of that reinsurer’s provisioned payout, and the reinsurer may therefore view incremental extra risk on a portfolio basis. Where a book is immature, where the new risk is very large in relation to total underwriting capacity, or where,

20

WEATHER DERIVATIVES AND REINSURANCE

$250,000

Amount

$200,000 $150,000 $100,000 $50,000

180 k–1 90k

160 k–1 70k

140 k –1 50k

120 k –1 30k

100 k –1 10k

80k –90 k

60k –70 k

40k –50 k

20k –30 k

0–1 0k

$0

Bracket

FIGURE 3.4 Payout Distribution—Excess for other reasons, the risk must considered in isolation, advanced modelling and an understanding of catastrophic as well as normal risk are required to make a proper risk assessment (see Figure 3.5). The incidence of many sets of outcomes in nature can be modelled using a Gaussian (or normal) distribution. The principle here is that the majority of numerical outcomes will be concentrated in the “body” of the curve, while extreme events are found in the “wings.” As alluded to in the introductory section, the evaluation of weather risk must take both median and extreme events into consideration; pricing the wings is a necessary skill in effective weather risk management. The skills of reinsurers in this area, and their necessary familiarity with weather and climatic data, have meant that reinsurance companies are natural participants in the market for weather risk management products.

WEATHER DERIVATIVES BROADLY DEFINED As a new market, weather derivatives have existed, under that name, since 1997. However, like credit derivatives and other products, they have existed in principle and practice (under other names) for much longer. Like credit derivatives, they essentially crystallize a component of risk that economic agents have been exposed to, and managed using combinations of other products, since the dawn of time. Simply put, a weather derivative is a formula: a structure whose ultimate

21

Key Weather Derivative Elements

wings body Number of Instances

30 25 20 15 10 5 0

Be

w lo

–1

0

–9

.5

to

–9 –8

.5

to

–8 –7

.5

to

–7 –6

.5

to

–6 –5

.5

to

–5 –4

.5

to

–4 –3

.5

to

–3 –2

.5

to

–2 –1

.5

to

–1 –

5 0.

to

0

Recorded Temperature

FIGURE 3.5 Historical Average Winter Temperature value (or pay-out) is determined by the outcome of one or more weather variables. Weather derivatives can vary greatly in terms of complexity and customization. Since their principal application is in the management of weather risk (rather than, for example, speculation), this chapter focuses on the hedger’s (or hedge provider’s) perspective. Depending on the nature of the risk to be hedged, a company may find a one-off tailored transaction, or a dynamically updated suite of individual trades to be the most appropriate course of action. Whatever the structure under consideration, a number of key elements must be identified and then quantified.

KEY WEATHER DERIVATIVE ELEMENTS The key elements to be isolated are ■ Weather variable: The general weather type in question must be narrowed down to quantifiable, readily observed, variables. Common examples include rainfall, sunshine hours, snow depth, air temperature, wave height, and wind speed. If appropriate, a combination of these may be employed.

22

■

■

■

■

■

■

WEATHER DERIVATIVES AND REINSURANCE

This may be an iterative process, as later work on quantification may reveal that one of a number of variables initially considered is statistically insignificant in its influence. The rain effect may be covered to an acceptable degree of accuracy by temperature hedging, for example. Data quality of different variables may also impact this decision. Location: The exact site for measurement of the above variable(s) should be determined where the risk is most concentrated. Again, this may prove iterative, for example where trade-offs exist between deal complexity and the perceived enhanced accuracy of multisite measurement. Data quality will again strongly influence these decisions. Period: It is important to establish the timeslice of economic exposure: Is the risk seasonal or year-round? If the exposure is known to exist for an extended period, it may prove economical to secure multiyear protection, since a lower probability of extreme outcomes will be priced in. Timing: The actual timing of the measurement needs to be specified. Is the exposure only at night? Or is it twice as great at weekends as during weekdays? Is hourly measurement necessary, or would a daily average2 be sufficient? Correlation: It is obviously critical to be able to model the nature of the exposure to weather. Are revenues proportional to weather over a period of time, and if so is this relationship linear? Are losses triggered by specific events, and is the risk then an absolute amount (as in the complete destruction of property) or is it proportional to the weather variable? Data processing: What filtering or processing of the data is needed? Closer to the correlation measurement, we must establish whether we are looking at exposure to a daily average or maximum value. Should extreme values within this series be smoothed or otherwise weighted differently? Amount: The size of the exposure, expressed in units of currency per unit of weather element. Many hedgers elect not to neutralize their entire measured exposure, but to put in place partial hedging. This can allow first-time participants to flush the product through their systems while they monitor the performance of the hedge. It also allows for a dynamic hedging policy, where changes in the company’s risk profile can be adapted to, and similar trades can be spread across, a variety of

Common Weather Derivative Structures

23

counterparties, thus limiting credit exposure. Splitting a deal into smaller tranches may also allow a hedger to take advantage of preferential pricing. However, this can backfire if the full deal size becomes known, and it is felt to hang over the market. In fact, the initially unattractive price quoted for the full exposure may have been the best deal. ■ Cap: In a convention taken from the reinsurance market, weather deals typically include a maximum payout amount. This limits the risk exposure of the hedger’s counterparty, the risk manager or market maker. Typically this may be expressed as a multiple of the deal amount, and this factor will vary according to the steepness of the distribution of expected or historical payouts for the specific deal in question.

COMMON WEATHER DERIVATIVE STRUCTURES While a one-off hedge is likely to prove to be a tailored, nonstandard transaction, it will generally be recognizably drawn from the following building blocks. Readers familiar with other derivative markets will recognize the basic derivative structures but should pay attention to differences in detail. These are the principal structures traded today. It is likely that their exact definitions will be subject to change as new entrants impose differing requirements on the products. This is particularly true of the standard underlying variables in use, such as heating and cooling degree days (HDDs; CDDs) (see section that follows), today’s predominant temperature metrics.

Swap Essentially a forward pricing transaction. A rate is quoted for the eventual outcome against the defined variable or structure. For example, 1,800 HDDs. If this is agreed by both parties to the transaction, this becomes the “fixed rate” for the deal. The “floating rate” is the actual outcome. At the end of the relevant period, the difference between the fixed and floating rates is calculated. This, multiplied by the deal amount, is the actual payment made under the deal (subject to the maximum payout cap). Typically, deals extend for only one payment period (one month or one season), and so involve a single payment at maturity.3 Observe that a swap involves a symmetrical liability for payment; either party to the deal may be required to make the settlement, depending on the net of the two rates.

24

WEATHER DERIVATIVES AND REINSURANCE

Swaps are the most basic of over-the-counter4 weather derivatives, and as such they could be regarded as a proxy underlying commodity for the other weather derivative products.

Call Also known as a “cap.” Here the symmetry of risk is removed. The hedger does not want the risk of any outgoing payment at deal maturity, so, in exchange for an upfront premium, they participate solely in any positive payout from the deal. If the hedger buys a call at $1,850, they will benefit from any outcome where the actual value exceeds $1,850, just as in a swap. The premium is the maximum amount paid out under this deal. Unlike a call option on a stock or a more traditional commodity, this product does not confer the right to buy the underlying instrument at a fixed price, this generally being nonsensical in the case of weather. Automatic execution is assumed, and the purchase is then effectively cash settled against the actual outcome. Excluding the effect of the maximum payout, the net result to a call buyer is the setting of a maximum level of financial exposure to an increase in the weather variable (i.e., the exposure is capped). Pricing such deals involves not only an idea of likely weather outcomes, but also an ability to price the volatility of these outcomes.

Put Also known as a “floor,” since it effectively puts a floor on losses. This is the inverse of the call. The buyer of a $1,750 put benefits from any outcome below $1,750, in the same manner as if they had sold a swap. Again, the buyer will never be out-of-pocket on the deal by more than the premium paid up front.

Collar A combination of a put and a call. It is possible to construct a collar that involves no exchange of premium (i.e., a costless collar). The profile is similar to a swap, but any variation within the range closest to the at-the-money price involves no payout either way. Swaps, calls, and puts may be combined in a number of ways, in order to achieve the desired payout profile.

25

Heating Degree Days (HDDS)

Future In essence, a weather swap traded on an exchange instead of Over-theCounter. Transacting on an exchange, a counterparty gains complete anonymity (although economic elements of the trade will be publicly recorded) and the positive credit aspects of transacting with a centralized counterparty. The trade-off is a loss of flexibility, as only a fixed set of standardized products will be available.

Digital Many less generic products are available in the Over-the-Counter market. These can provide, for example, for a fixed payment, only in the event that a certain critical level is breached, such as wind speed exceeding 70 miles per hour at any time within a certain period, or the occurrence of 3 successive days of rainfall within a specified period. As a rule, pricing such products is more involved, and the hedger should balance the benefit of a more tailored fit against the potential increased cost (See Figure 3.6).

HEATING DEGREE DAYS (HDDS) Degree days is a concept that originated in the calculations made by the energy and property industries to determine the cost of maintaining normal human conditions within a building, throughout the year. Conceptually, HDDs are the “area” defined by the summation of deviations below a predetermined reference temperature (typically 65°F or 18°C), over a period of time. Let H = HDD count for winter period (November 1–March 31, a total of 151 days in a nonleap year)  H = minC , A ×  where

∑

31March



h , i =1November i 

C = cap, the maximum payout allowable under the contract (e.g., $1,000,000) A = amount, the size of deal (e.g., $5,000 per HDD) hi = individual day’s HDD value, defined by: hi = max(0, Tref – Ti),

26

WEATHER DERIVATIVES AND REINSURANCE

250 200

Payouts

Thousands

150 100 50 0 00 12

50 12

00 13

50 13

00 14

50 500 14 1

50 15

00 16

50 16

00 17

50 17

00 18

swap call put collar straddle

–50

strangle

–100 –150 –200 –250

HDDs FIGURE 3.6 Comparison between Derivatives where Tref = reference temperature (e.g., 18°C) Ti = individual day’s actual temperature Cooling degree days work in a similar manner, the value being the excess of day’s temperature and a predetermined reference temperature. Other variants include energy degree days (HDDs + CDDs, the total needed to bring the temperature back to the reference point) and growing degree days, where we are concerned with a range of temperatures within which agricultural conditions are optimal.

IMPLEMENTING A WEATHER RISK MANAGEMENT POLICY Whether a company decides to purchase a tailored structure or embark on a modular risk management policy, the onus is on the decision maker and those providing the supporting analysis to understand and quantify the exact nature of the weather exposure to be managed. This includes not only the exact risks and their economic impact, but also the probability of occurrence. Weather derivatives are hedging tools that can be used as alternatives

Temperature—The Dominant Product

27

to, or in conjunction with, any number of alternative weather risk management strategies (not the least of these being physical avoidance or prevention, and the allocation of capital as self-insurance). As well as a solid grasp of the business processes exposed to weather risk, and the way in which they themselves may vary over time, a key element in the successful definition of a weather risk management policy is an intimate understanding of the nature of the weather variables isolated in the preceding checklist, and the quality of both the data provided for calculation of the derivative’s payout, and any historical data series, on which pricing will be based. In the latter, the availability of meta-data (nonnumerical information concerning, for example, the relocation of a measuring station) is almost as crucial as the provision of an orderly, uninterrupted series of data spanning multiple decades.

PROGRESS OF WEATHER DERIVATIVE MARKETS TO DATE The weather derivatives market is generally agreed to have been born in the United States in 1997. Trades were initially conducted between energy firms, with some participation from the agricultural sector as well.

TEMPERATURE—THE DOMINANT PRODUCT The dominance of energy companies in this market has led to the emergence of the degree day (principally winter heating degree days and summer cooling degree days) as the predominant metric for calibration of weather exposure. The general principle is that, assuming unit revenues are roughly constant (either through regulation, long-term contract setting, or by price hedging using power or gas derivatives), and unit costs are likewise constant (through similar mechanisms), then energy providers (gas or electric power) will face net revenue exposure to weather based on demand fluctuations. The unit margin may be constant but the volume is variable. Another reason for the current predominance of temperature within weather derivatives is that other data such as precipitation and wind speed are simply far more local in effect and the relevance of historical data is less strong. As a generalization, nontemperature products are harder to price with certainty, and have less direct correlation with the hedger’s actual exposure.5

28

WEATHER DERIVATIVES AND REINSURANCE

MARKET DEVELOPMENT Weather derivatives spread to Europe and Asia over the following years, and while the volume of trades has not reached anything like that in the United States, there has been a broad assortment of deal structures, catering to a wide variety of industries. Although hard numbers are hard to come by in this market, by the end of the year 2001 it is estimated that some 5,000 weather derivative transactions were closed globally, covering $7.5 billion in exposure. Europe’s activity is substantially lower, but the strong growth rate is highlighted by one market maker reporting deal volume in the second half of 2000 to be eight times the total traded up to that point. In all regions, activity has been split among the energy, insurance, and banking sectors. Typically the trading or market-making activity (viewed from an insurance perspective, the secondary market) has been dominated by energy companies, while the largest deals have been more readily absorbed by reinsurance firms.

PRICE ACTION DYNAMICS Perhaps uniquely among derivative markets, weather derivatives do not benefit from a market for their underlying components. This means on the one hand that (short of cloud seeding or tampering with the measurement equipment) it is not possible for a market participant to squeeze the underlying commodity in order to benefit in their derivatives position. Conversely, this lends a somewhat pure nature to price action in a liquid market; prices should respond solely to expectations of future outcomes. Since weather forecasts are generally held to be of rapidly diminishing utility when projected beyond the very short term,6 this means that mediumto long-term pricing should simply reflect variations in the modelling methods and assumptions, as well as risk capacity, of the market participants. Inevitably, though, pricing is also affected by supply and demand dynamics. The broader the acceptance of the product, the greater the liquidity of the market. In this context, a key determinant of future market success will be the extent that weather risk management is recognized as a vital tool for all areas of economic activity. This puts a high premium on efforts to market (in the fullest sense of the concept) the weather derivative product. It is a young market, and there is still much to be done.

Is This Trade an Insurance Product After All?

29

DERIVATIVES AND REINSURANCE: COEXISTENCE, COMPETITION, CONVERGENCE, AND SEGMENTATION So far, this text has presented the structural distinctions between the development of reinsurance products and derivatives. In practice, distinctions are blurred.

IS THIS TRADE AN INSURANCE PRODUCT AFTER ALL? In itself, there is nothing about a weather derivative, as just described, that commits the product to being solely viewed as an off balance sheet or capital market instrument. Weather derivative structures represent merely the mathematical definition of a payout profile. A precipitation call may be equally well-packaged as an insurance policy or a pure derivative product. Product use, and specifically the presence of an insurable interest, features strongly in how regulatory authorities determine whether to treat the product as insurance. Regarding insurable interest, the onus of proof can vary considerably. At the time of going to press, these issues are still being explored, but it is believed that in at least one European Union country, for example, a precipitation contract could attract insurance status without the need to demonstrate insurable interest. U.S. authorities have also yet to make definitive statements, but again, a “close correlation” may be sufficient to gain insurance status. Although there does not yet appear to be complete clarity on the situation yet, there is generally a degree of comfort within the market that gaming laws, for example, within individual major economies’ jurisdictions are not applicable to weather derivatives. Does it matter if the product is classified as insurance or not? The main difference in effect is generally the legal and taxation treatment under the jurisdiction concerned. A key element of this is the liability to pay an insurance premium tax. This can vary in extent and applicability from jurisdiction to jurisdiction, but levels of 9 percent of premium are not uncommon. Another is the financial statement reporting requirements. While the extensive discussion of the treatment of derivatives under FAS 133 in the United States is beyond the scope of this chapter, essentially the accounting method employed depends on whether a trade is seen as a “cash flow” hedge or a “fair value” hedge, the latter being eligible for

30

WEATHER DERIVATIVES AND REINSURANCE

mark-to-market treatment, as opposed to being posted as gains and losses through the income statement. Other benefits of the derivatives product include ease of unwinding, either through buyout pricing, which is generally readily provided to end users by market makers for structured transactions, or simply trading an offsetting deal in the market. Furthermore, the expanding use of (ISDA) International Swaps and Derivatives Association master documentation allows swift deal closure as individual deals need only be documented by a simple confirmation; this also allows netting of payments (and hence counterparty credit exposure). This may even apply across product groups. In practice, the determination of whether a trade is an insurance product or not is more pragmatic: Regulated and licensed insurance entities are the providers of insurance products in the marketplace. Where there is an economic advantage to the hedger in buying a derivative product instead, then this would be provided by an offshore entity. “Transformer” companies exist, whose function is to form a bridge between the derivative and insurance worlds. Major reinsurance players generally have their own inhouse transformer.

Overlapping Niches As a generalization it could be said that, by virtue of their different original motivations in approaching weather derivatives, the reinsurers and the derivative traders (financial and energy firms) have focused their expertise in different segments of the market: the tailored “wing risk” against the generic “body risk,” respectively. As a generalization this has held some truth to date. From the previous sections, it is clear that assessment process described for embarkation on a program of hedging with weather derivatives is not dissimilar from the analysis that an underwriter may perform in discerning the cost of providing insurance against a certain weather-related risk. Where a hedger discerns a complex weather risk exposure but is unable or disinclined to actively manage it, passing this risk over to a weather derivative market maker/structurer is analogous to the cession either of a single complex risk under facultative or a portfolio under treaty reinsurance. Likewise, the hedger who chooses to actively manage their risk portfolio is entering the derivatives trading marketplace.

Is This Trade an Insurance Product After All?

31

But these distinctions are being blurred by convergence in activities of the market participants. Many banks, investment banks, and other financial institutions are establishing Alternative Risk Transfer (ART) desks, parallel to similar practice in the reinsurance markets. Likewise, while reinsurers’ typical modus operandi has been to build portfolios of “primary” deals through direct origination activity, they have then rebalanced these by participation in the “secondary” market. Initially, this would involve seeking deals that looked as much like their primary product as possible as calls or puts with out-of-the-money strikes, preferably multiyear. This practice has changed somewhat, and some underwriters are approaching the secondary market as a valid primary source of weather business. This has even led to some marketmaking activity by reinsurers.

Capital Markets Overlap There is another element to this convergence process: the capital markets. On a fundamental level, self-insurance is an example of the crossover of capital markets and insurance: Here, instead of contracting external services for risk management (i.e., buying an insurance policy), a company may simply decide to raise extra capital and set that aside against the possibility of adverse circumstances. Another way of effecting this risk transfer to the capital markets can be as the direct transcription of an off balance sheet derivative structure to the coupon of a corporate bond issue, for example, a snowfall-linked coupon on a bond issued by a ski resort. Alternatively, a portfolio of weather derivatives may be presented as the underlying “asset” for an asset-backed securitization. The portfolio need not be confined to existing deals, but instead the prospectus may contain a definition of the nature of the portfolio’s constituents over a multiyear period, in a similar fashion to credit-linked notes and other asset-backed securities whose maturity exceeds that of the underlying assets. Within securitization, the initial focus had been in catastrophe (“cat”) bonds, but we have seen at least one “body risk” weather securitization, and many reinsurers as well as portfolio managers are investing in body risk. On both the buy and sell side, we see the direct analogy between portfolio management and reinsurance.

32

WEATHER DERIVATIVES AND REINSURANCE

WHERE DO WE GO FROM HERE? We are clearly seeing a convergence of reinsurance, capital markets, derivatives, and commodity trading functions around the central point of weather risk. Discrete skills will still apply even if these are spread across risk-transforming (i.e., market making and risk absorbing) firms from all shades of the spectrum. Specialists will approach normal and extreme risk segments in different manners. It is likely that the latter will always remain risk to be absorbed (although increasingly as a tradable yield product in the capital markets). Meanwhile, estimates already put the total volume of weather derivatives and securitizations extant as being greater than that of total catastrophe issuance and derivative trades.7 The core weather market has the potential to dwarf all other markets for risk transfer, so fundamentally does it affect economic activity. Ultimately, actual pressure to hedge weather risk (whether that be through internal realization or as a result of external pressure, for example from equity analysts) supported by extensive marketing and consulting by the weather derivatives players are needed to unearth the vast, latent enduser market. All participants have an active role to play in this development.

CHAPTER

4

Looking Forward: The Development of Bandwidth Market Liquidity J.P. Crametz RateXchange Labs

INTRODUCTION The year 2000 signaled the beginning of the bandwidth trading market with the first deal between Enron and Global Crossing announced in December 1999. There has been much activity in the bandwidth trading space since then, particularly in the establishment of market infrastructure: ever-evolving exchange models, electronic trading platforms, market makers ready to take positions, trading desks set up within many energy and utility companies, brokers to facilitate, and the establishment of pooling points/hubs. The parties are here, and invitations are extended. The real question remains: Are the guests showing up? This chapter examines the issue of market liquidity. What is a good measure of liquidity? In the absence of active trading activities, is the bandwidth trading market getting closer to being liquid, and how do we know? Are we seeing a hint of liquidity in this emerging market? Using trades/asks/bids data from various sources, we construct several dynamic (as in time series) indexes/metrics as a proxy to market liquidity, some of which are pure statistics, some of which follow from analytical models. Much like a doctor taking pulses, we are observing the symptoms. Are the symptoms consistent with a healthy path leading to a liquid bandwidth

33

34

LOOKING FORWARD: THE DEVELOPMENT OF BANDWIDTH MARKET LIQUIDITY

trading market? The answer is in the strong affirmative. This chapter tells a story of a market liquidity trend derived from available data. This chapter does not contain mathematical details of the analytical models. Interested readers can contact the authors for more information. There are many factors contributing to the development of bandwidth trading. The Telecommunications Act of 1996 (and the earlier split-up of AT&T) initiated the Big Bang of the new telecommunications era. Coupled with the World Trade Organization (WTO) decision to include telecommunications in the policy agenda, the globalization/competition genie of telecommunications can never be returned into the bottle again. An equally important policy factor is the privatization of the Internet, from a government run education/research/military facility to the innovative playing field of capitalistic genius. Companies began to liberalize state-run telecommunications during the 1990s. Governmental actions/policy alone could not have fueled the explosive development of the industry. A well-built kitchen needs its appliances. Technological advances in semi-conductors, computers, and lasers/fiber optics provide the tools to equip this kitchen. Business/consumer oriented applications (the World Wide Web, browsers, search engines, music, information, and e-commerce) round out the menu. Customers show up to patronize, which fuels competition. The stage is set for a new and expanding frontier in telecommunications: an inviting regulatory (or deregulatory) environment, enabling technology, creative applications, ready consumers, and innovative entrepreneurs. Various telecommunications-related applications lead to the explosive demand for bandwidth. In addition to business needs, the introduction and proliferation of broadband access technology compound the demand for long-haul capacity that is fast, reliable, secure, and redundant. Bandwidth trading emerged to fill a need.

CURRENT STATE OF MARKET DEVELOPMENT You do not trade bandwidth on a street corner. Trading infrastructure and intermediaries have emerged to enable market development: ■ Evolving “exchanges” models ■ Electronic trading platforms ■ Pooling points, or hubs

Current State of Market Development

■ ■ ■ ■ ■

35

Brokers Market makers Trading desks Bandwidth Trading Organization (BTO) Publications

The above list of factors contribute to the continuing development of the bandwidth trading market. These factors represent physical (pooling points/hubs), informational (electronic trading platforms, exchanges), and organizational (BTO) elements of the trading infrastructure. Some enabling elements play the role of facilitators (brokers), while others act as principals (market makers); some function as both (trading desks taking on risks as well as providing risk/asset management for their clients). To be truly commoditized, there have to be clearly defined standards: What is being traded? What are the consequences if there is nonperformance? Formed in March 2000, the Bandwidth Trading Organization (BTO) has taken on the task to establish standardized contract terms for the trading, measurement, and provisioning of bandwidth. It has been a long and difficult process as one moves from the old way of doing business to the new. Despite all the debates and apparent initial resistance, the effort to standardize is a precursor of market development. Even though final agreements on standardized contract terms have not been reached, the idea of contract standardization has taken its root. Many bandwidth trades over the past several months were completed under similar terms as those in the evolving BTO standard contract. McGraw-Hill was the first to provide daily coverage of the developing bandwidth market with its Bandwidth Market Report (BMR). In addition to news coverage of the market, the BMR also reports bid/offer prices between Los Angeles and New York on a weekly basis (through some form of informal data gathering process). Dow Jones Newswires recently started a DJ Bandwidth Intelligence Alert, with bid/offer information (NY–LA and NY–DC) and launched its first bandwidth indexes on July 2, 2001. The Scudder Publishing Group also has a Bandwidth Desk publication. Frequent and regular coverage by major news/publication organizations adds credibility and relevance to the developing bandwidth market. People are paying attention because the market will change the way bandwidth was transacted. The bandwidth market received a big boost from a new and unlikely

36

LOOKING FORWARD: THE DEVELOPMENT OF BANDWIDTH MARKET LIQUIDITY

type of player: energy companies. These companies include Duke Energy, Dynegy, El Paso, Enron, Koch, Aquila, Dynegy, and Williams. They all have a well-developed trading desk with risk management infrastructure. They can acquire the physical underlying with several market entry strategies: Acquire a telecommunications company, monetize current communications assets (existing networks), or build fiber optics routes along their pipelines. They have the capital and human resources to aggressively pursue a complementary new line of business. Brokers provide the lubricant to facilitate trades. They help potential new customers with education and market information. They also work the phone lines, use the electronic trading platforms to close deals. Notable brokers include Amerex, Tradition Financial Services (TFS), and Prebon Yamane.

IT IS ABOUT INFORMATION The reasons for the emergence of the bandwidth market have been articulated in different articles, which we will not repeat. Besides its role to facilitate trades, the bandwidth market provides information of price discovery and transparency. In fact, it is this price transparency role that provides the most value to its participants: buyers, sellers, market makers, and speculators. Financial derivatives will be created as hedging instruments to match the risk preferences of counterparties. As the bandwidth market matures and becomes liquid, credible (i.e., well-documented, verifiable, and reliable) market (price) indexes will play a central role in the creation of the associated derivative market. In an excellent exposition, Dr. Antoine Eustache details a methodological approach to creating power market indexes, which should provide a glimpse of what lies ahead in the bandwidth market (“Energy Indexation: Analyzing the Scope of Electricity Price Indexes,” in Energy Derivatives: Trading Emerging Markets, edited by Fusaro and Wilcox, Energy Publishing Enterprises, New York, NY, 2000). What can we learn today from limited market data, as the bandwidth market is developing? We, at RateXlabs, have been tracking market data from the days of the bulletin board, where the bid/offer postings were noncommittal, to the firm offers on RateXchange’s electronic trading platform (RTS—RateXchange Trading System) of today. We share the results of our continuing data analysis in the following section. The data sources include

37

Statistics and Metrics from Market Data

postings on the RTS, firm offers listings from a major carrier as well as published pricing information from McGraw-Hill’s Bandwidth Market Report and Dow Jones’ Newswire (now formalized as DJ Bandwidth Intelligence Alert). The observed data/metric trends suggest that certain arbitrage opportunities are shrinking, the bid/ask spreads are decreasing, and more routes are being offered in the open market. Market participants are observing and learning from market data and are being educated by it. More players are coming to the party. In addition to providing statistics and metrics, we also propose a simple analytical model to infer (from the statistics) that there are more ready buyers/sellers entering the market, a sign of market liquidity.

STATISTICS AND METRICS FROM MARKET DATA More Routes Are Being Offered on the Market We see the continuing trend that more routes are being offered on the market in the last half of 2000 as illustrated in Figure 4.1. There is a particularly substantial increase of routes between European cities. A

Intra-Asia Europe-Asia US-Asia US-Europe Intra-Europe 250

# of links offered

200

150

100

50

0 Jun-00

Aug-00

Oct-00

FIGURE 4.1 Increasing Number of Offerings (Different Market)

Dec-00

38

LOOKING FORWARD: THE DEVELOPMENT OF BANDWIDTH MARKET LIQUIDITY

second graph, Figure 4.2, presents the time series of increased offerings by capacity type.

FORWARD PRICES ANALYSIS We use time series forward bid/ask price data for one-month contract (for the months of March, April, and June 2001 delivery) DS-3, NY–LA, as an illustration. We observe similar trends with other parameters (other city–pairs, different capacity, or different contract duration). The data we used are from RTS, supplemented by price data from Dow Jones Newswire. The forward prices are time series data in the month of November 2000. It also includes some limited actual transitions. Prices are quoted in USD per DS0 mile, for the months of March, April, and June of 2001 between NY and LA. The following observations can be made from our analysis: ■ The number of bid/ask postings has increased substantially over the past year, pointing to increased interest and the need for a trading facilitator, particularly a clearing medium and mechanism. ■ The industry is comfortably moving toward shorter contract duration.

OC12 OC3 DS3 E1

250

# of links offered

200

150

100

50

0 Jun-00

Aug-00

Oct-00

Dec-00

FIGURE 4.2 Increasing Number of Offerings (Different Capacity)

39

Forward Prices Analysis

We see an increasing number of postings for one-month bandwidth contracts. The ability to provide near real-time switching/trading facilities is essentially to accommodate the needs of bandwidth trading development. ■ Prices for forward contracts are decreasing, indicating the continued trend of price erosion. However, prices are stabilizing. The one-month contract for April delivery declined by 9 percent from that of March. However, the monthly price decline from April to June has stabilized to 2 percent, suggesting a more balanced supply–demand relationship. ■ A more interesting metric is the ratio of bid/ask spread to that of the bid/ask average (or actual transaction price, if available), as a function of time, plotted individually for March, April, and June delivery. A narrower bid/ask spread suggests a more balanced supply–demand dynamic. When the bid/ask spread is normalized by the ask/spread average (or actual transaction price), it represents the percentage premium that a market maker is able to extract because of supply–demand imbalance. We observe the steady decline of this ratio, which gives a hint that the bandwidth market is moving toward more liquidity. Figure 4.3 is a simple bid/ask time series plot of forward prices for one-month delivery for New York and Los Angeles. The diamond data series are for March 2001 delivery, the triangle time series are for April

0.011

$ per DSO Mile

0.0105 0.01 0.0095 0.009 0.0085 0.008 31-Oct

5-Nov

10-Nov

15-Nov

FIGURE 4.3 NY–LA DS-3

20-Nov

25-Nov

40

LOOKING FORWARD: THE DEVELOPMENT OF BANDWIDTH MARKET LIQUIDITY

delivery, while the solid circle time series are for June delivery. The upper price lines are ask prices, while the lower price lines are the bid prices. The hollow circles are actual transactions. The differences between bid/ask prices (the spread) for the same delivery month are shrinking as we advance in time. The bid/ask spread (for the same delivery month) appears to be stabilizing (if we’re keeping in mind the limit of our data source). The forward prices (bid/ask brackets) also are declining as a function of the delivery month (from March to April to June), which points to price erosion. The next graph provides a more quantitative measure of this erosion trend.

TIME SERIES OF PERCENTAGE PRICE DECLINE We compute the monthly percentage price decline for the set of time series data we have. Since there are not enough transaction data, we will take the average of the bid/ask prices to represent forward price trends. For example, on November 5 the forward price for the April delivery is 4 percent less than that for the March delivery (the diamond shaped time series). On the same day, the forward price for the June delivery declines by 2 percent per month from that of the April delivery (the square shaped time series). The following picture shows such trends (see Figure 4.4). The monthly price percentage drop is most severe from March to April. April to June delivery prices drop has leveled off a bit to less than 2

2.00% 0.00% 31-Oct

5-Nov

10-Nov

15-Nov

20-Nov

–2.00% –4.00%

Mar-Apr Apr-June

–6.00%

Mar-June

–8.00% –10.00% –12.00%

FIGURE 4.4 Monthly Percentage Price Decline, One-Month Contract: NY–LA DS3

41

Time Series of Percentage Price Decline

percent per month (from a 9 percent drop between March and April). Figure 4.4 points to the need to investigate further such price trends: Is it due to some supply–demand dynamics or just the lack of ready transparency in this young market?

Bid/Ask Spread as a Percentage of Forward Prices The bid/ask spread tells a bit about the liquidity of the market (in an indirect way). With many buy/sell counterparties in the bandwidth market, the bid/ask spread tends to shrink. However, it is more informative to look at the bid/ask spread relative to the forward price (as a percentage). We plot this ratio in Figure 4.5 (the ratio of the bid/ask spread to that of the forward price). In the absence of substantial transaction data, we use the average of bid/ask prices as a surrogate to actual transaction price. This ratio (intuitively) normalizes the bid/ask spread. This ratio has an intuitive interpretation, similar to a statistical measure known as the coefficient of variation (defined as the ratio of standard deviation to the mean). We have used the bid/ask spread (or the range of prices) to replace the standard deviation. When more bid/ask offers are posted, we can convert our analysis to the coefficient of variation. This ratio shows that delivery in a more distant month (June as opposed to March) has a smaller value. First, prices are declining (declining forward curve, as implied by the surrogate forward prices, the

12.00% 10.00% 8.00%

March 6.00%

April June

4.00% 2.00% 0.00% 31-Oct

5-Nov

10-Nov

15-Nov

20-Nov

25-Nov

FIGURE 4.5 Ratio of B–A Spread to B–A Average

42

LOOKING FORWARD: THE DEVELOPMENT OF BANDWIDTH MARKET LIQUIDITY

average of bid/ask); thus, the denominator of this ratio is decreasing (which tends to increase this ratio). The fact that this ratio is declining (despite the decreasing denominator) is attributed to the even faster shrinking of the spread. This is a very positive development for the maturity/liquidity of the bandwidth market: Buyers and sellers are converging to a common price.

GEOGRAPHICAL ARBITRAGE A fundamental arbitrage principle arises from the geography of routing. When our first arbitrage analysis was performed in March 2000 between New York and various European cities, it was observed that such opportunities indeed exist. At that time, we suggested that as the market becomes mature (information becomes transparent), such opportunities will diminish. The analysis here can be viewed as a follow-up of our earlier report. We observe that such opportunities still exist today, but to a lesser extent. This suggests a learning process for market participants, and the fact that market data indeed provide price transparency to effect a more efficient market. The data we use are a combination of bid/ask postings from the RTS and pricing data from a major carrier.

United States–Europe Figure 4.6 shows the arbitrage opportunities for DS3 capacity expressed as a percentage cost of the direct route purchase over least-cost routing through an intermediate city. Note that there may be additional cost involved when purchasing (and connecting) the two segments from New York through the intermediate city to the desired European destination. Such overhead cost should be factored into the arbitrage calculation. When this is taken into account, the arbitrage potential should be discounted (e.g., by x%, or by $y to reflect such overhead costs) from that presented next. The opportunities presented are shown as a time series for the second half of 2000. We detect the shrinking of these arbitrage percentages over time, suggesting perhaps the impact of price discovery. The route listed as NY–PR (LO) represents the arbitrage opportunity that the direct purchase of DS3 capacity for the route from New York to Prague at October (the third bar in the first group) costs 59 percent more per month, than a two-segments route through London.

43

Geograhical Arbitrage

Jun-00 Aug-00 Oct-00 Dec-00 70% 60% 50% 40% 30% 20% 10% 0% ) ) ) ) ) ) ) ) ) ) ) ) O O R) O O LO LO LO LO LO LO LO LO (L (L /F I(L I(L L( A( V( A( A( U( U( O R( O M V T M P B L F B G D A W M R( NY NY NY NY NY NY NY NY NY NY NY NY -P NY

FIGURE 4.6 Geographical Arbitrage in U.S.-Europe Market (DS-3) (Route Cost Percentage over Least Cost Route)

When NY–PR (LO/FR) is used, the least-cost purchase may be routed through Frankfurt. A list of city name abbreviations appears at the end of this section.

Intra-European Market We observe less arbitrage opportunities in the Intra-European market. The major arbitrage interconnection points are Stockholm and Frankfurt. Figure 4.7 shows these opportunities for DS3 capacity in two months of 2000. The Intra-European routes listing is relatively new, showing less of a learning effect due to price transparency.

Europe-Asia and Intra-Asia New York acts as the intermediate arbitrage interconnection point between European and Asian cities. Figure 4.8 shows arbitrage opportunities for DS3 capacity. The learning effect over time is quite impressive, resulting in very slim arbitrage margins toward the end of 2000.

44

LOOKING FORWARD: THE DEVELOPMENT OF BANDWIDTH MARKET LIQUIDITY

Oct-00 Dec-00 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% LO-PR(FR)

FR-WA(LO)

LO-MO(SH)

LO-TA(SH)

FIGURE 4.7 Geographical Arbitrage in Intra-Europe Market (DS-3) (Route Cost Percentage over Least Cost Route)

Aug-00 Oct-00 Dec-00

80% 70% 60% 50% 40% 30% 20% 10% 0% LO-TK(NY)

LO-HK(NY)

LO-SG(NY)

FR-HK(LO,NY) AM-HK(LO,NY)

JP-SG(HK)

FIGURE 4.8 Geographical Arbitrage in Europe-Asia & Intra-Asia Market (DS-3) (Route Cost Percentage over Least Cost Route)

45

An Analytical Model to Track Liquidity

Oftentimes, the cheapest route between two cities (in Europe and in Asia) is constructed using New York and London as the interconnecting points, resulting in a three-segment route. In the Intra-Asia market, we found the case that Japan–Singapore can be obtained at a cheaper price by connecting the two segments via Hong Kong. See Table 4.1 for city abbreviations.

AN ANALYTICAL MODEL TO TRACK LIQUIDITY A market is liquid when a player can transact either by buying or selling at the market price in a timely manner. Liquidity is then the ability and ease to transact at the market price. If there are many players (buyers, sellers as well as speculators) actively involved in a market, it is likely that transactions will be “liquid.” At this stage of market development and in the absence of a formal definition, it is difficult to infer liquidity (or the lack thereof) or more importantly the speed that the market is approaching liquidity. We have created an analytical model that will compute the probability distribution of the bid/ask spread as a function of (among other things) the number of potential buyers and sellers. The model is based on the fact that different people will price a product differently, translating to different willingness to pay for a buyer and different willingness to part for a seller. Given this observation, as more serious buyers enter the market, the active

TABLE 4.1 Abbreviation of Cities AM BL BR BU DU FR GV HK JP LO MA MI

Amsterdam Bratislava Brussels Budapest Dublin Frankfurt Geneva Hong Kong Japan London Madrid Milan

MO NY PA PR SH SG SP TA TK WA VI

Moscow New York Paris Prague Stockholm Singapore St. Petersburg Tallin Tokyo Warsaw Vienna

46

LOOKING FORWARD: THE DEVELOPMENT OF BANDWIDTH MARKET LIQUIDITY

(i.e., highest) bid offer will be high. Likewise, as more serious sellers enter the market, the active (i.e., lowest) ask price will be low. When the active ask price is below the active bid offer, there will be a sale. On the other hand, a bid/ask spread will appear on the market (e.g., the RateXchange’s RTS) when the active ask price is above the active bid offer. Therefore, the bid/ask spread will vary according to the number of players (buyers and sellers) and their valuation of the commodity (bandwidth). We model the buyer’s (or seller’s) valuation of bandwidth as uncertain (or as random variables), each with a probability distribution. Assume that there are n(b) number of buyers and n(s) number of sellers. Buyer i has a valuation of bi (willingness to pay, or the bid price), and seller j has a valuation of aj (willingness to part, or the ask price). Note that the bi’s and aj’s are random variables. The active ask price (A) and active bid price (B) can be expressed as B = maximum {b1, b2, . . ., bn(b)}, and A = minimum {a1, a2, . . ., an(s)}. There is a spread when B < A. We are interested in the (probability) distribution of ∆ = A – B, when B < A: how likely it is for ∆ to be small, and how likely it is for ∆ to be large? Given the characteristics of the bi’s and aj’s and the value of n(b) and n(s), it is possible to derive the distribution of ∆. Our goal is to compare the distribution when n(b) and n(s) change. Without getting into the analytical details, we provide the following result of our numerical computation. There are three curves in Figure 4.9, each representing a distribution of ∆, for different values of n(b) and n(s), the number of players (or how liquid the market is). The total number of participants then equals n = n(b) + n(s). Curves 1, 2, and 3 represent increasing numbers of participants. We observe that as n increases, the distribution of ∆ becomes sharper toward the origin (when ∆ = 0). The conclusion is that as more players (sum of buyers and sellers) enter into bandwidth trading, it is more likely to observe a smaller bid/ask spread (the probability distribution is higher at the small value of ∆ when n increases). We also show the trend for the distribution of ∆ as n keeps increasing, in Figure 4.10.

47

An Analytical Model to Track Liquidity

Series 1 Series 2 Series 3

Bid/Ask Spread

FIGURE 4.9 Distribution of Bid/Ask Spread with Increasing “Liquidity”

Bid/Ask Spread

FIGURE 4.10 Distribution of Bid/Ask Spread with Increasing “Liquidity”

48

LOOKING FORWARD: THE DEVELOPMENT OF BANDWIDTH MARKET LIQUIDITY

Again we observe that (Figure 4.10) as n increases, it becomes more likely that the bid/ask spread will approach zero. The bid/ask spread trend statistics examined earlier agree with the analytical conclusion consistent with a market with increasing numbers of active players, the conclusion that the bid/ask spread is shrinking. Note that a player agreeing to participate (as in the signing of a legal contract to post binding bid/ask offers) does not translate into an active player; they can simply sign on to observe. The bid/ask spread trend we observe is, however, consistent with the unobserved number of active players through the willingness to pay (and to part) model introduced in this section.

CONCLUSION This chapter examines the issue of market liquidity. What is a good measure of liquidity? In the absence of active trading activities, is the bandwidth trading market getting closer to being liquid, and how do we know? Are we seeing a hint of liquidity in this emerging market? Using trades/asks/bids data from various sources, we construct several dynamic (as in time series) indexes/metrics as a proxy to market liquidity, some of which are pure statistics, some of which follow from analytical models. We began with a brief introduction to the events leading up to the emergence of the bandwidth trading market, from policy changes, to technological advances and the proliferation of web-based applications. A brief current state of the bandwidth trading development is also provided to examine the buildup of the trading infrastructure. This includes (1) the establishment of various exchanges, electronic trading platforms, and pooling points; (2) the involvement of bandwidth brokers and market makers; (3) the extension/creation of trading desks (from E&U’s, for example); (4) the establishment of the BTO to standardize trading contracts (among other functions); and (5) the attention from major news organizations. These events all contribute to the momentum buildup of an active bandwidth trading market. The time series metrics/statistics we construct include the number of route offerings (it has been increasing across the board) and forward price analyses to track the trend of bid/ask spread as well as price trend. We also provide geographical arbitrage analysis for the routes between U.S. and European cities, as well as those between Asian and European cities. Such arbitrage opportunities also exist between city–pairs inside Europe as well

Conclusion

49

as cities inside Asia. We observe the shrinking of such opportunities, suggesting the value of price discovery and information transparency. Finally, we introduce an analytical model to examine the bid/ask spread distribution (the likelihood of small versus large bid/ask spread) as the number of active players (both buyers and sellers) increases. The decreasing bid/ask spread trend observed in our data analysis is consistent with the scenario of increasing active participants. We can reasonably and confidently conclude that the bandwidth market is in a healthy path toward maturity (liquidity). The year 2000 was a constructive year for the establishment of the bandwidth trading infrastructure. There was every indication that 2001 and 2002 would see increased trades and more industry acceptance to trading; however, the collapse of bandwidth market value has indicated that market development will be later rather than sooner.

CHAPTER

5

New Techniques in Energy Options Robert Brooks, Ph.D., CFA President, Financial Risk Management LLC

INTRODUCTION Complex derivative contracts have been around for a very long time. Many complex financial transactions involve contingencies that can be interpreted as derivatives. Attempting to quantify the monetary value of these contingencies is relatively new. Louis Bachelier (1900) is usually credited as the first person to quantitatively value derivative securities. The seminal work of Black and Scholes (1973) helped to stimulate the modern derivatives markets. Surprisingly, energy derivatives markets are relative newcomers, with over-the-counter markets developing in the 1980s for oil and in the 1990s for natural gas and electricity. In this chapter we seek to balance the competing demands of energy risk management. Financial risk management is built on the twin pillars of valuation and management. Many theoreticians have advanced complex valuation models applied to energy options. However, these complex valuation models usually cause difficulty when one is seeking to deploy enterprise-wide risk management solutions. Hence, we seek accurate energy option valuation models that preserve the ability to solve risk management problems. Three topics related to energy options are addressed. First, the various approaches to valuation are reviewed with a focus on natural gas forward option contracts. The goal is to identify broad categories of valuation tech-

51

52

NEW TECHNIQUES IN ENERGY OPTIONS

niques and explore the benefits and costs of each category. Second, various energy option valuation techniques are covered with a particular focus on market models. The goal is to find an appropriate market model and related stochastic process to provide accurate descriptions of the complex energy markets. We illustrate some of the problems by focusing on the stochastic nature of the natural gas forward curve. Third, we reexamine the option valuation procedures in light of risk management needs. Our particular focus is on reducing the number of risk factors required to adequately model energy risk exposures.

APPROACHES TO VALUATION Various approaches to energy option valuation1 are covered and a framework for determining the correct approach for a particular energy valuation problem is provided. The focus here is on financial “value in exchange” as opposed to “value in use.” The objective is to offer three categories of approaches to valuation with a particular focus on establishing criteria for selecting the appropriate category to use for any given energy option valuation problem. We use a forward natural gas option contract to illustrate the issues. Williams (1938) and Gordon (1959) were among the pioneers in applying valuation techniques to financial assets. The capital asset pricing model (CAPM), introduced by Sharpe (1964), Lintner (1965), and others, was an early attempt to quantify the equilibrium adjustment for risk. Within the CAPM, risk was measured by the asset’s beta, and future expected cash flows were discounted at a risk-adjusted rate. We classify valuation models that discount at a risk-adjusted rate within the category of the discount factor adjusted approach. With the pioneering work of Arrow (1964), Black and Scholes (1973), Harrison and Kreps (1979), Cox and Ross (1976), and Hansen and Richard (1987), another approach to valuation has emerged that has been given a variety of names, such as state-claims valuation, equivalent-martingale valuation, stochastic discount factor valuation, or risk-neutral valuation. Within these methods of valuation, the adjustment for risk is taken in some way in the numerator of the valuation equation. For example, the typical way risk is adjusted in these methods is to adjust the probability measure. We will classify valuation models that adjust risk in this manner within the category called the cash flow adjusted approach. Cochrane (2000) demonstrates that these two general approaches to valuation can be reconciled

53

Approaches to Valuation

with each other within the state-claims framework. Before reviewing cash flow adjusted approaches and discount factor approaches, we review a much simpler approach.

Market Comparables Approach (MCA) One approach to valuation is based on the notion of comparability or substitution. Dewing (1941) expressed this approach as follows: “When several services or commodities satisfy a human want equally well, the value of each one of them is determined not by the sacrifice necessary to obtain each, but rather by the sacrifice necessary to obtain the one most easily available, which may be substituted for any one of the others.” If two investments will result in exactly the same future cash flows (same amount and timing) no matter what happens, then the appropriate values for these two investments should be the same. The two investments are assumed to produce exactly the same future cash flows regardless of the assumed underlying return distribution (presently known or unknown). We assign the label market comparables approach (MCA) to these types of methods. This approach is based on the law of one price and does not require any intermediate trading activities. The least imposing mathematical framework would involve a situation where the set of possible outcomes is not explicitly defined, that is, the circumstances for whatever reason involve future events that defy an easy mapping into a state-space. We cannot assign probabilities to future events nor even express what these future events might entail. Some may think that we are describing the spot market for electricity, concluding that within this setup it is not possible to derive a reasonable estimate of the market value of a particular derivative security. This is not true. Suppose the state-space is not well-defined, but there is a set of actively traded securities, such that s

CFi,t , j =

∑ α CF k

k=1 k≠ i

k,t , j

for all t and j

(1)

where s is the number of actively traded securities involved in replicating the cash flows (CF) for the ith security at time t for state j. Let αk denote number of units of security k, where positive implies long and negative im-

54

NEW TECHNIQUES IN ENERGY OPTIONS

plies short. That is, it is possible to replicate the cash flows for the ith security with a set of other actively traded securities. If there are no trading costs, no other market frictions, and short-selling is allowed, arbitrage activities will cause s

Pi =

∑α P

k k

k=1 k≠ i

(2)

where Pi is the market price of security i. Clearly, security i is comparable in cash flow to a portfolio of other securities. Thus, we call approaches to valuation based on employing other securities the market comparables approach. We emphasize the key assumptions using market comparable approaches are suitable when: ■ There exists a set of securities that produce future cash flows in each state identical to the security being valued (even states that are currently unimaginable). ■ Trading costs and other market frictions are minimal. ■ Short selling is allowed. The degree of confidence with the market comparables method will be directly related to the degree that these three key assumptions are reasonable. There are numerous examples of applications of the market comparables method. The value of a portfolio is merely the sum of the value of each security. Most finance and accounting theories hinge critically on this view. One can value options on natural gas forward contracts using the wellknown put-call parity for European-style forward options (no early exercise). Stoll (1994) established the relationship between puts and calls; however, this relationship was well understood as far back as Russell Sage in 1869. (See Sarnoff (1965).) Put-call parity with forward contracts states that the current price of a call option (ct) is equal to the difference between the current price of the forward contract (Ft,T) (observed at t and matures at T) and the strike price (X) discounted at the risk-free rate (r—annual compounding assumed or rc—continuous compounding) plus the current price of the put (pt). Let PV($1,T – t) denote the present value at t of a dollar at time T. Therefore, we assume

55

Approaches to Valuation

PV ($1, T − t) =

$1 (1 + r)T −t

= $1e − rc (T −t )

(3)

Both options are assumed to have the same expiration, T (where T–t is expressed in terms of fractions of a year). The forward put-call parity is ct = PV ($1, T − t)(Ft ,T − X) + pt =

Ft ,T − X (1 + r)T −t

+ pt

((4)

For put-call parity to hold, the previous three conditions must be reasonably true. If the put market is not liquid or if short-selling is not permissible, then we should not expect equation (3) to be consistently accurate in estimating the call price. One way to validate put-call parity is with a cash flow table. This is the way most arbitrageurs view this potential opportunity. Suppose you rearranged put-call parity, such that no investment was required at all ct −

Ft ,T − X (1 + r)T −t

− pt = 0

(5)

From this equation we construct a set of trades that exactly replicate these values. Specifically, +ct implies sell calls (positive cash flow means contract is sold), borrow (Ft,T < X) or lend (Ft,T > X) the discounted difference between the forward price and the strike price, and buy puts. Due to the net cash flows from these three trades, we also enter a long position in a forward contract. Table 5.1 illustrates the cash flows both today (t) and at expiration. How much should a portfolio that pays $0 for sure be worth today? No matter what discount rate you use, the present value is zero. If ??? is positive, you have a money machine or arbitrage profits. If ??? is negative in Table 5.1, then enter the opposite trades, and you have a money machine. Table 5.2 illustrates this case. Consider the following numerical example: Suppose the forward price for a one-year natural gas contract is $3.5 per MMBtu (million British thermal units), the strike price is $3.5 per MMBtu, the call option premium is $0.53 per MMBtu, the put option premium is $0.52 per

56

NEW TECHNIQUES IN ENERGY OPTIONS

TABLE 5.1 Forward Put-Call Parity Cash Flow Table Strategy Sell call Lend or borrow Buy put Net Long forward NET

Today (t) +ct –PV($1,T–t)(Ft,T – X) –pt $0 ???

At Expiration (T) FT,T < X

At Expiration (T) FT,T > X

$0 (Ft,T – X) (X – FT,T) (Ft,T – FT,T) (FT,T – Ft,T) $0

–(FT,T – X) (Ft,T – X) $0 (Ft,T – FT,T) (FT,T – Ft,T) $0

TABLE 5.2 Alternative Forward Put-Call Parity Cash Flow Table Strategy Buy call Borrow or lend Sell put Net Short forward NET

Today (t) –ct +PV($1,T–t)(Ft,T – X) +pt $0 + by assumption

At Expiration (T) FT,T < X

At Expiration (T) FT,T > X

$0 –(Ft,T – X) –(X – FT,T) –(Ft,T – FT,T) –(FT,T – Ft,T) $0

+(FT,T – X) –(Ft,T – X) $0 –(Ft,T – FT,T) –(FT,T – Ft,T) $0

MMBtu, the time to expiration is one year, and the continuously compounded interest rate is 5 percent. Because the forward price equals the strike price, in equilibrium, the call price should equal the put price. Therefore, put-call parity does not hold. Because the put price is less than the call price, we will sell the call, buy the put, and enter a long forward position. Consider Table 5.3. Thus we pocket $0.01 per MMBtu with no risk in the future. Notice that the arbitrage produces exactly the monetary difference based on the forward put-call parity equation. What makes the valuation category of market comparables approach so potent is the lack of any distributional assumptions regarding future uncertainty and how this uncertainty is priced. One security is created from trading others. We now review the cash flow adjusted approach in the context of energy options.

57

Approaches to Valuation

TABLE 5.3 Arbitrage Example with Forward Put-Call Parity Cash Flow Table Strategy

Today (t)

At Expiration (T) FT,T < X

Sell call

+ct = $0.53

Lend or borrow

–PV($1,T–t)(Ft,T – X) (Ft,T – X) –0.95238($3.5 – $3.5) ($3.5 – $3.5) = $0 = $0 –pt = $0.52 (X – FT,T) = ($3.5 – FT,T) (Ft,T – FT,T) = ($3.5 – FT,T) $0 (FT,T – Ft,T) = (FT,T – $3.5) +$0.01 $0

Buy put Net Long forward NET

$0

At Expiration (T) FT,T > X –(FT,T – X) = –(FT,T – $3.5) (Ft,T – X) ($3.5 – $3.5) = $0 $0 (Ft,T – FT,T) = ($3.5 – FT,T) (FT,T – Ft,T) = (FT,T – $3.5) $0

Cash Flow Adjusted Approach (CFAA) A second approach to valuation is also founded on the notion of comparability but requires active trading based on the principle of self-financing and dynamic replication. The seminal works of Black and Scholes (1973) and Merton (1973) are based on the idea of synthetically creating the cash flows of a risk-free bond from dynamically trading a stock and a call option on that stock. Although there have been a multitude of research papers written using this type of procedure, there is one common thread. The future cash flows can be discounted at the risk-free rate once either the cash flows or the probability distribution has been adjusted. These valuation methods are often referred to as risk-neutral valuation, because the discount rate is the risk-free rate. We refer to these adjusted probabilities as equivalent-martingale measures because the probabilities have been adjusted so that the stochastic process follows a martingale, after adjusting for the time value of money. Also the probability space of the original probabilities is equivalent to the probability space after the adjustment. We assign the label cash flow adjusted approach (CFAA) to emphasize that these types of methods require some adjustment to the numerator of the valuation equation. It is interesting to point out that this category of approach is only viable if the related

58

NEW TECHNIQUES IN ENERGY OPTIONS

securities have a high level of marketability. The CFAA approach is the category most dependent on marketability to be viable. When it is not possible to synthetically create the cash flows from existing securities without any assumptions about the state-space, it may be possible to synthetically create a particular security’s cash flows when there is sufficient structure assumed about the state-space. As we will observe, either the cash flow for state j or the probability of observing state j will be adjusted to account for risk. This structure has taken many different forms depending on the valuation needs. Underlying each of the valuation techniques classified under the CFAA is the ability to derive state-claims for all possible states in the sample space. A state-claim is the current price of receiving one unit ($1) at time t only if a particular outcome in the state-space occurs (state j) and zero units ($0) otherwise. Assuming the state-space is well-defined and enough structure exists to obtain state-claims, then the price of the ith security can be expressed as T

Pi =

m

∑ ∑ SC

t , jCFi,t , j

(6)

t =1 j =1

where SC denotes state claims and CF denotes cash flow. It can be demonstrated (see Cochrane (2000) for example) that the state-claim is equal to the discounted equivalent-martingale measure or SCt,j = PV($1,t) qt,j for all t and j

(7)

where r is assumed to be the appropriate continuously compounded riskfree rate and q denotes the equivalent-martingale measure. Substituting for this definition of a state-claim and factoring out the discount function yields T

Pi =

∑ t =1

m

∑

PV ($1, t)

j =1

T

qt , jCFi,t , j =

∑ PV ($1, t)E [CF ] q

i ,t

(8)

t =1

The current market price of security i is the discounted future expected cash flow based on equivalent-martingale measures and the discounting is at the risk-free interest rate.

59

Approaches to Valuation

The key assumptions to reasonably use the CFAA are: ■ There exists a stochastic process (or processes) that accurately depicts the future potential outcomes; that is, the state-space is well-defined. ■ There exists a trading strategy that produces future cash flows in each state identical to the security being valued. ■ Trading costs and other market frictions are minimal. ■ Short-selling is allowed. Cash flow adjusted approach to valuation is built on the ability to construct reliable dynamic hedges. The famous Black–Scholes (1973) option pricing model is based on the assumption that a dynamic strategy can be designed using call options and the underlying stock to simulate a risk-free payoff in the future. Many derivative valuation models are built on the CFAA. The essence of this approach is to alter the probability distribution of future cash flows as to achieve a risk-free rate of return. As such, this approach is often referred to as an adjusted probability measure. The CFAA is illustrated using options on natural gas forward contracts. A single period binomial framework is assumed with no market frictions of any kind and that a riskless asset exists. The binomial model assumes either an up state (u) or a down state (d).

F1,1 = Fu = uF0 F0 Node 0 or 0,0

Arc

1,1

F1,0 = Fd = dF0 1,0

| t=0

| T=1

60

NEW TECHNIQUES IN ENERGY OPTIONS

In this single period model, there are three nodes (states) and two arcs (paths). The following market data are assumed S0 = $31/3 per MMBtu (spot price of natural gas observed at t) Ft,T = $3.50 per MMBtu (forward price of natural gas, observed at t, expiring at T) X = $3.50 per MMBtu (strike price) r = 5% (annual compounded riskless rate) T – t = 1 year (time to expiration of forward contract) σ = 40% (standard deviation of continuously compounded, annualized percentage price changes of forward contract) Now several intermediate parameters are calculated. Remember, the objective is to value the call option. The price relative of the forward contracts when the up and down states occur (consistent with the standard option valuation assumptions) as well as the equivalent-martingale probabilities (q) are calculated u=

{

}

{

}

Fu = exp σ T − t = exp 0.40 1 = 1.491825 Ft

(9)

(forward price relative if the up event occurs) d=

Fd 1 1 = = = 0.670320 Ft u 1.491825

(10)

(forward price relative if the down event occurs) qu =

1− d 1 − 0.670320 = = 40.13123% u − d 1.491825 − 0.670320

(11)

(equivalent-martingale probability if the up event occurs) qd =

u −1 = 1 − qu = 1 − 0.4013123 = 59.86877% % u−d

(equivalent-martingale probability if the down event occurs) Note that the expected value of the forward price relatives is one.

(12)

61

Approaches to Valuation

F  Eq  T  = quu + qd d = 0.4013123 (1.491825)  Ft  + 0.5986877 (0.670320) = 1.0

(13)

Hence, q is an equivalent-martingale measure. Based on these parameters, the binomial tree is F1,1 = Fu = uF0 = $5.2213 1,1 F0 = 3.5 0 F1,0 = Fd = dF0 = $2.3461 1,0 | t=0

| T=1

Therefore, the binomial tree for the call option is C1,1 = Cu = max[0, Fu – X] = $1.72139 1,1 C0 = ? 0 1,0 C1,0 = Cd = max[0, Fd – X] = $0.0 | t=0

| T=1

62

NEW TECHNIQUES IN ENERGY OPTIONS

To find the value of the call option, an additional parameter is needed. The call option delta (∆c) measures the sensitivity of option prices to changes in the underlying forward price. ∆C =

Cu − Cd $1.72139 − $0 = = 0.598688 $5.22139 − $2.34612 Fu − Fd

(14)

Consider the unusual trading strategy of buying 1/∆C call options, going short one forward contract, and borrowing the following amount (B*) B* =

Ft − dFt $3.5 − (0.670320) $3.5 = = $1.09893 1+ r 1 + 0.05

(15)

Hence, the portfolio (Πt) is valued at time t as (remember the cost of entering a forward contract is zero) Πt =

1 1 F − dFt 1 Ct − B* = Ct − t = Ct − $1.09893 ∆C ∆C 1+ r ∆C

(16)

The values of this portfolio for the up and the down states are ΠT ,u = =

ΠT ,d = =

1 CT , u + (Ft − uFt ) − B *(1 + r) ∆C

(up state) (17) 1 $1.72139 + ($3.50 − $5.22139) − $1.09893 0.598688 (1 + 0.05) = $0 1 CT , d + (Ft − dFt ) − B *(1 + r) ∆C (down state) (18) 1 $0.0 + ($3.50 − $2.34612) − $1 .09893 0.598688 (1 + 0.05) = $0

Due to the zero future portfolio value, the value of the portfolio at t should also be zero. Therefore, the option price is

63

Approaches to Valuation

Πt =

1 1 F − dFt = $0 Ct − B* = Ct − t 1+ r ∆C ∆C

(19)

F − dFt = $0.657918 Ct = ∆ C t 1+ r

Equation (19) is referred to as the no arbitrage method of valuing the option. There are two other perspectives that yield the same result. The equivalent-martingale method takes the expected future call value and discounts it at the riskless rate. 1 1 [quCu + qdCd ] Eq [CT ] = 1+ r 1+ r (20) 1 [0.4013123 ($1.72139) + 0.5986877($0)] = $0.657918 = 1 + 0.05

Ct =

Alternatively, the state-claim method above can be deployed. Here the state-claim values for up and down states are SCT , u = SCT , d =

1 1 qu = 0.4013123 = $0.382202 (up state) 1+ r 1 + 0.05

1 1 qd = 0.5986877 = $0.570179 1+ r 1 + 0.05

(21)

(down state) (22)

Therefore, the value of this call option is Ct = SCT,uCT,u + SCT,dCT,d = 0.382202($1.72139) + 0.570179($0) = $0.657918

(23)

It is possible to demonstrate that these valuation procedures can be generalized to a multiperiod setting. However, in the multiperiod setting intermediate trading is required to dynamically replicate the option payoffs (called a self-financing, dynamic replicating strategy). Clearly, in order for these valuation methods to yield reasonable results, the ability to actively trade the underlying asset (forward contract in this example) is required. The final category of valuation approaches, generically called the discount factor adjusted approach, is now covered.

64

NEW TECHNIQUES IN ENERGY OPTIONS

Discount Factor Adjusted Approach (DFAA) The traditional approach to valuation is to forecast some future expected cash flows and then to take the present value of this future expected cash flow stream. John Burr Williams (1938) is usually credited with first articulating this procedure for common stocks. Williams states “The investment value of a stock [is] the present worth of all the dividends to be paid upon it adjusted for expected changes in the purchasing power of money.” Interestingly, Williams goes on to argue “that neither marketability nor stability should be permitted to enter into the meaning of the term investment value.” (See Ellis [1989], pp. 153 and 156.) In 1959, Gordon, when introducing the now famous dividend discount model bearing his name, argued that the appropriate discount rate increases with the degree of uncertainty related to the future dividend stream. Hence the “stability” of Williams does influence market value. From this foundational paper, a vast literature has developed extending and testing various aspects of this approach to valuation. We assign the label discount factor adjusted approach to these types of methods. The identifying criterion for a valuation method to fall in the DFAA category is that the adjustment for risk is made in the denominator of the valuation equation. The higher the risk (however defined), the higher the interest rate will be for discounting. When the nature of existing securities and/or the structure of the statespace does not afford the ability to derive state-claims, then the valuation method typically adjusts for risk in the denominator by assuming a specific risk premium. This approach is the least favored due to the difficulty in accurately estimating required inputs and the resulting prices’ sensitivity to these estimated inputs. The discount factor adjusted method does not alter the cash flow probability distribution; rather, the risk adjustment is taken in the interest rate at which the cash flows are discounted through time. There must be sufficient structure imposed upon the state-space to compute at least the expected future cash flows and the appropriate risk premium.

T

Pi =

m

1

∑ ∑ (1 + r + RP t =1 j =1

t

t i ,t , j )

pt , jCFi,t , j

(24)

Approaches to Valuation

65

where pt,j denotes the subjective probability based on a particular individual’s perspective on future cash flows. Also, the size of the risk premium is a function of the compounding method. When sufficient structure exists to use the CFAA to achieve valuation, using the DFAA requires a direct mapping between the risk premium and the assigned probabilities for future states; otherwise, multiple values for the same security are obtained. In some sense, such one-to-one mapping does not always hold, due to the vast amount of trading that occurs daily. Obviously, when a trader’s probability beliefs and risk premium result in valuations sufficiently different from market prices, trading will occur. A simple example of the DFAA is the standard Gordon growth model for valuing common stocks, P0 = D1/(k – g), where k is the cost of equity capital or the investor’s required rate of return. The typical way the investor’s required return is estimated is by using the risk-free rate plus a risk premium (for example, CAPM k = r + βi(E(rm) – r)). Other examples of this approach are valuing mortgage backed securities with the option adjusted spread. These methods are extremely sensitive to parameter estimation error and are hard to externally verify. Because the DFAAs are used widely in practice, one would conclude that there is currently insufficient structure in some markets to apply either the MCA or CFAA. The DFAA is placed within the CFAA using the binomial framework. Consider again the simple one period binomial framework in the previous section. The difference here is that each investor will impose their own subjective beliefs about the probability of the up and down states. For example, suppose an investor believed that the probability of an up event was 43 percent (as opposed to the equivalent-martingale probability of 40.13123% identified earlier). Now we have two issues to address. First, what is the appropriate risk premium? Second, what is the appropriate value for the call option? Consider a constant risk premium of 7.5061 percent.

1 1 [ puCu + pdCd ] Ep [CT ] = 1 + r + RP 1 + r + RP 1 [0.43 ($1.72139) + 0.57($0)] = $0.657918 = 1 + 0.05 + 0.075061

Ct =

(25)

66

NEW TECHNIQUES IN ENERGY OPTIONS

which is the same result as CFAA methods. Clearly, they are the same by selecting the appropriate risk premium. Alternatively, we can solve for the implied risk premium.  Ep [CT ]  RP =    Ct 

1 / T −t

 0.740198  − (1 + r) =    0.657918 

1/ 1

− (1 + 0.05) = 0.075061 (26)

By combining the CFAA and DFAA approaches, interesting information can be gleaned from derivatives market values. The CFAA approach can be used to establish the appropriate volatility (or binomial tree), and the DFAA approach can be used with an investor’s view to determine the implied risk premium. The implied risk premium is a useful measure for assessing hedging and speculative trading activities.

Selecting the Best Approach to Valuation Three categories of valuation methodologies encompass virtually all methods of valuation: market comparables approach (MCA), cash flow adjusted approach (CFAA), or discount factor adjusted approach (DFAA). From a confidence perspective, market comparables is the best, followed by the cash flow adjusted method. Only as a last resort does one wish to go with a discount factor adjusted method. However, within energy markets considering the DFAA is reasonable due to lack of liquidity or other trading problems. Table 5.4 summarizes the major assumptions and their importance within the various approaches to valuation. For MCA, the existence of a

TABLE 5.4 Major Assumptions of the Three Approaches to Valuation Assumptions

MCA

CFAA

DFAA

Short selling allowed with full use of proceeds Trading cost minimal Set of securities exist to replicate payoffs Stochastic process to model risk variable Trading strategy exist to replicate payoffs Explicit risk adjustment

Strong Weak Strong NR NR NR

Strong Strong NR Strong Strong NR

NR* NR NR Weak NR Strong

*Not relevant

67

Energy Option Valuation Models

set of securities that exactly replicate the future payoffs of a particular security and short selling are the critical assumptions. Is a public utility willing to short power in July? For CFAA, there are several assumptions that are critical; however, we no longer need the existence of a replicating set of securities. Finally, the critical assumption of DFAA is the ability to explicitly adjust for risk when discounting the future expected cash flows. We turn now to review several standard option valuation models, illustrating the various approaches to valuing energy options.

ENERGY OPTION VALUATION MODELS Option Valuation Framework The following are key underlying assumptions for deriving option valuation models. ■ There exists a stochastic process (or processes) that accurately depicts the future potential outcomes, that is, the state-space is well defined. Specifically, we start by assuming that the underlying variable (for example, derivative contracts or underlying assets) is assumed to follow geometric Brownian motion. dS = µSdt = σSdz

(27)

■ There exists a trading strategy that produces future cash flows in each state identical to the derivative security being valued. ■ Trading costs and other market frictions do not exist. ■ Short selling is allowed. ■ There are no storage costs, cash flows (like dividends), or convenience yields related to owning the underlying index. If the option can be reproduced by trading the underlying variable and the risk-free bond, then we can assume the underlying asset grows at the cost of carrying it (here, only the financing costs or the risk-free interest rate) and hence the valuation procedure falls in the CFAA. Therefore, the expected future stock price can be expressed as Eq[ST] = Ster(T–t)

(28)

68

NEW TECHNIQUES IN ENERGY OPTIONS

or St = Eq[ST] e–r(T–t)

(29)

Let us consider the value of a call option whose payoff at expiration can be expressed as cT = max[0,ST – X] As illustrated in equation (20), the call value at t (t
(30)

Taking the expectation in equation (30) results in the standard Black– Scholes option pricing model (BSOPM), which can be expressed as ct = e–r(T–t)[Eq[ST]N(d1,q) – XN(–d2,q)]

(31)

pt = e–r(T–t)[XN(–d2,q) – Eq[ST]N(–d1,q)]

(32)

where

d1, q =

 Eq [ST ]  2 ln  + (0.5σ )(T − t) X   σ T −t

=

S  ln t  + (r + 0.5σ 2 )(T − t)  X

d2, q = d1, q − σ T − t

(33)

σ T −t (34)

St is the asset price at time t X is the strike price r is the annualized, continuously compounded risk-free interest rate, T is the calendar expiration, expressed in years, hence T–t is the time to maturity, σ is the annualized, continuously compounded standard deviation of percentage changes in the stock price (known as volatility), and

69

Energy Option Valuation Models

d

N(d) =

∫

e−x

2

/2

dx

(35)

2π

−∞

which is the area under the standard cumulative normal distribution up to d (see the following discussion). Note that N(–d) = 1 – N(d) because the normal distribution is symmetric. See Brooks (2000) for an accurate approximation for solving N(d) using C++. We could also assume that the underlying asset grows at a risk-adjusted rate and hence the valuation procedure falls in the DFAA. That is, Ep{ST} = Ste(r+RP)(T–t)

(36)

and thus the option valuation models are ct = e–(r+RP)(T–t)[Ep[ST]N(d1, p) – XN(d2, p)]

(37)

pt = e–(r+RP)(T–t)[XN(–d2, p) – Ep[ST]N(–d1, p)]

(38)

where

d1, p =

 Ep [ST ] 2 ln  + (0.5σ ) (T − t) X   σ T −t

=

S  ln t  + (r + RP + 0.5σ 2 ) (T − t)  X (39)

d2, p = d1, p − σ T − t

σ T −t (40)

For energy commodities, it is difficult to estimate the appropriate risk premium, RP. It is common to model the underlying asset incorporating storage costs and convenience yield. Practically, the convenience yield is a number that makes the model work in the same way as implied volatility and option valuation models. Let u denote the continuously compounded storage costs on the underlying instrument and let y denote the convenience yield. Within this framework we have the expected future value of the underlying instrument equal to the forward price (observed at t that expires at T):

70

NEW TECHNIQUES IN ENERGY OPTIONS

Ft,T = Eq[ST] = Ste(r+u–y)(T–t)

(41)

The expected growth of the underlying instrument is the risk-free rate plus the cost of storing it less the benefit from owning the underlying, the convenience yield. Substituting this result into the option pricing model preceding, we have

[

]

ctF = e − r (T −t ) Ft N(d1F, q ) − XN(d2F, q )

[

(42)

]

ptF = e − r (T −t ) XN(−d2F, q ) − Ft N(−d1F, q )

(43)

where

d1F, q =

F  ln t  + (.5σ 2 ) (T − t)  X

(44)

σ T −t

d2F, q = d1F, q − σ T − t

(45)

These option valuation models hinge on the assumed stochastic process generating the underlying instrument, geometric Brownian motion. With different stochastic processes, you will have different option valuation models. There are many stochastic processes worth considering. Often these models are termed “market models,” because we are seeking the best representation of the stochastic nature of the underlying instrument. We carefully examine one potential market model.

LSC Market Model 2 Identifying the family of stochastic processes that best depict the joint relationship between various maturity forward contracts as well as different markets is sought. We introduce the LSC model, denoting level, slope, and curvatures of changes in forward prices. Several risk management issues related to energy options are illustrated. The main problem is adequately modeling the stochastic behavior of forward curves. Energy option valuation models take the forward curve as given and incorporate

71

Energy Option Valuation Models

information about the appropriate volatility. The natural gas forward curve comprises many forward contracts all tied to natural gas, unique only in timing of delivery. Forward prices exhibit clear seasonality as seen in Figure 5.1. Any effort to fit the natural gas forward curve with a small number of parameters would be extremely difficult and will result in estimation error when valuing a portfolio of simple forward contracts. When implementing risk systems, it is important that the models correctly value the portfolio initially. That is, you must have accurate mark-to-market valuations. Notice, however, from Figure 5.2 that the behavior of forward price changes is relatively smooth. Forward price changes capture the embedded risk of forward contract portfolios and are influenced by a relatively small number of factors based on principal components analysis. Our objective is to select a market model to capture the stochastic behavior of a portfolio of natural gas forward and forward option contracts. One approach would be to assume that each forward contract is a separate risk variable and estimate a relatively large correlation matrix. From the analysis above, forward prices are highly correlated. Alternatively, assume that, say, the first four principal

$2.8

Dollars per MMBtu

$2.7

$2.6

12/3/99 12/2/99

$2.5

$2.4

$2.3

$2.2 0

0.5

1

1.5 Maturity in Years

2

FIGURE 5.1 Natural Gas Forward Prices

2.5

3

72

NEW TECHNIQUES IN ENERGY OPTIONS

$0.00

–$0.02

Dollars per MMBtu

–$0.04

–$0.06

–$0.08

–$0.10

–$0.12

–$0.14 0.0

0.5

1.0

1.5

2.0

2.5

3.0

Maturity in Years

FIGURE 5.2 Natural Gas Forward Prices First Differences from December 2, 1999 to December 3, 1999

components are the risk variables. Unfortunately, the principal components lack economic interpretation. A balance is sought between a small number of risk factors and economic interpretation. Economic interpretation is important because we seek to not only measure risk but also to actively manage risk. Although there are a variety of methods for estimating the stochastic behavior of forward curves, we illustrate a modified version of Svensson’s model (1995), which we call the LSC model (denoting level, slope, and curvatures). The LSC model is originally based on research by Nelson and Siegel (1987) related to the term structure of interest rates. The LSC model has been applied in the bond market by Willner (1996) as well as Brooks and Yan (1999). The LSC model is a linear, n-factor model where the factors have economic interpretations. The LSC model applied to natural gas forward contracts can be expressed as follows

(

)

∆f˜ m j ; b, t = ∆f˜j,t =

N fac −1

∑ b˜

f i ,t C i , j

i =0

(m ; τ ) + ε˜ j

f i −1

j ,t

(46)

73

Energy Option Valuation Models

where

~ = the variable is stochastic across calendar time mj = maturity of the forward contract j, expressed as a fraction of a year Nfac = the number of factors associated with changes in the forward curve (notice the counter starts at zero in the equation above) f = a constant term associated with the maturity τi–1 coefficients Ci,j defined below ∆fj,t = the change in the mj – maturity forward contract price between t and t + ∆t (= fj,t+1 – fj,t)3 bi,tf = the stochastic sensitivity coefficient associated with the maturity coefficients Ci,j observed from t to t + ∆t and f Ci,j(mi;τi–1 ) = the maturity coefficients defined as follows (where τ0f = τ1f ) C0, j(mj ;τ–1f ) = C0(mj) = 1

(47)

 m j  C1, j m j ; τ0f = C1, j m j ; τ1f = exp − f   τ1 

(

(

)

)

Ci, j m j ; τ if−1 =

(

mj τ if−1

)

 m j  exp− f ; i = 2,K, N fac − 1  τ i −1 

(48)

(49)

εj = the residual error remaining after accounting for the maturity coefficients. The LSC model is linear in the coefficients and hence ordinary least squares regression can be applied. The change in the forward price as m goes ∼ ∼f ∼f to infinity is ∆f(mj→∞;b,t) = b0,t. Hence, we call b0,t the “level” of change in the ∼ forward curve. The forward price as m goes to zero is ∆f(mj→0;b,t) = ∼f ∼f ∼f b0,t + b1,t. Hence, we call b1,t the “slope” of change in the forward curve. Notice if the slope is positive, then longer maturity forward price changes are lower ∼f than the short-term forward price changes. The other terms, b1,t,i=2, . . . , Nfac –1, measure the “curvature” of the changes in forward prices. This framework assumes that the maturity of the forward contract does not change over ∆t. When a contract expires and the nth contract becomes the (n–1)th contract, an adjustment is necessary. However, the focus is on modeling the stochastic behavior over the next day or some relatively short period of time. Figure 5.3 illustrates the sensitivity of the maturity coefficients for the four parameter LSC model with τ 1f = 0.25 and τ 2f = 0.75. The maturity coefficients attach greater weight to observations at different maturities. Applying the LSC model to energy forward curve, we have introduced estimation error into the risk management system. Figure 5.4 illustrates the

74

NEW TECHNIQUES IN ENERGY OPTIONS

1.2

Regression Variables

1.0

0.8 Level Slope Curve 1 Curve 2

0.6

0.4

0.2

0.0 0

0.5

1

1.5

2

2.5

3

3.5

Maturity in Years

FIGURE 5.3 Four Parameter LSC Model Variable (τ 1f = 0.25 and τ 2f = 0.75)

$0.00

—$0.02

Dollars per MMBtu

—$0.04

—$0.06

—$0.08

—$0.10

—$0.12

—$0.14 0.0

0.5

1.0

1.5 Maturity in Years

2.0

2.5

FIGURE 5.4 Actual Change in Forward Curve with 3-Parameter LSC Model on December 3, 1999

3.0

75

Energy Option Valuation Models

fit of the three-parameter model illustrated in Figure 5.3 with December 3, 1999 data. Although the fit contains error, it does not appear severe. Of course, severity depends on your perspective. Based on the two-parameter LSC model applied to a historical data set, the average cross-sectional R-square was 0.6753 suggesting that a lot of variability was left unexplained. The R-square increases with the third parameter to 0.8013 and with the fourth parameter to 0.8706. Figures 5.5–5.8 exhibit the cumulative distribution function (CDF) based on actual data as well as based on the LSC model with from one to four parameters for the most volatile nearby contract. Clearly, one- and twoparameter LSC models are inadequate to estimate the CDF and hence risk measures such as value-at-risk (VaR). For example, the VaR at the 95 percent confidence level is actually $0.128/MMBtu, whereas the estimated VaR with the two-parameter LSC model is $0.109/MMBtu. The VaR estimation error is approximately 15 percent of the actual VaR. However, it is unclear whether the addition of parameter four justifies the added computational cost. For example, the VaR with the three-parameter LSC model is $0.122/MMBtu and the VaR with the four-parameter LSC model is $0.129/MMBtu. The VaR estimation error for the three-parameter LSC model is approximately 5 percent 1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65

CDF

0.60

Actual Predicted

0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 —0.5

—0.4

—0.3

—0.2

—0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

$/MMBtu

FIGURE 5.5 Actual Cumulative Distribution Function and Estimated Based on 1-Parameter LSC Model

76

NEW TECHNIQUES IN ENERGY OPTIONS

1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65

CDF

0.60

Actual Predicted

0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 –0.5

–0.4

–0.3

–0.2

–0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

$/MMBtu

FIGURE 5.6 Actual Cumulative Distribution Function and Estimated Based on 2-Parameter LSC Model

1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65

CDF

0.60

Actual Predicted

0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 –0.5

–0.4

–0.3

–0.2

–0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

$/MMBtu

FIGURE 5.7 Actual Cumulative Distribution Function and Estimated Based on 3-Parameter LSC Model

77

Energy Option Valuation Models

1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65

CDF

0.60

Actual Predicted

0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 –0.5

–0.4

–0.3

–0.2

–0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

$/MMBtu

FIGURE 5.8 Actual Cumulative Distribution Function and Estimated Based on 4-Parameter LSC Model

of the actual VaR, whereas the VaR estimation error for the four-parameter LSC model is approximately 1 percent of the actual VaR. Although a threeparameter LSC model appears adequate, the LSC model provides the flexibility to achieve higher levels of precision by adding curvature terms.

Alternative Stochastic Processes LSC Model The stochastic processes reviewed here start from basic multivariate normal and move to an increasingly higher level of complexity. It remains primarily an empirical decision as to the appropriate model to ultimately adopt. We start with a simple example of applying the LSC model to changes in the power forward curve. Suppose we are long 4 forward contracts with maturity m1 = 1 month and short 3 forward contracts with maturity m1 = 5 years. For illustration purposes only, suppose the forward curve at time t = 0 could be represented as fm ,0 = g(mj, . . .) j

(50)

78

NEW TECHNIQUES IN ENERGY OPTIONS

where g( ) represents a generic power forward curve function as complicated as necessary. The forward curve at time t = 1/52 could be represented as (we assume no estimation error for illustrative purposes) f˜m j ,1 / 52 = g(m j ,K) + ∆ f˜m j ,0

(51)

where we assume

(

~ ∆f m j ,0 =

)

(

)

b˜0f ,0 + b˜1f,0C1 m j ; τ1f = 0.5 + b˜ 2f ,0C2 m j ; τ1f = 0.5

(52)

Figure 5.9 illustrates two power forward curves one week apart. The power forward curve is assumed to have changed, based on the following functional form: ~ ∆fm j ,0 = $2.0 − $10C1(m j ; τ1f = 0.5) + $10C2 (m j ; τ1f = 0.5) In this case, the one-month contract will change

$70.00 $60.00

$ per MWh

$50.00 $40.00 $30.00 $20.00 f(t=0) f(t=1/52) df

$10.00 $0.00 –$10.00 0

0.5

1

1.5

2

2.5 Maturity

3

3.5

4

4.5

FIGURE 5.9 Example of Forward Curve for Power

5

Energy Option Valuation Models

~ ∆f m

j

=1 / 12 ,0

79

= $2.0 − $10C1(m j = 1 / 12 ; τ1f = 0.5) + $10C2 (m j = 1 / 12 ; τ1f = 0.5) = $2.0 − $10(0.84648) + $10(0.14108) = −$5.054 / MWh

and the five-year contract will change ~ ∆fm j = 5,0 = $2.0 − $10C1(m j = 5; τ1f = 0.5) + $10C2 (m j = 5; τ1f = 0.5) = $2.0 − $10(0.00005) + $10(0.00045) = $2 / MWh Hence, if we were long 4 one-month contracts and short 3 five-year contracts, our portfolio gain or loss could be represented as  4 N =  2 x1 −3 −$5.054 ∆F =   2 x1  $2  −$5.054 ∆Π = N' ∆F = [4 − 3]   = −$26.216 / MWh 1x 2 2 x1  $2  In the following sections, several candidate market models are offered for the stochastic behavior for forward curves.

Multivariate Normal Model Assuming the LSC model coefficients follow a multivariate normal distribution ~ ~ bif,t = µ i (t)dt + σ i (t)dz i

(53 )

where µi(t) = the expected change in the stochastic sensitivity coefficient (bi,tf ) over time period dt and may change over time (for example, to incorporate seasonality) σi(t) = the standard deviation of the stochastic sensitivity coefficient

80

NEW TECHNIQUES IN ENERGY OPTIONS

(bi,tf ) over time period dt and may change over time (for example, to incorporate seasonality), and d~ zi = the standard Gauss–Weiner process (mean zero, standard ∼ ∼ deviation dt, or d ∼ zi = dtεi , where εi ∼ N(0,1), and ρdz ,dz i j denotes the correlation between stochastic sensitivity coefficients). Substituting equation (53) into equation (52), we have N fac −1

∆fm j ≈

∑

bif,tCi

i =0 N fac −1

=

(

N fac −1

m j ; τ if−1

) = ∑ {µ (t) dt + σ (t) dz } C (m ; τ i

i

i

i

f i −1)

i =0 N fac −1

∑ µ (t)C (m ; τ )dt + ∑ σ (t)C (m ; τ )dz i

j

i

f i −1

j

i

i =0

i

f i −1

j

(54)

i

i =0

=d µˆ j(t)dt + σˆ j(t)dz j where =d denotes equal in distribution N fac −1

µˆ j (t) =

∑ µ (t)C (m ; τ ) i

i

j

f i −1

(55)

i =0

and N fac −1 N fac −1 2

σˆ j (t) = =

∑ ∑ σ (t)C (m ; τ )σ (t)C (m ; τ )ρ i

k= 0 i =0 N fac −1 σ i (t)2 Ci i =0

∑

(

i

(

m j ; τ if−1

)

f i −1

j

)

2

k

k

j

f i −1

dzi , dz j

N fac −1 N fac −1

+2

∑ ∑ σ (t)C (m ; τ ) i

i =0

k= i +1

i

j

f i −1

(56)

σ k (t)Ck m j ; τ if−1 ρdzi ,dz j With Nfac stochastic sensitivity coefficients, we can generate Nc distributions (denoting different maturity forward contracts and forward option contracts), where Nc >> Nfac (much greater). Within this framework, changes in forward prices are normally distributed. As m → ∞, from equation (54) we have

81

Energy Option Valuation Models

∆f(m∞;b,t) = ∆fm = b0,f t = µ0(t)dt + σ0(t)dz0

(57)

∞

and when m → 0, from equation (54) we have ∆f(m0;b,t) = ∆fm = b0,f t + b1,f t = {µ0(t) + µ1(t)}dt + σ0(t)dz0 + σ1(t)dz1 0 (58) =d {µ0(t) + µ1(t)}dt + [σ0(t)2 + σ1(t)2 + 2σ0(t)σ1(t)ρdz ,dz ]1/2 dzj=0 0

1

One hypothesis for the behavior of the forward curve can be phrased as “today’s forward curve is our best proxy for tomorrow’s forward curve.” We set aside for the moment seasonality effects. That is, over the next short time horizon, the nth maturity contract is the same contract at the beginning and ending of the horizon (no contracts expire over the short time horizon). Under this hypothesis, the expected change in forward prices is zero. Therefore, µi(t) = 0 for all i and all t From equation (54), we have N fac −1

∆fm j =

∑ σ (t)C (m ; τ )dz i

i

j

f i −1

i

d = σˆ j (t)dz j

i =0

Once again we consider the portfolio long 4 forward contracts with maturity m1 = 1 month and short 3 forward contracts with maturity m1 = 5 years. Suppose we have the following parameters (Γ denotes the correlation matrix): σ0(t) = 20% σ1(t) = 50% σ2(t) = 30%

From equation (56) we have

 1.0  Γ = −0.5 −0.8 

− 0.5 1.0 0.0

− 0.8  0.0 1.0

82

NEW TECHNIQUES IN ENERGY OPTIONS

(

2

σˆ j =

∑

σ i (t)2 Ci m j ; τ if−1

i =0

) ∑ ∑ σ (t)C (m ; τ )σ (t)C (m ; τ ) 2

2

2

+2

i

i =0

i

f i −1

j

k

k

j

f i −1

k= i +1

ρdzi ,dz j

( ) + 0.3 C (m ; τ ) 0.2(1)0.5C m ; τ −0.5 + 0.2(1)0.3C m ; τ −0.8 + 0.5C  ( )( ) ( )( )   + 2   (m ; τ ) 0.3C (m ; τ ) (0.0)   = 0.04 + 0.25C (m ; τ ) + 0.09C (m ; τ ) − 0.1C (m ; τ ) − 0.096C (m ; τ ) 2

= 0.2212 + 0.52 Ci m j ; τ1f 1

j

2

1

j

j

f 1

j

2

f 1

j

f 1

2

2

j

f 1

2

f 1

j

1

f 1

2

2

j

f 1

2

1

j

f 1

2

f 1

The standard deviation of the one-month contract is σ^1(t)2 = 0.04 + 0.25(0.84648)2 + 0.09(0.14108)2 – 0.1(0.84648) – 0.096(0.14108) = 0.12273 σ^1(t) = 0.350332 and the standard deviation of the five-year contract is σ^1(t)2 = 0.04 + 0.25(0.00005)2 + 0.09(0.00045)2 – 0.1(0.00005) – 0.096(0.00045) = 0.03995 ^ σ (t) = 0.19988 2 Thus we observe the higher volatility at the short end of the forward curve and the five-year forward rates behave almost identically to the perpetual forward rate, σ0(t) = 0.2.

Mean-Reverting, Multivariate Normal Process Mean reversion can be incorporated by assuming ^

b tf = pi [µi – bif ]dt + σi(t)dzi

(59)

83

Energy Option Valuation Models

where µi = the long-run expected value of bi,tf (usually zero), pi = the pull parameter that regulates how fast the stochastic coefficient, bi,tf , is drawn back to µi Substituting equation (59) into equation (52), we have N fac −1

∆fm j ≈

∑ b C (m ; τ ) f i ,t

i

i =0 N fac −1

=

∑ {p [µ i

i =0 N fac −1

=

∑

f i −1

j

i

} (

]

− bˆ if dt + σ i (t)dz i Ci m j ; τ if−1

[

] (

)

pi µ i − bˆ if Ci m j ; τ if−1 dt +

i =0

)

N fac −1

∑ σ (t)C (m ; τ )dz i

i

j

f i −1

i

(60)

i =0

d = µˆ j (t)dt + σˆ j (t)dz j

where N fac −1

∑ p [µ

µˆ j (t) =

i

i

] (

− bˆ if Ci m j ; τ if−1

i =0

)

(61)

N fac −1 N fac −1

σˆ j (t)2 =

∑ ∑ σ (t)C (m ; τ )σ (t)C (m ; τ )ρ i

i

j

i =0 k= 0 N fac −1

=

∑ σ (t) ( i

i =0

(

2

Ci m j ; τ if−1

)

)

f i −1

k

2

N fac −1 N fac −1

+2

k

j

f i −1

dzi , dzk

∑ ∑ σ (t)C (m ; τ ) i

i =0

i

j

k= i +1

f i −1

(62)

σ k (t) m j ; τ if−1 ρdzi ,dzk Thus we see that within this framework, changes in forward prices can be represented as normally distributed. As m → ∞, from equation (60) we have ^

∆f(m∞;b;t) = ∆fm∞ = b0,tf = p0[µ0 – b0f ]dt + σ0(t)dz0 and when m → 0, from equation (60) we have

(63)

84

NEW TECHNIQUES IN ENERGY OPTIONS

∆f (m0 ; b, t) = ∆fm j = b0f ,t + b1f,t

{[

]}

] [

(64) = p0 µ 0 − bˆ 0f + p1 µ1 − bˆ1f dt + σ 0 (t)dz0 + σ1(t)dz1 ( Within this framework, changes in forward prices can be represented as normally distributed with mean reversion incorporated in the drift term.

Mean-Reverting, Multivariate Normal with Poisson Jump Process Jumps can also be incorporated into the stochastic process. Clearly, markets such as power need to admit jumps in the stochastic process. ^

bi,tf = [µi(t) – λi(t)ki(t) – b if ]dt + σi(t)dzi + γi(t)dqi

(65)

where λi(t) = the average rate at which jumps occur per unit of time, dt ki(t) = the average jump size per unit of time, dt γi(t) = a stochastic variable that captures the nature of the jump process, assuming a jump has occurred. (Merton [1976] calls this type of parameter the “impulse function”) dqi = the change in qi(t), dqi = qi(t + dt) – qi(t) qi(t) = an independent Poisson process The Poisson probability density function is given by P(X = x) =

λx e − λ ; x = 0,1, 2, K x!

(66)

where λ is the only parameter and is the average number of events occuring in the relevant unit of time. Typically, dt is assumed to be sufficiently small as to have only two possible events, x=0 (no jumps) and x=1 (a jump). Notice that if x=0, then P(X = 0) =

λ0e − λ = e −λ 0!

and therefore (assuming a maximum of one jump)

(67)

85

Conclusion

P(X = 1) = P(X > 0) = 1 − e − λ

(68)

It is an empirical exercise to select the appropriate parameter values.

Other Processes There are numerous other processes that could be contemplated within energy markets. One could address stochastic volatility, for example. Stochastic volatility could be handled within a GARCH(p,q) framework. Identifying the appropriate market model remains more of an art than a science within energy markets due, in part, to the lack of quality empirical data. As we increase the complexity of the market model, it becomes increasingly difficult to estimate the parameters. Successful selection of a market model seeks a balance between accurately depicting reality and model simplicity.

CONCLUSION We examined the three categories of approaches to valuation illustrated with options on natural gas forward contracts. The market comparables approach can be used when a liquid put market is available through putcall parity. The cash flow adjusted approach can be used when there is insufficient liquidity in the put market; we illustrated this approach using a simple one period binomial model. Finally, subjective beliefs (probability of up and down states) can be incorporated and related implications examined within sight of the discount factor adjusted approach. By combining both the cash flow adjusted approach and the discount factor adjusted approach, useful additional information can be acquired. For example, the combined approach can yield the implied risk premium, given a trader’s market view. Next, basic energy option valuation models were reviewed in the context of both the cash flow adjusted approach and the discount factor adjusted approach. The option valuation models were presented in the context of the discounted expected future asset or forward price. The role of the risk premium was identified as unique to the discount factor adjusted approach. Finally, attention was given to constructing a market model that cap-

86

NEW TECHNIQUES IN ENERGY OPTIONS

tures the variability within energy forward curves. A parsimonious model was explored that permits a variety of shocks to the current forward curve. Specifically, the forward curve is assumed to change based on stochastic processes surrounding level, slope, and curvatures (LSC model). For the natural gas market, a three-parameter model appears adequate. We conclude with a review of different stochastic processes to apply to the LSC model, multivariate normal, mean reversion, jump diffusion, and stochastic volatility.

REFERENCES Arrow, K.J. “The Role of Securities in the Optimal Allocation of RiskBearing.” Review of Economic Studies 31 (1964), 91–96. Beckstrom, Rod A. “Value at Risk: Theoretical Foundations,” Capital Market Strategies (November 1995) (London: IFR Publishing, 1995). Beder, Tanya. “VaR: Seductive but Dangerous.” Financial Analysts Journal (September/October 1993), 12–24. Best, Philip. Implementing Value at Risk (Chichester, England: John Wiley & Sons, 1999). Black, Fischer, Emanuel Derman, and William Toy. “A One-Factor Model of Interest Rates and Its Application to Treasury Bond Options.” Financial Analysts Journal 46 (January/February 1990), 33–39. Black, Fischer, and Myron Scholes. “The Pricing of Options and Corporate Liabilities.” Journal of Political Economy (May 1973), 637–659. Brooks, Robert. “Approaches to Valuation Illustrated with Interest Rate Swaps.” Derivatives Quarterly 4(3) (Spring 1998), 51–62. Brooks, Robert. Building Financial Derivatives Applications with C++ (Westport, CT: Quorum Books, 2000). Brooks, Robert. Interest Rate Modeling and the Risk Premiums in Interest Rate Swaps (Charlottesville, VA: Research Foundation of Institute of Chartered Financial Analysts, 1997). Brooks, Robert. “Value at Risk Applied to Natural Gas Forward Contracts,” Working Paper, University of Alabama (2000). Brooks, Robert, and David Yan. “London Inter-bank Offer Rate (LIBOR) versus Treasury Rate: Evidences from the Parsimonious Term Structure Model.” Journal of Fixed Income 9(1) (June 1999), 71–83. Campbell, John Y., Andrew W. Lo, and A. Craig MacKinlay. The Econo-

References

87

metrics of Financial Markets (Princeton, NJ: Princeton University Press, 1997). Cochrane, John H. Asset Pricing (Princeton University Press, Forthcoming). Cox, J.C., and S.A. Ross. “The Valuation of Options for Alternative Stochastic Processes.” Journal of Financial Economics 3 (1976), 145–166. Dewing, Arthur Stone. The Financial Policy of Corporations (New York: John Wiley & Sons, Inc., 1941), 275–277. Reprinted in Charles D. Ellis, Classics: An Investor’s Anthology (Charlottesville, VA: Institute of Chartered Financial Analysts, 1989). Dothan, Michael U. Prices in Financial Markets (New York: Oxford University Press, 1990). Dowd, Kevin. Beyond Value at Risk (New York: John Wiley & Sons, 1998). Duffie, Darrell, and Kenneth J. Singleton. “An Econometric Model of the Term Structure of Interest-Rate Swap Yields.” Journal of Finance 52(4) (September 1997), 1287–1321. Durand, David. Basic Yields of Corporate Bonds, 1900–1942, Technical Paper No. 3 (Cambridge, MA: National Bureau of Economic Research, 1942). Ellis, Charles D. Classics: An Investor’s Anthology (Charlottesville, VA: Institute of Chartered Financial Analysts, 1989). Energy Information Administration. Energy Information Sheets (July 1998) (Washington, DC: National Energy Information Center, U.S. Department of Energy, 1998). Energy Information Administration. International Energy Outlook 2000 (March 2000) (Washington, DC: Office of Integrated Analysis and Forecasting, U.S. Department of Energy, 2000) (www.eia. doe.gov/oiaf/ieo/index.html). Gordon, M.J. “Dividends, Earnings and Stock Prices.” Review of Economics and Statistics 41 (May 1959), 99–105. Hansen, Lars Peter, and Scott F. Richard. “The Role of Conditioning Information in Deducing Testable Restrictions Implied by Dynamic Asset Pricing Models.” Econometrica 55(3) (May 1987), 587–613. Harrison, J., and D. Kreps. “Martingales and Arbitrage in Multiperiod Securities Markets.” Journal of Economic Theory 20 (1979), 381–408. Hull, John C. Options, Futures and Other Derivatives, 3rd ed. (Upper Saddle River, NJ: Prentice Hall, 1997).

88

NEW TECHNIQUES IN ENERGY OPTIONS

Jorion, Philippe. Value at Risk: The New Benchmark for Controlling Market Risk (Chicago, IL: Irwin Professional Publishing, 1997). Levy, Haim. Introduction to Investments (Cincinnati, OH: South-Western College Publishing, 1996). Lintner, J. “Security Prices, Risk and Maximal Gains from Diversification.” Journal of Finance (December 1965), 587–615. Merton, Robert C. “Option Prices When Underlying Stock Returns Are Discontinuous.” Journal of Financial Economics 3 (January–March 1976), 125–144. Merton, Robert C. “Theory of Rational Option Prices.” Bell Journal of Economics and Management Science 4 (Spring 1973), 141–183. Nelson, Charles R., and Andrew F. Siegel. “Parsimonious Modeling of Yield Curves.” Journal of Business 60(4) (1987), 473–489. Sarnoff, Paul. Russell Sage: The Money King (New York: Ivan Obolensky, 1965). Sharpe, William F. “Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk.” Journal of Finance (September 1964), 425–442. Stuart, Alan, and J. Keith Ord. Kendall’s Advanced Theory of Statistics. Vol. 1: Distribution Theory, 5th ed. (New York: Oxford University Press, 1987). Stoll, Hans R. “The Relationship Between Put and Call Option Prices.” Journal of Finance (December 1969), 801–824. Sturm, Fletcher J. Trading Natural Gas: Cash Futures Options and Swaps (Tulsa, OK: PennWell Publishing Company, 1997). Svensson, Lars E. O. “Estimating Forward Interest Rates with the Extended Nelson & Siegel Method.” Institute for International Economic Studies, Stockholm University, Reprint Series, No. 543 from Quarterly Review (Swedish Central Bank) No. 3, 1995. Williams, John Burr. The Theory of Investment Value (Cambridge, MA: Harvard University Press, 1938). Reprinted in Charles D. Ellis, Classics: An Investor’s Anthology (Charlottesville, VA: Institute of Chartered Financial Analysts, 1989). Willner, Ram. “A New Tool for Portfolio Managers: Level, Slope, and Curvature Durations.” Journal of Fixed Income (June 1996), 48–59.

CHAPTER

6

The New Accounting Rules for Derivatives: FAS 133 and Its Impact on Energy Convergence Dr. Nedia Miller, Options Principal MILLER CTA, member of NYMEX

INTRODUCTION This chapter is designed to give an overview of the new accounting rules for derivatives, Financial Accounting Standards (FAS 133) and their effect on the energy markets as well as on related secondary markets. FAS 133 was implemented on January 1, 2001 and are the new rules for hedge accounting in the United States. It seems evident that these standards will be adopted globally in the future. First, we give a brief explanation of why the new accounting rules have emerged. Secondly, we introduce techniques for performing effectiveness testing, which is the main criteria for the derivative instruments to qualify for special accounting treatment, defined as “hedge accounting.” Third, we display the conceptual framework of FAS 133, to familiarize the reader with the different types of hedge accounting and the resulting new elements of the financial statement.

This chapter is dedicated in loving memory to Lori Lopez, a close friend and exceptional colleague. Lori was always an inspiration to all who came into contact with her and we miss her spirit and talent. She was a former KPMG employee, and passed away during Christmas 2000.

89

90

THE NEW ACCOUNTING RULES FOR DERIVATIVES

Finally, we are keeping an open mind regarding the industry’s positions (both pro and con) concerning the implementation issues of the new hedge accounting rules for derivatives. At this point we will leave it up to the reader to make his/her own conclusions about how to make economically sound decisions for conducting business, trading/selling commodities, and implementing risk management practices in this new environment of financial reporting. These business decisions include what instruments to use and how to structure risk management strategies in order to comply with the new hedge accounting rules. In the previous chapter of this book, the author introduces the reader to the challenges of pricing financial instruments such as options. Now, after the enforcement of the new accounting rules for derivatives by the Financial Accounting Standards Board, risk managers, traders, and salespeople are facing the following questions: what derivative instruments should they choose for their hedging practices, which instruments should they trade, and should they choose more plain vanilla type instruments, which eventually would facilitate the process to qualify for hedge accounting. Along with this, they have to consider a number of other related issues that were not part of their concerns before the introduction of FAS 133. We can not answer all these questions but will try to give an objective and informative picture of where the derivatives industry is evolving at the present time, how the new accounting rules for derivatives are evolving, and the interdependence of these two developments. The reasons why the FASB came up with these new accounting rules for derivatives are numerous. However, foremost is the fact that over the last 10 years a large number of derivative instruments fled the global futures exchanges for the OTC derivatives markets, which added significantly to the already existing lack of transparency in these markets. There is a general consensus that the development of derivatives has outpaced the development of standards to account for them. The general practice for dealing with trading or market making positions is to mark them to market, valuations being included in the bottom line of the firm. It has been harder to pinpoint how to account for derivatives that have been used to hedge (hedging inventory in the commodity markets, import/export transactions in the foreign exchange markets, etc.). Without clear accounting standards for derivatives, practitioners have been left to interpret and extend standards to cover hedging derivatives as best they can. As a result, derivatives have, in general, been treated as items off the balance sheet.

Introduction

91

Since there has been a tremendous growth of derivatives instruments and not enough transparency in the marketplace, there were a number of large failures during the 1990s (Metallgesellschaft, Orange County, Barings, and others), for which derivative instruments got the blame. Over the past several years, two accounting organizations responsible for defining consistent standards for financial statements have developed standards to address some of the issues specific to the reporting of derivatives. The U.S.based Financial Accounting Standards Board (FASB) came out with the FAS 133, effective from the beginning of 2001. The London-based International Accounting Standards Committee (IASC) issued IAS 39, effective June 1, 2001. The new standards are designed to address the rules surrounding mark-to-market, or the fair value accounting approach. Both organizations agreed that trading books should be marked-to-market. However, there are differences in the agendas of the two organizations regarding the new accounting rules for derivatives. For example, while FAS 133 focuses specifically on hedge accounting for derivatives, IAS 39 covers both cash and derivative instruments. One can look up other distinctions in the organizations’ websites. This chapter specifically addresses the accounting treatments for derivatives according to FAS 133 and its implications in the energy converging markets. According to FAS 133, the new accounting rules will determine which derivatives qualify for hedge accounting, or the so-called short method, and which instruments do not qualify, or cannot be used to hedge a position. In addition, the FASB introduced a new definition for derivatives as well as three methods to perform hedge accounting. These new accounting rules include an entirely new model for accounting for derivative hedges. It does away with deferral and synthetic accounting by making all derivatives instruments appear on the balance sheet at “fair value.” In addition it requires companies to maintain hedge program accountability by posting any hedge ineffectiveness in current income. It is important to mention that FAS 133 focuses on hedge tools and not on the type of risk that has been hedged, which is a fundamental change in hedge accounting as it had been practiced under FAS 52, (currency), FAS 80 (interest rates and commodities). Under the old accounting rules, hedge accounting treatment revolved around the extent to which and under what circumstances Treasury can defer recognition of gains and losses on hedges. Under FAS 133, hedge accounting treatment revolves around where to recognize the change in value. Ultimately FASB

92

THE NEW ACCOUNTING RULES FOR DERIVATIVES

would like to have all financial instruments on the balance sheet at fair value, but such a radical step would have a massive effect on the income statement of many companies. Therefore, the Board compromised, offering hedgers who are using derivatives a certain amount of flexibility in the form of Other Comprehensive Income. The OCI category of the financial statement was specifically created for derivative hedge accounting use. By introducing this new category and keeping the traditional current income category, the gains and losses on the hedged transactions are being parked in OCI until it is time to recognize them in current income. This new category has the role of keeping all changes in fair value as they occur, during the life of the transactions, and post them to current earnings at the end of the transaction. This is to allow the revenues from derivatives hedged transactions to be identified and to be more transparent on financial statements. In other words, income from derivative hedges posted into the OCI account is recycled into recognized income to coincide with the recognition of the hedged asset. If the transaction fails to qualify for hedge accounting treatment, the overall change in fair value is posted directly to earnings. As already mentioned, under FAS 133 the new definition of derivatives is much broader. It goes way beyond options and custom tailored derivative instruments to include futures and forward contracts. Thereby a derivative instrument is now defined as a contract with all the following characteristics: The derivative has an underlying and either a notional amount or payment provisions or both; and it does not require initial investment, but net settlement or its equivalent. If the above requirements are met, then this instrument is a derivative and FAS 133 applies (see Table 6.1). Most critically, derivative instruments must be recorded in financial statements at their fair values; and beyond that, additional analysis is required for even the most basic accounting entries.

TABLE 6.1 Underlying and Notional Derivative Stock option Currency forward Commodity future Interest rate swap

Underlying

Notional

Stock price Exchange rate Commodity price Interest index

Number of shares Number of currency units Number of commodity units Dollar amount

Introduction

93

The use of the derivatives falls under one of the following categories. ■ For speculative purposes. ■ To hedge the exposure associated with the price fluctuations of an asset, liability, or firm commitment. ■ To hedge the exposure associated with an uncertain forecasted cash flow. ■ To hedge the exposure associated with the currency component of a net investment. ■ In the first case where derivatives are used to perform speculative application, derivatives gains and losses must be marked-to-market and will be recognized in the current earnings category of the financial statement. Fair value accounting is used for fair value hedges, which are derivatives that are used to hedge transaction, that are marked-to-market, for example, trading assets. The accounting treatment of the hedge follows the accounting treatment of the asset. Both are marked-to-market and both cash flows are recognized in current earnings and recorded through the earning cycle right away. Since both the asset/liability position and the hedge are marked-tomarket, the effectiveness of the hedge is directly reflected in earnings of the current period. The accounting for the derivative is the same as that for speculative applications, but in this case, due to the risk being hedged in the underlying, the hedged exposures must also be marked-to-market, after which the results must be posted to current income. Ideally, the hedge will be a perfect hedge whereby the gains and losses will offset the losses /gains on the underlying exposure, so there will not be any impact to earnings. For cash flow hedges, derivative gains/losses must be evaluated to determine how much is ineffective. The ineffective portions of the hedge must be realized right away in current income, while the effective portions of the hedge are initially sent to Other Comprehensive Income. These effective portions are also posted to earnings later in the same time frame in which the forecasted cash flows affect earnings. For hedges of the currency exposure of a net investment in foreign operations, hedges have to be marked-to-market. Effective hedge results must be consolidated with transaction adjustments in other comprehensive income. The ineffective portion (differences) between total hedge results and the translation adjustments being hedged will be posted to earnings. Once the hedging instrument meets the above characteristics (for the

94

THE NEW ACCOUNTING RULES FOR DERIVATIVES

transaction to qualify as a hedge), the derivative instruments, which are going to be used for the hedge have to also show a close correlation with the underlying instrument (80%–20% rule). In order to measure the relationship between the derivative and the underlying position to be hedged, FASB demands an effectiveness test, which has to be performed at the inception of the transaction and on an ongoing basis (at least quarterly). In other words, if the hedge proves to be highly effective the transaction not only qualifies for hedge accounting, but the test determines what amount of the change in value will have to be recorded immediately on the income statement. For example: If the hedge is up $50 and the position is down $45, the $5 must be marked-to-market right away in current income/earnings. The assumption of a strong correlation between the hedge and the asset leaves room for interpretation and gray areas of the law. In addition, measuring the effectiveness of a hedge with an unstable or changing correlation proves problematic. The chosen methodology (statistical method) that will be applied to perform the calculation significantly affects the hedge’s effectiveness test. One should keep in mind that hedge accounting remains permissible only if both the specific definition and specific procedures are in place to document and measure the effectiveness of the hedging relationship. The most popular statistical methods that the industry has been using to perform this test are regression analysis and dollar offset. If dollar offset is used for assessment, one can use period by period or cumulative. Both of these statistical methods have their pitfalls, which are not going to be discussed. Just to give an idea of how much leeway the FASB leaves to the hedger to decide what method to use to perform the effectiveness test and what to include in the test, we would like to give an example of hedging with an option. The following are possibilities to measure the effectiveness of option hedge. ■ Measure the effectiveness based on the intrinsic value, and exclude changes in time value from the test. ■ Measure the effectiveness test based on changes in the option’s minimum value (intrinsic value plus the effect of discounting), and exclude the change of volatility value. ■ Measure the effectiveness of forward/option hedge assessed based on changes in fair value, spot, and forward.

Introduction

95

In each case, excluded portions must be recognized in earnings right away as they occur and the effectiveness of “similar hedges” generally should be assessed in the same way. It should be mentioned that the standard does offer a safe harbor in the form of the short cut method, which says that if the hedge and the underlying exposure match perfectly (based on a series of rules provided in FAS 133, paragraph 68), companies can assume there will be no ineffectiveness in their hedge. This means that the hedger would not have to go through the long haul method of measuring the hedge and the exposures separately. It can be concluded that in general, effectiveness works on two levels and the measures are different for each. 1. To get a position to qualify as a hedge (i.e., to be able to designate a derivative as a hedge of some underlying exposure), the company must be able to prove “prospective high level effectiveness.” Those are the expectations of high level offset. 2. To be able to keep this hedge accounting treatment, companies need to measure the fair value of both the derivative and the hedge periodically (at least every quarter). The change in fair values must offset each other on a dollar offset basis. If they do not, companies must reassess prospective effectiveness to insure that the position still meets the close correlation requirement and the hedge overall can remain under the hedge accounting rules. Companies will also need to identify any ineffectiveness and record it in income or OCI, depending on the hedge relationship. After the company performs the effectiveness test and the correlation between the derivative and the underlying instrument is established, the transaction has to be treated according to the FAS 133 accounting framework. Derivative hedges are classified as explained on the previous page. For example in the energy industry ■ Fair value hedges are typical for the hedging of energy inventories, fixed assets/liability, firm commitment, plant, property, and equipment. ■ Cash flow hedges are typically used for forward floating rate purchases or sales, forecasted events, and underlying events that are probable and derivative. Counterparty must be likely to perform (i.e., during the California energy crisis of 2000/2001, the counterparty did not perform).

96

THE NEW ACCOUNTING RULES FOR DERIVATIVES

EXAMPLE 1: INVENTORY HEDGE On November 1, Utility X buys 10,000 dekatherms of gas and injects it into a Northeast storage facility. The utility wishes to hedge the price exposure (fair value) of this gas in inventory by entering into a NYMEX futures contract to sell this gas in March. Specifics of the transaction are as follows. ■ Inventory cost (spot price at November 1): 4.00 ■ NYMEX March futures price at inception (November 1): 3.50 ■ Basis at inception: +.50 The futures contract represents a derivative instrument which Utility X designates as a fair value (FV) hedge of natural gas inventory. As an FV hedge, the change in the FV of the futures contract will be recorded to income as will the change in the FV of the hedged item (inventory). At November 30, assume prices are as follows (see Table 6.2). ■ ■ ■ ■ ■ ■ ■

Entries at November 30: Record derivative at fair value: Hedging loss (P&L): 1,000 Derivative liability: 1,000 Record increase in FV of inventory: Inventory: 2,000 Hedging gain (P&L): 2,000

The derivative liability will be settled and the written-up inventory will be booked to COGS (P&L) when the inventory is sold. This example clearly shows how the hedge exposure is being marked-

TABLE 6.2 Inventory Hedge At November 1 At November 30 Increase/Decrease

Nearby Spot Price

NYMEX

Basis

4.00 4.20 +.20 Inventory Gain

3.50 3.60 –.10 Derivative Loss

+.50 +.60 +.10

Example 2: Production Hedge

97

to-market, or recorded at fair value at the same time as the asset being hedged. In our next example in order to demonstrate how cash flow hedge is being recorded, we choose production hedge.

EXAMPLE 2: PRODUCTION HEDGE A producer wishes to hedge a portion of its anticipated gas production for the month of January 2001 by entering into an OTC gas swap contract in November. The swap calls for the producer to receive $4.20 and pay NYMEX L3D. The producer will physically sell his volumes at the Chicago City Gate. Specifics of the swap contract and Chicago pricing at November 1 are as follows. ■ Jan. NYMEX price is $4.20 ■ Forward Chicago City Gate price is $4.50 ■ Thus, basis = 4.50 (physical) – 4.20 (swap price) = .30 Accounting entries: ■ At inception—No entry, futures contract is at fair value. ■ At November 30—Assume prices are as follows: NYMEX Jan. contract $4.60 Chicago Jan. forward $4.80 Thus, basis equals .20. ■ OCI :3,000 ■ (10,000 dekatherms*(4.80–4.50)) ■ Hedging loss (P&L): 1,000 ■ Liability from price risk management: 4,000 ■ (10,000 dekatherms × (4.20–4.60)) Note: Because the unhedged basis went from .30 to .20, .10 of the .40 movement in price of the futures contact was not effective and thus was recorded to P&L currently. Also, assume that prices did not change upon settlement of the January contract (settles at end of December). ■ Liability from price risk management: 4,000 ■ Cash: 4,000

98

THE NEW ACCOUNTING RULES FOR DERIVATIVES

Settlement entries: ■ ■ ■ ■

Cash (10,000 × 4.80): 48,000 Gas sales revenue (P&L): 48,000 Hedging loss (P&L): 3,000 OCI: 3,000

Net P&L effect: ■ Gas sales revenue: 48,000 ■ Hedging loss: 4,000 ■ Net P&L effect: 44,000 Net P&L effect represents $4.50 Chicago forward price of deal at inception less .10 basis movement times 10,000 dekatherms. According to FAS 133, cash flow hedges are measured by comparing the derivative’s gain or loss with the change of the cash flows of the associated hedged item. In our example since the unhedged basis went from .30 to .20, .10 of the .40, movement in the price of the futures contract was not effective and therefore was recorded to P& L (was posted to earnings). As we already have seen in both examples, we made the assumption that these transactions had already qualified for hedge accounting, based on other results from the effectiveness test, which had shown a high degree of effectiveness (80%–20%). In contrast to the first example (the inventory hedge, which is a fair value type of hedge), the production hedge, which is a cash flow type of hedge, is measured by comparing the derivatives gain or loss with the change in the cash flows of the associated hedge item. FASB leaves it up to us to decide what type of technique we should use in order to perform the effectiveness test, and at first glance it appears that the rules are quite generous. However, the problem is that even if two price levels are highly correlated, this is not really statistically valid information, because there is no guarantee that this correlation will remain steady during the life of the hedge. In the energy industry, in order to hedge the exposure of natural gas (West Texas) we use a derivative with a price based on a Henry Hub (the NYMEX contract delivery point). To validate the “highly effective expectation,” under FAS 133, all we need to do is to prove that the two respective prices (the price in West Texas and the price in Henry Hub) are highly correlated by performing effectiveness testing. Again, this

Example 2: Production Hedge

99

does not necessarily prove that the correlation will remain highly effective in the future. Further, price changes might not remain highly correlated during the life of the hedge (even if the hedge proves to satisfy the intended economic objectives of the new rules of FASB and the effectiveness test shows a close correlation between the derivative and the item being hedged). FASB tried to correct this by requiring that the effectiveness test should be performed at least quarterly during the duration of the hedge and if the correlation does not show the required effectiveness, the hedge should be terminated right away and the gains/losses posted to earnings. FAS 133 also has a section on embedded derivatives, which are frequently seen in the power markets. Transactions including embedded options, particularly those used in the power industry are difficult to account for. In the section about embedded derivatives it is explained which transactions are regarded to have embedded derivatives, and therefore no bifurcation of the derivatives is required, in which cases the derivative has to be accounted for separately. Basically, if the hedging transaction consists of embedded derivatives that are not clearly related to the economic characteristics of the host contract, we have to separate the derivative from the host contract and account for it separately. This means that the hedging transaction has to be frequently bifurcated or separated into different components in order to be properly recorded. We will not go into many details but would like to mention just a few cases when the derivative has to be accounted for separately from the host contract. For example, after the deregulation of the U.S. electric power industry in 1996, the industry faced numerous hedge transactions left with longterm contracts and different firm commitments linked to the host contract, which were considered contracts with embedded derivatives before FAS 133. The industry had a major problem with pricing these embedded derivatives. Under the new FAS 133 accounting rules this industry had to deal with both pricing these derivatives transactions and accounting for them. With the current discussions about deregulating bandwidth it is to be expected that similar challenges will be faced by risk managers and other financial professionals. In bandwidth trading we typically see transactions with linked long-term commitments. These are transactions that have embedded and nonembedded derivatives. There will probably be even more significant downfalls than the ones we have already seen in the power industry. For example, it was reported after the first quarter 2001 of the implementation of FAS 133 that El Paso had lost $800 million. Of course, we

100

THE NEW ACCOUNTING RULES FOR DERIVATIVES

do not know to what extent these losses are actual losses and what part of them is purely a result of the financial recording technique required by FAS 133. This can be seen only in the future, when we can compare accounting periods where equal accounting rules apply. It is important to keep in mind that under FAS 133 derivative hedge transactions must be recorded when they occur. In addition the organization has to have the ability to value derivative positions on a mark-tomarket basis, since this is the basis for every procedure that determines effectiveness. Overall the main contribution of the new rules for hedge accounting is that they define timing and manner of revenue recognition in a reporting period, (i.e., the bottom line of the firm that acts as its primary performance indicator). In addition, the transparency of the use of derivatives in the financial statements will be improved. Positions will be represented on the balance sheet with more transparency than was previously required. It clears the way for further developments on reporting derivatives and risk. On the other hand, we should mention that the new accounting standards are not always easy to implement. In some businesses, obtaining a sound and consistent P&L statement for derivative transactions during the course of regular business hours may be difficult. Many derivative markets may not be liquid (for example the power market), and these standards do not address the method to value these assets fairly. It is to be expected that FASB will make some adjustments in order to facilitate the implementation of these new accounting rules as they evolve. It is expected that FAS 133 standards for hedge accounting will be applied throughout the world as the accepted accounting standard especially in light of the Enron financial debacle.

CHAPTER

7

The Central and Eastern European Energy Sector Reforms: Convergence versus Divergence Dr. Markus Reichel President, EconTrade Deutschland GmbH

INTRODUCTION It is generally agreed upon that the fall of the Communist system in Central and Eastern Europe (CEE) took place as the result of a long-lasting process of degradation of the political and economical regime, but it happened in a historically unique spirit of openness to change.1 While the political hegemony of the Soviet Union led to far-reaching similarities of all macroeconomics and political systems in CEE, a “window of opportunity” opened after 1990 and made, starting from a fairly equal initial state, individual changes and development paths of the CEE countries possible. The respective economies seemed to have reached a bifurcation point.2 The bifurcation of the respective national development paths concerning institutions, and political and economical system initiated a divergence process in the whole former Communist bloc, and thus also in CEE. Nevertheless, this divergence process was influenced by the existence of a worldwide opinion that after the fall of the Communist system only political and economical structures that are based on market forces have a chance to be successful.3 The European Union seems to work as “attractor”

101

102

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

for the transformation process of the CEE countries. Due to the economical dominance of the EU, far-going convergence to the EU legislation and economical integration is desired. This means that convergence and divergence processes interfere and historical experience will show which countries will remain locked into a disadvantageous post-socialist system,4 which ones will build up an autonomous successful “third way,” and which ones will manage to fully converge to and integrate with the Western European economical and political community. The energy sector serves as a good indicator for forecasts of those results; as it was looked upon as strategic, any opening of this sector to new organizational schemes is the result of a successful transformation of formal and informal institutions and makes an extrapolation of this development to the future and to other sectors possible. This chapter aims to help the reader understand to which extent convergence and divergence of the CEE energy sector to the ongoing worldwide market liberalization process. The chapter is structured as follows: Section 2 presents some energy economical tendencies of CEE and describes qualitatively the prospects and obstacles of this region (compared to Western Europe) to converge to the liberalized and technically modernized energy economies. Section 3 gives a rough description of the idea of path dependence within economical systems. Section 4 describes the degree and results of international cooperation as a means for enhancing convergence. Section 5 identifies indicators in order to judge the degree of convergence of CEE energy economies to the EU. Finally, Section 6 gives an in-depth description of Poland and Ukraine as extreme cases and their degrees of convergence or divergence to the EU, which is here symbolizing a liberalizing—or liberalized—energy economy. Finally, the prospects for electricity trade will be sketched out.

ENERGY AND ECONOMIC TENDENCIES WITHIN THE CENTRAL AND EASTERN EUROPEAN COUNTRIES (SECTION 2) At the beginning of the 1990s, the CEE energy sectors were faced with ■ Low growth of consumption ■ Severe problems resulting from the general transition process

Energy and Economic Tendencies within the Central and Eastern European Countries

103

■ Very bad state of the infrastructure of the energy sector ■ Ecological problems as a result of emissions and lack of security in nuclear energy ■ Bad financial situation Having solved similar problems that resulted from relatively parallel or converging development paths between 1950 and 1990, similar remedies were seen as helpful. These include the creation of financial incentives for private investments by developing macroeconomical and jurisdictional frameworks, abolishment of cross-subsidies, and limitation of political influences on the energy sector. Any of these steps was hard due to the existing social situation; more or less they all would lead to rising energy prices, while the net income of households sank. Thus, the political will for reforms at the beginning of the transformation period of the 1990s can be judged by watching the price development. Certainly, it is hard to convert the success of the restructuring into one key figure. A possible indicator has to be seen in the ratio of the growth rates of industrial and private consumer prices (Table 7.1). An index >1 means that the general subvention of private electricity consumption by industry was (at least partially) stopped. An index of 2, for example, means that household prices rose twice as fast as industrial prices in the respective country (compared with 1990). Obviously, Poland and Hungary seem to have quite a unique position. According to leading economic indicators, one can divide the Central and TABLE 7.1 Development of Household and Industrial Prices in Eastern Europe since 1990 Country Poland Czech Republic Hungary Slovakia Bulgaria Romania Lithuania Latvia Estonia

1990

1992

1994

1996

1998

1999

1 1 1 1 1 1 1 – –

1.79 0.55 1.39 0.89 0.68 0.53 0.65 – 1

2.11 0.56 1.93 0.58 0.97 0.82 0.58 1 1.19

2.75 0.69 2.87 0.62 0.76 0.82 0.59 1.32 1.4

2.96 1.16 2.91 0.64 0.93 1.26 (1997) 0.57 1.32 1.84

No info. 1.16 3.01 1.01 1.03 No info. 0.56 1.32 No info.

Source: Riesner (2000)

104

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

Eastern European countries in three groups: Poland and Hungary belonging to the leading group, followed by the bulk of the CEE countries, among others the Czech Republic and Slovakia, which started the restructuring process but still have severe deficits. In Section 6, Poland and Ukraine will be characterized with an in-depth analysis. Finally, a country like Belarus is still on a pre-reform level (no data available here). Related to these country-specific positions, the EU is negotiating with most of the Central and Eastern European countries concerning a timetable for accession. For energy-related matters, this concerns mostly the convergence to the EU Directives regarding the creation of a common electricity and gas market. In the rest of this chapter we focus on the electricity sector as it indicates very well the development of the whole energy sector. It is interesting enough, that the directive governs the deregulation of the member countries in a similar way as in the case of the countries that want to enter the EU. So, it is already clear that the European unification process is the best motor for liberalization in every Eastern European country and a clear means of their convergence to the Western European market model. Nevertheless, there are some qualitative factors to be seen as the main chances and obstacles to a general restructuring and harmonization of the CEE energy sector with EU level. They refer to the current and future competitiveness of the sector. ■ Technical know-how in Central and Eastern Europe is very good. ■ Currently there are low production costs due to low labor costs and mostly amortized generation capacities that can be run on the basis of variable costs and partially subsidized fuel costs. ■ The electricity sector is still understood as strategic. ■ The general choice between privatization within a liberalized (existence of TPA for eligible customers) and a not yet liberalized (long-term power delivery contracts that are politically influenced) framework is not yet taken. ■ Lacking legal framework. ■ Financial engineering know-how is not yet accustomed to the needs of a liberalized energy market. ■ Energy is not yet understood as a commodity to do trade with. ■ The need for methods-to-market energy is not yet understood.

Path Dependencies in Economic Systems (Section 3)

105

■ In some countries, there is a severe lack of liquidity in order to create a functioning energy market. ■ Social problems caused by opening of the electricity market (and resulting lower demand for local fuel sources) have not yet been anticipated. ■ Future production costs will have to rise in order to finance the modernization of the existing power plants. ■ A liberalized international electricity trade with countries from the EU will be partially impossible due to reciprocity. The lacking convergence of the CEE electricity sectors to the EU market has already resulted in paradoxes, for example, some CEE countries like Poland, Ukraine, and the Czech Republic are seen as possible sources for cheap energy. The industrial sector has partially lost competitiveness on the international markets as industrial energy prices are today higher than in some Western countries.

PATH DEPENDENCIES IN ECONOMIC SYSTEMS (SECTION 3) Under certain circumstances why do economical systems not evolve toward an obviously “better”—the theoretically optimal—state? This question gave rise to an ongoing discussion on the existence of path dependencies, that is, the question to what extent the evolution of such systems is limited by the initial conditions and if such limitation does lead to sub-optimal results.5 In order to describe such effects, the concept of lock-ins, lock-outs, and lock-in/lock-out-breaks has been developed.6 An economical system is “locked-in” to a certain trajectory,7 if more efficient trajectories are existing, but cannot be reached without external influence.8 On the other hand a certain development state is “locked-out” if it cannot be reached without external influences, but would be more efficient, and if this increase of efficiency would even lead to a payback of the (economical, social) costs of the lock-out break. Within the context of this paper, the question is if, concerning the development of the CEE energy economical systems, such path dependencies can be identified. It is not the aim of this paper to discuss if the goal of most countries, convergence to the common Pan-European energy market and its structures, is the optimal one.

106

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

Let us symbolize the total social costs that are related with a certain market structure x by the curve Ct = Ct(x) and let us assume that the dynamics of the system are driven by the potential that is symbolized by P = (X). In that case the dynamics of the system can be compared to a ball that P is running to the next (local) minimum of P. Let us assume that x represents a proxy like “efficiency,” “flexibility” of market structures. If we assume that the market structure of the EU energy sector is more efficient/flexible than those of the CEE countries (i.e., xEU > xCEE) and that Ct(xEU) < Ct(xCEE) we might assume the following situation (see Figure 7.1). If such constellations would occur, we could describe the following qualitatively different situations: ■ Country CEE1 has not yet reached the critical mass in order to be on a self-sustaining road to further reforms. There is a tendency that it will fall back to a pre-reform level. ■ Country CEE2 has arrived at the critical point; it can be seen as the bifurcation point as it will either fall back to a pre-reform level, that is, diverge, (x → 0) or continue its reform path, that is, converge to the EU (x → xEU ).

?

P

Ct

X CEE1

CEE2

CEE3

XEU

Xmax

FIGURE 7.1 Path Dependence in Economical Processes

International Cooperation (Section 4)

107

It will not be the aim of this paper to prove or find out the existence and shape of Ct and P, this has to be left for further research. Nevertheless, we want to give some indication that relatively marginal differences of the respective development might lead to radical changes of the resulting further development paths, which are hardly revocable.

INTERNATIONAL COOPERATION (SECTION 4) International cooperation is a central element of successful convergence. For a Western judgment of the successes of cooperation between East and West, refer to the respective reports from World Bank’s or European Bank for Reconstruction and Development’s side. How do CEE experts themselves judge the successes of East–West cooperation? In 1999 a survey was undertaken in order to find out how energy experts from different countries in Central and Eastern Europe (Romania, the Czech Republic, Slovakia, Bulgaria, Hungary, Poland, Russia, Ukraine) judged the energy-related cooperation among their respective countries and the main Western countries. The study represents 414 answers which cannot be judged as statistically representative, but give a good overview about the tendencies and problems of international cooperation.9 The questions concerned among others the following topics (1: very low/bad; 5: very high/very good). No. 1: Interest in complex cooperation. No. 2: Interest in strategic cooperation. No. 3: Knowledge of the economical reality in the respective transformation country. No. 4: To what extent are representatives from transformation countries seen as partners? No. 5: Degree of correspondence between the activities of a Western country’s government and its enterprises. No. 6: To what extent are specific influences of the transformation period taken into account by the Western partners? No. 7: To what extent is there a readiness to take risks? No. 8: To what extent are the Western partners ready to help measures in preparing common projects?

108

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

Figures 7.2 and 7.3 present the average overall questions and a comparison to a similar research project from 1995. Summarizing, one can see that above all Denmark, but also France and Germany, are judged relatively well. Comparing the results with those from a similar survey from 1995, it can be seen (amongst others) that Germany has improved its position.

INDICATORS FOR CONVERGENCE TO MODERNIZED AND LIBERALIZED MARKET STRUCTURES (SECTION 5) Within this chapter some basic and derived indicators for convergence to or divergence from the relatively liberalized and modernized market structures of the EU are sketched. We differentiate between the following parameters: ■ Institutional framework The underlying institutions of the energy market create the framework for a market environment that can be

Germany

3,27

Austria

3,24

France

3,44

Netherlands

3,25

U.K.

3,04

U.S.A.

2,96

Japan

2,32

Belgium

2,70

Italy Spain

2,75 2,11

Canada

2,59

Greece Sweden

2,42 2,06

Switzerland

2,41

Denmark Turkey

3,66 2,17

FIGURE 7.2 Successes of International Cooperation, 1999 Source: Riesner (1999)

109

Indicators for Convergence to Modernized and Liberalized Market Structures

Results 1995

Results 1999

3,27 Germany 2,52

3,04 U.K. 2,78

2,96 U.S.A. 3,00

FIGURE 7.3 Successes of International Cooperation, 1995–1999 Source: Riesner (1999)

compatible to the demands of the EU common energy market. In this section, we are not going into an in-depth analysis of the whole energy related law system, but ask for the degree of market transparency, liberalization, and privatization. ■ Demand/supply framework The economical counterpart to the institutional framework is the development of supply and demand, and related matters. The state of the modernization of power plants is essential for the capability to successfully converge to the EU energy sector. Capacity and demand development and the state of energy exchange with neighboring countries above all from the EU have to make a factual integration and thus convergence possible. ■ Evolution of a market environment While the institutional system lays the ground for evolution of a modernized and liberalized market, the factual development might be different due to the dominance of market-preventing informal institutions. Central indicators are here the degree of competition and the absolute and relative price development (e.g., existence of cross-subsidies). See Figure 7.4.

110

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

Complexity of indicator

Evolution of market environment • degree of competition • price development

Institutional framework • market transparency • liberalization • privatization

• structural changes

Demand/Supply • capacity development • demand development • energy exchange Indicator

FIGURE 7.4 Indicators for Convergence

RELEVANT TRENDS—TWO EXTREME CASES (SECTION 6) This section characterizes two relatively extreme cases within the group of CEE energy economies. While Poland, even though severe problems and mistakes during the restructurization process cannot be denied, is controlling the process, Ukraine is facing an internal situation that leads to a nearly total loss of hope of foreign investors and institutions in the future renaissance of the once in the past very powerful energy sector.

POLAND Market Structure On April 10, 1997 the new Polish energy law was adopted. It forms the legal background for the current reforms of the Polish energy market. It is more or less built up according to the EU Directive of December 19, 1996 (see Figure 7.5). Since 1996 several implementing regulations were adopted. Due to the Polish Energy Act, a vertically separated market model was introduced to the Polish energy sector. The sector is opened for internal (national) competition according to the following harmonogram.

111

Poland

Regulatory Office EXPORT

IMPORT

Power Producers

Contract market Power exchange

PSE

DISTRIBUTION COMPANIES

Tariff

CUSTOMERS

Local Market

TPA FIGURE 7.5 Structure of the Polish Electricity Market ■ ■ ■ ■

Release customers with a purchase of 40 GWh from January 1, 2000. Release customers with a purchase of 10 GWh from January 1, 2002. Release customers with a purchase of 1 GWh from January 1, 2004. Release all remaining customers from December 5, 2005.

External competition is not yet possible due to the Energy Act. This situation will change on the day of Poland’s entry into the EU, which can be expected for January 1, 2003 or January 1, 2004.

Demand Development Electricity demand will probably develop according to the following scenario of the Polish energy political strategy until the year 2020 (See Table 7.2). District heating and industrial auto generation will keep its current level of about 18 GWh. The share of small-scale CHP (Combined Heat & Power) is politically intended to rise. The share of renewable energy sources and other decentralized energy sources (including CHP) is forecasted to rise to approximately 9 percent to 10 percent. According to the strategy of

112

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

TABLE 7.2 Gross Demand for Electricity in Poland Scenario Gross demand (TWh) Avg. Ann. Growth rate [%]

1997

2005

2010

2015

2020

140.5

167.6

186.9

204.4

233.2

2.2

2.2

1.8

2.6

—

Source: Energy Strategy until 2020, Warsaw 2000

the development of renewable energies, their share alone shall be 7.5 percent in the year 2010. Even though they are currently totally marginal, a significant rise of the share of renewable energies is very probable due to the current policy of the EU, which already has to be respected by Poland. Thus, the share of renewable energies can be foreseen to rise as presented in Table 7.3 (see also Figure 7.6).

Capacity Development The Polish power system has an overcapacity of approximately 5.400 MW. Total installed capacity amounted to 34.112 MW on December 31, 1999. Until 2005 the decommissioning and replacement of about 3.500 MW generation capacity is expected. New capacities will probably be fueled by gas or hard coal. New lignite capacity is not expected to start production before 2007 (Belchatow II, 830 MW; see Figure 7.7). Due to the forecasted demand rise there will not be any overcapacity after 2005. New significant capacities on hard coal base are currently not planned; nevertheless, the replacement of existing capacities can be expected. New hard coal-fired capacities might be expected at the cost of imported coal. Upper Silesia itself will only remain a coal location during the next two decades if there is active influence from the political side; due to

TABLE 7.3 Share of Renewable Energies of the Electricity Demand year share

2001 2,4%

2002 2003 2004 2005 2006 2007 2008 2009 2010 2,5% 2,65% 2,85% 3,1% 3,6% 4,2% 5,0% 6,0% 7,5%

Source: Strategy on Renewable Energies, Warsaw 2000

113

Poland

Public Utility Hard Coal and Other FuelFired Power Stations

Public Utility Hydro Power Stations 4 140

Public Utility LigniteFired Power Stations 7 363

Public Utility Thermal Power Station 130 464

Industrial Auto Producers Thermal Power Stations 795

Public Utility Power Station 130 604

Small Hydro Plants 152

Gross Electric Energy Production 142 119 Net Electric Energy Production 127 861

Electricity Imports 3 491

Net Usable Production 124 993 Electric Energy Supplied to the Network 128 484 Net Electric Energy Consumption 105 086

Electricity Exports 8 426

Electric Energy Sold to Final Customers from Public Power System 99 554

Railway Traction (PKP) and City Traction 4 417

Customers Supplied from Low-Voltage System 44 209

FIGURE 7.6 Electric Energy Balance in the Year 1999 (GWh) Source: www.pse.pl/en/statistics/s5f.html, 2001–02–12

Electric Energy Sold to Customers Directly from Power Stations 13 463 Auxiliary Consumption of Plants 13 463 Electricity Consumption for Pumping 2 868 Overall Power Station Demand (s/s Demand) 305 Losses and Statistical Differences 14 667 Electric Energy Used from Own Generation by Industry 5 532

Customers Supplied from High-Voltage System and Medium-Voltage System 50 928

114

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

Industrial Auto Producers Thermal Power Stations 8%

Public Utility Lignite-Fired Power Stations 27%

Public Utility HydroPower Stations 6%

Public Utility Hard Coal-Fired Power Stations 59%

FIGURE 7.7 Distribution of Installed Capacity in Poland, 1999 social reasons this has nevertheless to be expected. Pressure of substitution of coal by gas on the one hand or indirectly by imported electricity on the other hand will keep the expected profitability of the coal sector low. Coal prices will be slightly above or even possibly below production costs. The first new power plants can be expected after 2005. Gas will determine the new entrance price, also for coal power plants, and thus of the future coal price. The new entrance is forecasted to be 130 PLN/kWh (this refers to a gas price of appr. 13.9 PLN/GJ). Additional base-load capacity of min. 3.000 MW and about 2.000 MW of peak-load capacity is expected to be necessary until 2010, but probably to be covered by import capacities.

International Energy Exchange Currently the total import/export of capacity of Poland is estimated to be approximately 2000 MW. New transmission capacities (Lithuania, Belarus, Ukraine, Germany) will not be built before 2005. In 2010 the total import/export capacity might reach 3 GW. Currently there exist the following interconnections (see Table 7.4). Poland has a shortage of peak capacity. Especially in the future peak load will probably be imported as there is excess peak load capacity on the EU market. A 600 MW-link to Sweden started operation in November 2000. At the moment there are realistic plans for a 110 kV-link between Poland and Lithuania.

115

Poland

TABLE 7.4 Interconnections Between Poland and Neighboring Countries Country

Voltage Level

Germany Belarus Ukraine Czech Republic Slovakia

2*220 kV, 2*400 kV 220 kV 1*220 kV, 1*750 kV (out of work) 2*110 kV, 2*400 kV, 2*220 kV 2*400 kV

Liberalization and Competition The Polish power sector is distorted by long-term power purchase agreements (LTPPA) between several power stations and the Polish power grid company, Polskie SiECi Elektroenergetyczne S.A. (PSE S.A.), that were created from 1993 to 1998 in order to secure ecological investments in various power stations. The main problem for the sector is the fact that ■ They dominate the total sales volume. ■ They are contracts between the utilities thus not neutral and independent grid operator PSE S.A. ■ Their prices are too high. Table 7.5 and Figure 7.8 show the total amount of contracted electricity deliveries 2000–2010. So the LTPPA have to be converted into bilateral contracts between power stations and consumers or distribution companies. In order to cover the otherwise stranded costs, a Stranded Costs Compensation System (SOK) is planned to be introduced. SOK is thought to provide a marketoriented solution to the problem of long-term contracts. The share of approximately 80 percent of the annual sales volume that still is controlled by the PSE will be replaced by a system of guaranteed payments. The additional costs will be covered by increased transmission tariffs. While this system will not lead to a significant change of the current situation that is

TABLE 7.5 Development of Sales According to LTPPA Year 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 GWh 82508 81494 83055 81891 76456 77008 49876 38610 36536 35562 34715

116

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

GWh 90000 80000 70000 60000 50000 40000 30000 20000 10000 0 Year

2000 2001 2002 2003 2004

2005 2006 2007

2008 2009 2010

FIGURE 7.8 Development of Sales According to LTPPA

advantageous for those power plants with LTPPA, the qualitative step is the elimination of the role of PSE that during the past led to manifold discrimination of market participants. Thus, SOK will work as an instrument of the regulator ■ To level out distortions between power plants with and without LTPPA by letting those without LTPPA benefit from the SOK-system. ■ To reduce the distortions of the evolving market environment by the LTPPAs themselves to a bearable level. ■ Nevertheless, a total disappearance of the LTPPA cannot be expected. The Warsaw Power Exchange (Gielda Energii S.A.) began operations during the summer of 2000. Currently the sales are marginal and not driven by market forces. The management of Gielda Energii S.A. postponed the planned introduction of energy derivatives until the end of 2001 at the earliest. As long as the market is controlled by the LTPPA in their existing form, the Warsaw Power Exchange does not have any impetus to develop, since the grid operator plays a double role by being one of the major shareholders of the Power Exchange, by being the main energy trader holding the LTPPA, and by controlling the transmission grid. The introduction of SOK (even if taking place gradually) will nevertheless create the room for a competitive non-OTC market. The good news is that the pure existence of Warsaw Power Exchange puts pressure upon the regulator to create devel-

117

Poland

opment perspectives for the exchange, as it has to be seen as a prestigious project of the Polish government.

Price Development The future price development of the Polish market is hard to predict. The following picture shows several scenarios. ■ The development of the approximate market price on the German electricity market until 1997–2010 (as reference). ■ The development of the Polish wholesale market price until 1997–2000. ■ The short and long transition period scenarios, respectively, concerning the market price development in Poland until 2010. ■ A high-market price scenario concerning the market price development in Poland until 2010. ■ These scenarios differ in several aspects: The depth of the price slump, that is, toward the marginal costs or above this level (mostly as the costs of amortization would have to, at least partially, be covered). The turning point and the length of the transition period. The projections are presented in real terms of 2000 (see Figure 7.9). The reasoning behind the low price scenario is that due to the introduction of competition in Poland prices will collapse to a level near variable

140

Development of generator wholesale prices

120 100 80

High-case market prices

60 40 20

Low-case long transition 2010

2009

2008

2007

2006

2004

2005

2003

2002

2000

2001

1999

1998

1997

0

Low-case short transition

FIGURE 7.9 Historical and Projected Market Prices (Germany, Poland)

118

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

electricity production costs of approximately 70–75 PLN/MWh (depending on coal prices). A relevant interaction between the German and the Polish electricity system is not assumed. This means that a development similar to the development in Germany after liberalization in 1998 would be found where market prices collapsed to a level at variable production costs or even below. The high market price scenario on the contrary says that the Polish development will not be similar to the German due to an unanalogous situation at the beginning of the respective liberalization, and the interactions between the Polish and the German electricity system. The following are reasons for an unanalogous development in Poland compared to the German one. ■ The German market was liberalized at once: In Germany (at least theoretically), full liberalization started right after the introduction of the Energy Law in April 1998. In Poland there are several groups of eligible customers. ■ No clear regulation after liberalization: In contrast to Germany, a regulatory office controls the liberalization process of the Polish energy market. ■ At the beginning of liberalization, the price level was much higher in Germany than in Poland. ■ Huge financial resources and written-off power stations with high technical condition in the German power companies gave them the resources for a price war. ■ Total inexperience of main actors in the market: The main actors in the German market had no experience on a deregulated market and believed that by starting a price war they could significantly change their market share. ■ Vertical integration of German companies provided income on the distribution side, so a price war on the generation side could be cross-subsidized. On the other hand, the Polish energy sector is totally separated.

UKRAINE The power sector of Ukraine is diverging from the liberalizing European markets. Some power stations are able to compete on the Western Euro-

Ukraine

119

pean markets, due to low operating costs. But the restructuring of the sector, even though started on the institutional level, is not possible as a result of a severe lack of financial resources.

Market Structure In Ukraine there are four thermal power generation companies comprising a total installed capacity of 36.6 GW, two hydropower generation companies with a total installed capacity of 4.7 GW, and five nuclear power stations with total installed capacity of 11.8 GW. They deliver power to the 27 distribution companies and independent electricity supplier via the wholesale market. Figure 7.10 illustrates the vertical separation of the sector.

Demand Development Table 7.6 shows the demand development since 1990. Demand fell to a bottom level that is 63 percent of demand in 1990, due to the macroeconomical crisis of the country. Experts do not foresee any significant rise in near future, as a macroeconomical recreation cannot be expected, and Ukraine still belongs to the most energy-inefficient countries of CEE and, thus, has huge capacities for energy saving.

Capacity Development On the one hand, Ukraine has extra capacities (installed capacities are higher than consumption). On the other hand, there is a lack of fuel and money for equipment’s upgrade at the power stations. Thus, more than 30 percent of the capacities at thermal power stations are currently not in operation and need serious upgrade. The same problems apply to hydrostations. So, it is difficult to make any forecast of future capacity development. At present there are many projects for upgrade and modernization of existing power stations as well as for completion of construction of those stations, construction of which started in Soviet times.

Liberalization and Competition Formally competition has already started. The state enterprise Energorynok is responsible for the organization of the Wholesale Market. Ac-

120

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

ENERGY GENERATION COMPANIES (electricity generation)

WHOLESALE ELECTRICITY SUPPLIER Payment for electricity transit through electricity networks owned by Oblenergos

REGIONAL POWER DISTRIBUTION COMPANIES (Oblenergos) (electricity transmission, distribution, supply at regulated tariff, production to own)

INDEPENDENT ELECTRICITY SUPPLIERS (electricity supply at nonregulated tariff)

FINAL ELECTRICITY CONSUMERS Electric power sale/purchase contracts

FIGURE 7.10 The Structure of the Ukrainian Energy Market Source: Djafarova 2000

TABLE 7.6 Trends of Electricity Demand in Ukraine, 1990–1999 (in TWh) Year 1990 1991 1992 Supply 268,3 262,5 246,3 Losses 15% 16% 16%

1993 1994 1995 1996 1997 1998 1999 227,2 200,6 189,7 179,5 176,9 171,3 168,1 17% 18% 20% 22% 24% 25% 26%

Source: State Statistic Committee, Energy & Electrification Magazine

Ukraine

121

cording to legislation and a sector agreement all members of the Wholesale Market must sell their electricity produced or imported for the Ukrainean market exclusively on the Energorynok. There are just a few exceptions from this rule, mainly electricity produced for their own needs or in CHPs. On the Energorynok there are two groups of electricity generation companies. ■ The thermal power stations except for nuclear units perform their operations on the basis of price bids. These price bids include information about the price for the electricity delivered for startups and the operation at the point of idle running. The administrator of the Wholesale Market determines the equilibrium price. ■ The power generators that perform their operations on the basis of tariffs that are set by the National Electricity Regulatory Commission (NERC). These are the hydropower companies, the CHPs, and the nuclear power generation companies. NERC uses for tariff determination the companies’ costs, allowing a specific level of profitability. This level is currently set too low, which leaves the companies no financial resources for their modernization. In fact, some sort of competition exists only among thermal power generators. Tariffs for nuclear and hydrostations are set by the government. Distribution and supply of electricity are executed by Oblenergos’ being natural monopolists in their region. Independent power suppliers (IPS) existed by 2000. Currently due to the changes in legislation requiring 100 percent prepayment from IPS, all of them are out of business. So, there is almost no competition in all spheres except for generation. The main problem of the Ukrainian power sector that makes modernization and privatization very difficult is the extremely low liquidity. The level of cash collections has increased (from 12 percent in August 1999 to 60 percent in August 2000), but still is too low to create financial resources for a normal operation of the power companies. Furthermore, the tariffs can hardly be raised due to the resulting social problems and pressure on price stability. As an effect of low liquidity, barter trades dominate the whole Ukrainian energy sector and prevent efficiency gains from the modern structure of the electricity market.

122

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

Price Development Electricity tariffs for households are subject to regulation by the government, that is, their formation and establishing are far from the market priciples. Tariffs for the industry are being established on the basis of a “cost plus” method for each power distribution and supply company individually. It is difficult to foresee the development of prices for electricity in Ukraine. Mainly it depends on the results of privatization and increasing share of privately owned companies. For sure, the tariffs will grow but to an extent that doesn’t cause social problems and industry crisis. A growth of 20 percent–30 percent can be expected.

Privatization The privatization of the Ukrainian power sector has up to now been difficult to realize. The underlying reasons are poor technical conditions of the power companies’ equipment and the disastrous financial situation of the enterprises, due to low tariffs and cash rates. Nevertheless, the government plans to privatize all distribution companies until 2002. Until the end of 2000, 7 distribution companies could be privatized, but in each of them the state still holds 25 percent + 1 of the shares. The privatization for the nonnuclear thermal power stations is expected to start in 2001, but the state plans to retain a controlling majority (50% + 1 share) in each company. Hydropower stations, nuclear power stations, and CHPs are planned to remain state property (see Tables 7.7 and 7.8).

International Energy Trade Until 1990, Ukaine was one of the major energy suppliers of Eastern Europe and exported approximately 30 TWh/yr. This situation changed dramatically due to the political changes. Currently the energy system of Ukraine works practically in an autonomous mode even though interconnections to the neighboring countries on the 330–750 kV-voltage level exist. An operation of the energy system in a parallel mode with the CENTREL/UCTE systems will require a preparation time of approximately 10–15 years. Currently only some units at the power stations in Burshtyn and Dobrotvir work in parallel mode to CENTREL/UCTE. Thus, Ukraine exports mainly to its neighbors Slovakia, Hungary, and Poland.

123

Convergence versus Divergence

TABLE 7.7 Power Production Costs of Ukrainian Power Plants

Power Station

Power Production Costs [$/MWh]

Fossil Power Plants Burshtyn 23.47 Kurakhov 26.14 Ladizhin 23.62 Zuev 22.90 Krivorozn 24.55 Lugansk-8 24.55 Lugansk-2 29.21 Mironov-1 34.99 Mironov-2 32.38 Pridniprovsk-4-1 24.46 Pridniprovsk-4-2 26.99 Slavyansk-1-1 24.50 Slavyansk-1-2 26.99 Slavyansk-2 36.26 Starobeshevo 26.49 Tripoli-4 23.88 Tripoli-2 27.11 Zmiev-4 24.17 Zmiev-6 25.71 Uglegorsk-4 23.36 Uglegorsk-3 26.08 Zaporozhe-4 23.10 Zaporozhe-3 25.27 Dobrotvorsk-2 31.10 Dobrotvorsk-3 30.16 Average 25.08

CONVERGENCE VERSUS DIVERGENCE Table 7.9 tries to compare and contrast the differences in market developments in Poland and Ukraine. Poland’s membership in CENTREL already directly links their economy to the Western European electricity market. This factor seems to have reached a farreaching convergence to the European model that works as attractor for the whole region. Even though the existing formal and infor-

124

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

TABLE 7.8 Power Production Costs of Ukrainian Power Plants

Power Station

Power Production Costs [$/MWh]

Nuclear Power Plants Zaporozhe1 10.54 Zaporozhe2 10.54 Zaporozhe3 10.54 Zaporozhe4 10.54 Zaporozhe5 10.54 Zaporozhe6 10.54 So.Ukraine1 8.73 So.Ukraine2 8.73 So.Ukraine3 8.73 Rovno 1 9.46 Rovno 2 9.46 Rovno 3 9.46 Khmelnitsky1 9.46 Khmelnitsky2 19.48 Rovno4 19.48

mal institutions—such as LTPPA, lacking opening of the internal energy market to external competition, the unsolved problems with the restructuring of the coal sector, still existing cross-subsidies, lacking privatization— still do not support full convergence, the gap between the two markets (and market models) will certainly be overcome during the next few years. The presented description of the electricity sector of Ukraine proves that the formal and the informal institutions severely differ. The structure of the sector is built up according to the English power market and could serve as an incentive to a competitive market. But a dangerous macroeconomical environment, combined with the bad technical and financial state of the sector itself, make such a development impossible. ■ Taking into account the presented description, it is obvious that Ukraine is diverging from the Western (and Central) European market and market model, and has hardly any chance to overcome this situation. It is locked into a divergent development. Due to this situation, a further integration of the Ukrainian energy sector (for example by inte-

125

Convergence versus Divergence

TABLE 7.9 Poland versus Ukraine—a Comparison Poland Basics Institutional

Demand-supply framework

Effects

• relatively well-structured law • clearly structured market (vertical separation) • ongoing privatization • ongoing modernization • partially underfinanced companies • CENTREL/UCTE integrated electricity system • Rising degree of competition due to existence of Power Exchange, new market participants • Possibly collapsing electricity prices after internal and external opening of the electricity market • Price structure relatively free of cross-subsidies

Ukraine • relatively well-structured law • clearly structured market (vertical separation) • worn-out capacities • illiquidity, low cash rate • electricity system forced to work in autonomous mode • Metering systems do not allow real competition

• Market participants • Low electricity prices for social reason do not allow restructuring measure • Cross-subsidies in price structure

gration to the UCTE) cannot be expected. It thus will remain autonomous and not take part at progressive developments in neighboring countries like Poland, Slovakia, or Hungary. ■ Summarizing, 10 years after 1990 each CEE country has gone along its development path. At the beginning certainly existing ambiguity due to relatively close initial conditions of the further development has disappeared at least if one looks at the most prominent and extreme examples. ■ Unclearness only starts if one looks at less extreme countries like Bulgaria. The further development of their energy sector is probably strongly linked to the prospects of a soon EU membership. Interaction between endogeneous energy economical developments and exogenous

126

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

political decisions is possible and shows that these countries still are confronted (or have to count on) with a bifurcation point of their further development path—convergence versus divergence.

PROSPECTS FOR ELECTRICITY TRADE Convergence of the electricity markets will lead to their integration; divergence might lead to their disintegration. What will this finally mean for the development of electricity trade in the CEE region? Poland can be taken as a reference point for the whole region. No global player who wants to be active in Eastern Europe can afford not to be present in Poland, either by being a strategic investor in one of the power companies that are now on the way to privatization, or by trading electricity in and via Poland. On the other hand, Warsaw Power Exchange can serve as an indicator for the situation of electricity trade in Poland. There is currently hardly any turnover on PolPX due to battle between the still weak forces that fight for a real opening of the internal energy market (only formally it is already well-opened) and the incumbent nearly-monopolists that try to defend their situation as long as only possible. To the first ones belong the new market entrants mainly from abroad, like Vattenfall and Eon, but also those companies that would benefit from a fast opening of the markets (e.g., power stations like Rybnik that don’t have any long-term power purchase agreements). The latter ones are represented by the grid operator PSE that still is the controller of two-thirds of the power market by the LTPPAs and by those companies that benefit from those contracts due to their profitability. Nevertheless, it can be assumed that the still small free market will grow rapidly. This will be driven by the underlying fact that those who have to pay for the ongoing discrimination of new market entrants are the customers. A recent, not yet published, survey from Hungary proves for example that the biggest industrial customers pay significantly more for energy than their counterparts from EU countries like Germany. The same fact applies to Poland. Thus, in order to prepare the energy-intensive industries for the common EU market a fargoing liberalization is indispensable. This will lead to a relatively fast internal and external opening of the energy markets. The still existing problems on this way (e.g., lacking metering systems, etc.) will be solved quickly, especially if foreign capital is interested in this,

References

127

which will certainly apply for the candidates for quick EU accession. Nevertheless, it can be doubted that this will be the case in countries that are not on their way to integration, like Ukraine, as detailed and described previously. Lacking security of investments due to unstable frameworks, combined with the bad technical state of the infrastructure on all levels of the electricity sector, will prevent most market players from their entrance into that market. But without this the real liberalization impulse will remain too weak to create the conditions for electricity trade. So such countries can only serve, and only for limited time, as exporters of cheap electricity, produced in worn-out power stations. They will belong to an internal open and flexible electricity market only in the long term.

REFERENCES Arthur, W.B. “Competing Technologies, Increasing Returns and Lock-In by Historical Events.” Economic Journal 99 (1989), 116–131. Balmann, A., and M. Reichel. Pfadabhängigkeit and Lock-In. To be published in C. Herrmann-Pillath, M. Lehmann-Waffenschmidt, Handbuch zur Evolutorischen Ökonomik Heidelberg: Springer, 2001. Batov, S., and D. Popov. Restructuring of the Energy Sector in Bulgaria. (Vilnius: World Energy Council Regional Forum), 1999. Centre for Renewable Energy Sources (CRES), CEC/DGXVII. Investment Guide for the Energy Sector in Balkan Countries (Athens: CRES) 1998. David, P.A. “Clio and the Economics of QWERTY.” American Economic Review (Papers and Proceedings) 75 (1985), 332–337. Djafarova, Y. “Energy Trade in Ukraine.” Presentation, Power Bridge 2000 (Warsaw, 2000). Erdmann, G. Elemente einer evolutorischen Innovationstheorie (Tübingen, 1993). Erdmann, G. Ein nichtlineares Modell zur mittel-bis langfristigen Prognostizierbarkeit des Erdölpreises. In E. Fulda, and M. Härter (Hrsg.), Neue Ansätze der Prognostik (Frankfurt: Peter Lang, 1997), S. 103–131. Gros, D., and M. Suhrcke “Ten Years After: What Is Special about Transition Countries?” Working paper No. 56 (London: EBRD, 2000). Hannapi, H. Endogenisierung von Institutionen, Beitrag zur Jahrestagung des Ausschusses Evolutorische Ökonomik, Gesellschaft für Wirtschafts- and Sozialwissenschaften (Verein für Socialpolitik) (Osnabrück, 4.–6.7. 1996).

128

THE CENTRAL AND EASTERN EUROPEAN ENERGY SECTOR REFORMS

Hare, Paul G., and R. Davis (eds.). Transition to the Market Economy. 4 Vols. (London: Routledge, 1997). Jankauskas, V. “Unbundling, Privatization and Tariff Reform in the Energy Sector.” Proceedings of the conference: Lithuania: from Transition to Convergence, September 23–24, 1999 (Vilnius, Lithuania, 1999), 378–398. Kapala, J., V. Miskinis, U. Rudi, M. Caikovska, and N. Zeltins. “Actual Economic and Energetic Problems of Baltic Countries.” Latvian Journal of Physics and Technical Sciences (1999). Kolodko, G. Od szoku do terapii: Ekonomia i polityka transformacji, poltex (Warszawa, 1999). Lehmann-Waffenschmidt, M., and M. Reichel, Kontingenz, Pfadabhängigkeit and Lock-In als handlungsbeeinflussende Faktoren der Unternehmenspolitik. In T. Beschorner and R. Pfriem, Evolutorische Ökonomik and Theorie der Unternehmung, metropolis (Marburg, 2000), 337–376. Liebowitz, S.J., and Stephen E. Margolis. “Network Externatilities.” wwwpub.utdallas.edu/-liebowit/palgrave/network.html, vom 10.12. 1999, 21:07, veröffentlicht in: The New Palgraves Dictionary of Economics and the Law (Macmillan, 1998). Liebowitz, S.J., and Stephen E. Margolis. “Path Dependence.” wwwpub. utdallas.edu/-liebowit/palgrave/palpd.html, vom 10.12. 1999, 19:37, veröffentlicht in: The New Palgraves Dictionary of Economics and the Law ( Macmillan, 1998). Mayhew, Alan. Recreating Europe: The European Union’s Policy towards Central and Eastern Europe (Cambridge: Cambridge University Press, 1998). Myant, Martin (ed.). Industrial Competitiveness in East-Central Europe (Cheltenham: Edward Elgar, 1999). OECD. The Competitiveness of Transition Economies (Paris: OECD, 1998). Reichel, M. “Realisierbarkeit der Potentiale von erneuerbaren Energien— Zur Relevanz von Lock-Out—Effekten für die Markteinführung der Photovoltaik—and Windenergienutzung.” Dissertation TU Dresden 1998, erschienen unter dem Titel “Markteinführung erneuerbarer Energien” (Wiesbaden: Gabler, 1998). Riesner, W. “Entwicklungstendenzen in den Energiewirtschaften der EU— Beitrittskandidatenländer Mittel—and Osteuropas.” In: Zur deutschen Energiewirtschaft an der Schwelle des neuen Jahrhunderts, Schriften-

References

129

reihe des Instituts für Energetik and Umwelt, B. 6 (Stuttgart-Leipzig, Teubner-Verlag, 2000). Riesner, W. “Results of an International Survey.” (Zittau, 1999, unpublished). Riesner, W. Proceedings of the 10th Zittauer Seminar. Zittau, 2000. Witt, U. (1997): “‘Lock-In’” vs. “‘Critical Masses’”—Industrial Change under Network Externatilities.” In: C. Antonelli, and P.A. David (eds.), The Economics of Path-Dependence in Industrial Organization. International Journal of Industrial Organization 15(6), 753–773.

CHAPTER

8

An Italian Road Map to the New Energy Markets Alessandro Mauro* Risk Analyst, Energia SpA

INTRODUCTION The European Union is currently in the midst of a liberalization process of its energy markets, aiming to introduce competition and free flows of gas and electricity, as has already happened for many goods and services in EU member countries. These transformations were introduced by two European Energy Directives whose purpose was to shape the general features. Nonetheless, each country’s market has distinctive features that are reflected in the current market environment and changes taking place. Italy adopted the European rules by two EU legislative decrees, implemented in February 1999 (electricity) and August 2000 (gas). As for the overall EU, the process will transform the industry from a monopoly into a free market. The following pages introduce features specific to the Italian environment, the shape of its new energy markets, and the business model companies are currently adopting.

TURMOIL IN THE ELECTRICITY SECTOR In 2000, Italian net electricity production amounted to 253,000 GWh that, along with 45,000 imported GWh, satisfied a consumption of 298,000

*The author wishes to thank Regina Jain for her assistance.

131

132

AN ITALIAN ROAD MAP TO THE NEW ENERGY MARKETS

GWh. Until 1999 Enel, the integrated state-owned monopoly created in 1962, controlled about 80 percent of the total generation capacity. The liberalization decree established the complete unbundling of the monopoly, specifically between the production and distribution arms. Moreover, in order to promote competition in power generation, Enel has been required to sell at least 15,000 MW of installed capacity to other operators by 2002. These MWs have been given to three new generation companies (Gencos): Eurogen (7,400 MW), Elettrogen (5,500 MW) and Interpower (2,600 MW); in 2001 Endesa of Spain won the first bid and bought Elettrogen. Thanks to these and other smaller sales, plus the forecasted increase in Italian installed capacity, Enel expects its production share to drop below 50 percent by January 1, 2003 as stated by the decree. Thanks to a revamping or repowering program, the older generation units belonging to the three Gencos (electricity generators) will be transformed into modern combined cycle gas turbine (CCGT) units. Moreover, the majority of genuine new capacity forecasted in the next years is expected to use CCGT technology, resulting in an overall boost in future gas demand. As far as electricity distribution is concerned, in 1962 Enel was granted a legal monopoly of electricity sales to final customers. Only some public LDCs (local distribution companies) , owned by municipalities, retained similar monopoly rights in their towns, mainly located in northern Italy. Today the liberalization decree states that customers consuming more than 9 GWh/y are considered eligible, meaning that they are free to choose their electricity supplier; moreover, provided that they form a consortium company, the minimum single consumption becomes 1 GWh/yr. The current size of the eligible market is more than one-third of total demand. On the other side of the market there are non-eligible customers, smaller consumers that will not be allowed to choose the electricity supplier they prefer until 2030. They will instead be supplied at regulated tariffs by the Single Buyer, a new body designated to procure electricity for their needs. At the dawn of liberalization, the first eligible customers were offered electricity with large price discounts. Unfortunately, this process met a bottleneck when non-Enel capacity plus limited uncontracted importation capacity from neighboring countries were no longer ample enough to supply all eligible demand. The main consequences were prices that soared to previous monopolistic levels and a general loss of confidence in the final results of the ongoing liberalization process. During 2002, an electricity pool will be created in order to concentrate

Turmoil in the Electricity Sector

133

short-term physical exchanges. The market will be centrally dispatched according to a merit order: The electricity will be produced by generators that present the lowest price bids. The electricity will be sold to the Single Buyer (i.e., wholesalers, traders, and eligible customers). However, the liberalization decree does not forbid operators to bypass this auction process and instead sign bilateral contracts for long-term electricity supply. Apart from these physical exchanges, there are plans to launch a financial electricity market, a place where operators will be able to hedge, arbitrage, or speculate on electricity prices. In order to foster the development of a broader competitive environment, the government has decided to create new eligible customers by lowering the minimum consumption threshold to just 0.1 GWh/yr from 2003. Anyway, the main problem will continue to be the lack of sufficient non-Enel electricity. As a byproduct of this limitation, it seems quite probable that the bidding process in the pool will not be really competitive initially, as Enel will retain relevant market power even after the Gencos sales, slowing down the desirable price reduction process. In order to address this problem, it is generally believed that new structural measures are needed, mainly the increase of connection lines with other countries and the simplification of rules to build new power generation plants. Notwithstanding, the market is seeing new entrants from Italy, other areas of Western Europe, as well as the United States. They are positioning themselves at all levels of the electricity chain, as producers, wholesalers, and traders; consequently, their business strategies are quite differentiated. Currently, the most active among them are the producers, and their first short-term objective is to buy one of the Enel Gencos. In this regard, the common strategy has been to create joint ventures among companies, including Italian and foreign players. At the same time, many are considering building new green-field or existing brown-field capacity. Among newcomers is Eni, the Italian oil and gas major, which is setting great efforts on the revamping and repowering projects of some already owned electricity plants. The plants were endowed to Enipower, a newly formed company, in order to reach 5,000–7,000 MW of installed capacity by 2005 to be run as merchant power in the electricity pool. The explicit aim of Eni is to add the transformation of gas into electricity to the value chain but at the same time, allow an easing in the limit that has been set in the Italian gas market, as is noted in the next paragraph.

134

AN ITALIAN ROAD MAP TO THE NEW ENERGY MARKETS

SUNRISE ON GAS LIBERALIZATION Domestic gas consumption amounted to 69 Bcm (billion cubic meters) in 2000, with indigenous production accounting for only 22 percent, the rest coming mainly from Algeria (38%), Russia (28%), and the Netherlands (8%). As demand is expected to rise to 82 Bcm in 2005 and 89 Bcm in 2010 and national production continues to decline from the second part of 1990s, new importation infrastructures are needed. Eni is completing a new pipeline from Norway, and starting to build an underwater line linking Libya with Sicily. Also LNG regasification terminals will play an important role: The Italian Edison will build a plant in conjunction with Exxon–Mobil in the northern Adriatic Sea, and British Gas is planning to construct another one in southern Italy. Potential sources of incremental LNG supply are Algeria, Egypt, Nigeria, and the Middle East. This probable scenario would be altered if Eni were to be allowed to develop relevant gas reserves in the northern Adriatic Sea, currently frozen due to environmental concerns. The gas liberalization decree was passed in May 2000. It stated that all customers consuming more than 200.000 cubic meters/year are considered eligible and therefore free to choose a gas supplier. Newcomers have indeed the right to access the high-pressure gas transmission system, mainly owned by Eni, and supply these big customers. This right is also granted for LNG regasification terminals and storage infrastructures. As of January 1, 2003 all customers will become eligible and the right of access will be consequently extended to medium- and low-pressure pipes. In order to promote effective competition, the liberalization decree sets two antitrust limits. From 2003 no one will be allowed to sell more than 50 percent of the total annual gas consumption. Moreover, from 2002 it is forbidden to introduce into the high-pressure transmission system more than 75 percent of total annual consumption; the limit will be lowered by two percentage points each year until it reaches 61 percent. This second prohibition is aimed to increase the availability of gas to new operators, at the same time reducing the monopolistic position of Eni in gas production and import. Given current forecast on future consumption, domestic production, and take-or-pay importation obligations, the limit is likely to have a real impact on Eni gas activities. Taking into account the fact that Eni can ease the limit by burning gas to produce electricity in plants owned by the group, it will anyway have to release or divert some Bcm/yr in the short term and more at the end of the decade. Eni has already moved to cope

Sunrise on Gas Liberalization

135

with this problem, selling all its Libyan equity gas and the contracted gas coming from Norway. Turning our attention to downstream in the gas chain, we must say that at the dawn of the gas industry in Italy, municipalities reserved gas distribution and sales in urban areas to public or private LDCs through concessions, securing a local monopoly for each of them. The result is that nowadays it is possible to count more than 700 gas LDCs, many of them very small and quite inefficient. In this field the liberalization decree introduced the unbundling of gas distribution and sale activities from January 1, 2002. Distribution, which is the management of the pipeline network, will have no competition within the market but there will be competition to enter the market. In fact, new concessions will be granted with tenders for a maximum period of twelve years, while existing ones will be reserved to incumbent LDCs only until 2012. Another measure aimed to foster competition is the public ownership of the network, which will reduce financial expenses for new entrants. Nevertheless, in the short term, it seems that the only way newcomers can enter the distribution sector is through the acquisition of already existing private LDCs, considering that publiclyowned LDCs are not yet on the market. The gas sales business, on the contrary, seems to offer bigger opportunities. Some Italian firms, like Edison, Enel, and Energia, have started to win contracts for big industrial customers to the detriment of Eni, which retained a de facto monopoly in high-pressure deliveries. Moreover, many Italian LDCs, as well as some startup trading companies and foreign newcomers are moving their first steps. The most urgent requirement for them is to find a source of gas at a reasonable price, in order to be able to initiate fierce competition for big customers and, from January 1, 2003, for commercial and residential ones. It has to be considered that some entry points in the transmission system could be congested and, in this case, capacity auctions should be introduced. Moreover, if they acquire gas produced outside the European Union, they will need an authorization from the Ministry of Industry. All new and existing non-EU gas imports will have to secure 10 percent of yearly total imported quantity as compulsory strategic storage in the Italian territory. Finally, the liberalization act made no provision for either the creation of a gas wholesale market or the identification of a physical or virtual trading hub. This could prevent spot transactions and slow down the abandonment of oil-indexed pricing; nevertheless, they are both very probable in

136

AN ITALIAN ROAD MAP TO THE NEW ENERGY MARKETS

the medium term. Currently, some operators are already trying to organize gas deals at the borders and short-term LNG imports. Moreover, Eni gas storage infrastructures are already managed using a virtual hub for injections and withdrawals.

THE ITALIAN MULTI–UTILITY MODEL Throughout the world, wherever there has been an unbundling of a vertically integrated energy supply chain, the incumbent monopolist has tried to recreate opportunities of value creation providing new products and services at different levels of the chain. This has been an obvious response to the forced abandonment of cross-subsidization practices, replaced by the exploitation of scale and scope economies in liberalized markets. Currently, this phenomenon of horizontal integration is very evident at the bottom of the chain, that is at the distribution and sale levels, where operators are striving to shape themselves according to a new business model by now known as multi-utility. Many Italian firms are following with determination this general trend. Enel is probably the most convinced of them; considering its recent shopping list, it nowadays controls the biggest Italian (and European) water system, gas LDCs, a pay TV channel, the third Italian mobile telephone operator, and the second Italian fixed telephone company. The declared objective of the firm is to diversify its activities and widen its geographic extent with the aim to focus on the relationship with final customers. Eni, currently strengthening core upstream oil and gas activities, has never been a utility company. Regardless, it currently has interests in telecommunications and it controls Italgas, the biggest Italian gas distributor selling near 8 Bcm/yr, that is now active in telecommunications and water sectors. Edison, the biggest Italian private energy company producing oil, gas, and electricity, is diversifying itself toward energy distribution, offering also telecommunications products and acquiring stakes in water management systems. Besides the current moves of these three big companies, public LDCs offer various examples of already formed smaller-scale multi-utilities. As already stated, their existence relied on the possibility of producing and selling electricity even after the creation of Enel. However, many of them also distribute the gas they buy from Eni to final customers, while others are active in water services and waste management. Recently, the largest public LDCs have been transformed into joint stock companies and, even if the majority is still publicly owned, they are expected to be more business oriented and are

The Italian Multi–Utility Model

137

preparing themselves for the new competition that will arise from the longawaited liberalization of local public services. In the new energy markets, they are forming trading branches to tackle eligible market shares. As far as private LDCs are concerned, their principal business is gas distribution, but many are also operating water services and waste management; as already stated, local electricity distribution is currently forbidden. Apart from Italgas and a few others, the majority of them are too small to compete in new markets; recently, they have begun to merge or form some kind of aggregation in order to increase their market power; this has been the case of Gas IT, Plurigas, and Blugas. Currently Enel, Edison, and others are striving to form joint companies with LDCs, in order to shape local multi-utility ventures selling electricity and gas and adding telecommunications products, especially in broadband applications (e.g., pay-TV, video-on-demand, etc.). The advantage of these joint ventures seems to rest on the possibility, for all parties, of preserving their own customer bases while entering new markets and offering new products, bridging the gap represented by the current lack of technical capabilities and financial resources. Nonetheless, LDCs are also moving on their own, and the preferred field is again telecommunications. Many are laying down local networks of fiber optics, while others are opting for wireless solutions. Other players think it will be better to focus on the sales of products using the existing telecommunications infrastructure. Multi-utility companies, both public and privates ones, are fostering another structural change, often referred to as convergence among markets. Markets’ extents are widening and energy-to-energy competition is appearing also in Italy; Italian electricity is competing with imports and the same is happening among imported natural gas. Energy markets are also becoming more interrelated. Upstream in the chain, gas will be the primary fuel of new power generation capacity; generators will soon be able to choose on a daily basis whether to sell gas or burn it to sell electricity, and they will be offered interruptible contracts. This will, in turn, give rise to an increase in price correlation and the creation of new financial instruments, like the spark spread that will enable operators to hedge, arbitrage, or speculate against the evolution of gas–electricity relative price movements. Downstream in the chain the convergence phenomenon is also self-evident; operators have started to sell both electricity and gas as bundled products. This strategy rests on the possibility of offering, in addition to traditional products, a broad array of services; it is through the bundling of

138

AN ITALIAN ROAD MAP TO THE NEW ENERGY MARKETS

useful and reliable services that utilities will add new value in the future. During past years we have seen the offering of the first simple services, as traders and wholesalers deliver data on consumption, offer energy saving management and e-billing services. They share common features, such as being mainly distributed via the Internet to large customers and representing typical business-to-business applications. These kinds of customers will soon adopt e-procurement systems, and energy providers will have to be prepared in advance. The name of the game is, again, new telecommunications solutions; they represent new services and, at the same time, the new mediums through which both old and new products are delivered. Regardless, it is not true that all future strategies will rest on telecommunications; an example of other innovative services is presented in the next paragraph. Residential customers form, at least for now, a totally different market. In Italy, the Internet is not so widespread, and e-commerce is an uncommon practice. In the near future, energy sales practices may more closely resemble those common in the retail sector. In 2003, gas will be completely liberalized and it will not be a long time before we see special offers in department stores. Obviously, the future of this demand segment will rely partly on the Internet as well. There are good outlooks for the success of e-energy commerce, given that energy supply belongs to that category of products possessing standardized features, such as books or airline tickets, that were the first successful products sold on the Internet.

NEW RISK MANAGEMENT PRODUCTS Today it seems that the most important risks new operators are facing are structural, undermining the very existence of their startups. They risk being squeezed between supply and demand conditions, especially in terms of prices and/or quantities. For example, many traders are negotiating considerable quantities of gas that will come from outside Italy, in take-or-pay terms for decades, with no customer base at the moment. Moreover, they are agreeing on prices without having a clear picture of the competitive environment that will emerge in the future. Consequently, it is essential for them to acquire market share as soon as possible, and the only way to achieve the goal is to adopt the best marketing strategy. A cheaper price will not be the unique or best lever. It will

New Risk Management Products

139

have limited effectiveness, as others will simply replicate the move and depress profits. As already stated, the right approach rests on the possibility to offer new value-added services in conjunction with energy supply. In the last paragraph we gave a general introduction to the theme; here we want to focus on something more specific, linked to the large amount of price risk currently inherent in Italian energy supply, whose prices are almost totally linked to oil price movements. In this context, Italy shares the weakness of other hydrocarbon importing countries, regarding the prices of gasoline, gas oil, and other fossil fuels. Moreover, gas prices are set as formulas indexed to oil and refined products prices, due to the lack of a liberalized market. If you also consider the impact of the euro/dollar exchange rate, the result is a nightmare. Nonetheless, the position of Italy is even more serious, as thermal power generation utilizes only hydrocarbons; in fact, a referendum in 1987 banned the use of nuclear power. Therefore, electricity tariffs fully reflect both oil price movement and the evolution of the euro/dollar exchange rate. In the future a great challenge for suppliers will be to transfer a part or all of this price risk from customers to themselves or to third parties. This move will represent a winning strategy only if it is included in a comprehensive risk management procedure: risk assessment, risk measurement, and risk control. To consider an example of such a strategy, a supplier could sell gas to a power generator with innovative pricing. The new price could consist of the current formula price only up to a ceiling price; whenever it is exceeded, the customer will pay only the ceiling price. It is obvious that the supplier is taking risk upon himself and must be aware of what kind of product he is offering. This is considered selling an exotic option, that is, an option where the underlying right is complex and difficult to evaluate. In fact, the existing formula price is calculated taking the arithmetic mean of oil, gas oil, and fuel oil prices for the previous three months. Therefore, we have at the same time a basket option and an average option. In order to know the fair value of this kind of insurance sold to the customer, the famous standard Black–Scholes model is useless. More sophisticated approaches are needed, like Monte Carlo simulation. After calculating this fair value, it will be a marketing choice whether to invoice it to the client or not; in the second case the seller must decide whether to bear the risk directly, to hedge in the financial market, or to transfer it to a financial institution, paying a price.

140

AN ITALIAN ROAD MAP TO THE NEW ENERGY MARKETS

Besides valuation, another necessary step is to assess the amount of risk concerning the ceiling contract. This, in turn, implies that a general risk measure must be chosen. Currently, the most fashionable is Value-at-risk (VaR), which reports the maximum expected monetary loss on a risky position within a certain period of time, at a certain probability level. As for valuation, it is impossible to use a straightforward closed-form VaR approach. Instead, the construction of a Monte Carlo VaR model is required; it will simulate the real future evolution of prices and calculate the total Value-at-risk of the contract. Figure 8.1 shows the monthly VaR numbers from month m through month m+6, calculated with a ceiling price of 14.5 euro cents for every cubic meter of gas sold, when the actual gas formula price was slightly lower. Obviously, price risk is increasing over time as uncertainty does the same. It is also worth noting that the absolute risk level is quite considerable, being the ceiling price chosen in this contract probably too low. Anyway, a broad set of alternative choices is available: establishing a higher ceiling price, sharing the risk with other parties, or hedging on the financial market. The answer must come from a sound risk assessment procedure and this, in turn, will require the explicit choice of a general risk management system. We have kept things very simple, in order to show the result of a simple single move. But the real future environment will be more complex. For

4.40

5 3.71

(euro cents / cubic meter) 4 2.90

3 2.03

2 1.10

1 0.02

0 m

m+1

m+2

m+3

m+4

m+5

Months

FIGURE 8.1 VaR at 95% Probability Level (Euro Cents/Cubic Meter)

New Risk Management Products

141

example it must be recalled that the gas buying price for the supplier is also indexed. The supplier will have to consider all single exposures jointly, in order to calculate the total price risk. Moreover, a greater price risk is likely to emerge for electricity, as price volatility in the pool is expected to be higher than in the case of the gas formula. In the near future, the pool will be the center where risk management will be developed in Italy and it will represent one of the hottest issues in coming years.

CHAPTER

9

Freight Trading: The Emerging Commodity Market Kirk H. Vann CEO, Freight Advanntage and Advannce Energy

INTRODUCTION Transportation has been an integral part of commodity trading for hundreds of years. The cost of transportation is one of the main factors in arbitrage which is profitability moving a commodity from one location to another. The early trading of coal, grains, and other dry cargos identified freight as the major factor impacting cargo profits or losses. Other commodity markets have realized the advantage of arbitrage, and in certain cases such as electricity, natural gas, rail, or trucking have established that the transportation component of the commodity price can be traded independently of the commodity. In effect, another commodity has been created. In the case of oil products, however, freight, for a variety of reasons has been viewed as a service rather than as a tradable commodity. It seems likely that with the advent of Business to Business (B2B) exchanges on the Internet that freight will become more visible and ultimately more liquid. These changes will promote freight trading with freight emerging as a true commodity market. The same financial tools that have been used to revolutionize other commodity markets are slowly being introduced to the freight industry as futures, swaps, options, and other financial instruments. Freight will ultimately succumb to the same market forces that have

143

144

FREIGHT TRADING: THE EMERGING COMMODITY MARKET

changed all other commodities. It should be remembered that oil was considered a true commodity until the emergence of OPEC’s power in the 1970s, and the subsequent start of energy futures contracts on the New York Mercantile Exchange beginning in 1978. These two factors alone have made the oil market both visible and have promoted price volatility. These changes in turn have prompted the creation of more sophisticated risk management by the NYMEX, International Petroleum Exchange (IPE) and the Over-the-Counter energy markets from their inception to the present. Oil is now one of the most liquid and highly tradable commodities in the world, and has been developed in a relatively short period of time. Freight will soon follow these commodity market developments in the energy complex.

MARITIME INDUSTRY OVERVIEW Until recently, the maritime industry has been virtually unchanged from the days when ship captains were “principals” in merchant companies hundreds of years ago. The major changes were mostly in communications (i.e., from cable/telex to phone/fax to satellite/e-mail). B2B now becomes the next step in this technology process and will be an integral part of the evolution of freight as a commodity and the subsequent development and acceptance of risk management tools. The emergence of middlemen such as brokers, agents, and other service providers has contributed to the fragmentation of the shipping industry creating inefficiencies. Conversely, B2B will promote efficiency and centralization, and ultimately commoditization. Freight has been considered a service rather than a commodity for a variety of reasons. The two most common reasons are environmental concerns, which is a political minefield in most companies and the vetting process that determines which ships are acceptable or unacceptable for service. The insinuation is that no two ships are alike and thus do not meet the definition of a commodity. Environmental concerns are obviously important, and fortunately technology and environmental directives are easing some of these concerns, but more importantly, having freight treated as a commodity will not affect its environmental importance. In fact, it may highlight environmental issues further as they continue to gain more visibility.

Maritime Industry Overview

145

The argument of vetting is overstated since vessel particulars generally known as “Q-88” data, regarding size, single or double hull, age, vessel dimensions, and equipment can be classified relatively easily into distinct categories replicating a standard commodity. Today, this process is generally done by brokers, charterers, and ship owners who identify how many vessels of a certain category are available for shipping dirty cargoes such as Panamax, Aframax, Suezmax, or VLCC. Similar categories exist for clean tankers. The various quality differences within these categories can be handled quite easily and therefore would allow them to be traded as a commodity. In addition, various shipping routes are recognized as standard, such as Caribbs–USGC/USEC or WAF–USGC/USEC and have in fact been identified in the development of the swaps markets that are discussed later in this chapter. So what does all this mean for the maritime industry? B2B will eventually provide the visibility and subsequent liquidity to expose the inefficiencies of the industry. This will promote freight trading and the emergence of market makers to take the inefficiencies out. Finally, the further development of more sophisticated risk management tools will complete the evolution. The question is not if but when and how long will this take? Before we talk about current progress in freight trading, derivative development, and risk management, let’s define a few critical concepts as simply as possible. arbitrage The ability to profitably move a commodity from one location to another. Freight is usually the key component. commodity An article of trade or service that can be bought and sold. futures contract A supply contract between a buyer and a seller, whereby the buyer is obligated to take delivery and the seller is obligated to provide delivery of a fixed amount of a commodity at a predetermined price at a predetermined location. Futures are traded exclusively on regulated exchanges like the NYMEX, IPE, and Baltic. liquidity The relative ease or ability to buy and sell. If it is relatively easy, then it is a liquid market; if not, then it is an illiquid market. options A contract that gives the buyer the right but not the obligation to buy or sell the underlying commodity at a specified price within a specified time in exchange for a one-time premium payment. The contract also obligates the writer/seller, who receives the premium, to meet

146

FREIGHT TRADING: THE EMERGING COMMODITY MARKET

these obligations. Options can be bought and sold on a regulated exchange or over-the-counter. risk management The combination of futures, swaps, options, and physical tools on the underlying commodity to maximize profit potential and minimize loss potential. Hedging is a way to manage basis risk, which is the uncertainty of time, location, or quality. swap A customed-tailored, individually negotiated transaction designed to manage financial risk over a specified period of time. The swap writer/seller assumes the market risk and the swap buyer fixes their risk. Price discovery and payments are based on an agreed to market index by both parties. All swaps are over-the-counter and may involve principals, banks, and/or brokerage houses. volatility The market’s price range and movement within that range. Historical volatility refers to how the market has traded in the past, while implied volatility is an indication of perceived future trading ranges.

FREIGHT MARKET PRICE VOLATILITY Every commodity has price volatility; however, the range of that volatility depends on several factors specific to that commodity. Freight is very volatile with historical price volatility of 50% to 75% on an annualized basis under normal circumstances, but it can soar to 200 percent to 300 percent during special events. Major factors affecting freight are weather, oil economics, political tension, environmental factors, and supply/ demand. Some of these factors are more predictable than others, particularly seasonal factors, but in all cases, risk management techniques can be used effectively. Time chartering is the leasing of a tanker to a third party for a specific period of time at a specified rate and is the only mechanism currently used to fix income. Ship owners get utilization of their vessels at a specific cash flow but do not participate in any upside gain if market conditions change. However, they can buy bunker swaps or fix physical bunkers to set an operating margin in certain cases. The third party who leases the vessel is satisfied as long as the spot market price remains above the current time

Current Risk Management Initiatives

147

charter rate, but they lose opportunity margin if the spot market price falls below the time charter rate. The problem facing the freight industry today is that there exists much price volatility and risk, but there are ineffective methods to manage both. Several companies have recognized that the skill sets in freight should be the same as in oil trading, and in some cases have promoted oil traders to head freight groups. However, the purely marine and operational jobs are still staffed with traditional thinking personnel. Many in the freight industry argue that freight is too unique a service to be commoditized. They need only to look at the emerging weather derivatives market to see that if weather can be hedge, so can freight.

CURRENT RISK MANAGEMENT INITIATIVES More changes have happened during the past 5 years to make freight a visible and tradable commodity than in the previous 500 years. Various aspects of the changing landscape of the maritime industry will now be examined. In the over-the-counter swaps markets, ship brokers Mallory, Jones, Lynch, and Flynn in conjunction with Citibank have developed a freight swap based on various standard routes and vessel sizes in order to provide financial hedging for ship owners and ship charterers. While this new risk management tool has had limited success to date, it still provides a financial hedge to lock in attractive fright rates for ship owners or charterers. The major problem has been that ship owners are always optimistic that freight rates will be higher at the same time as the charterer feels just the opposite. So far, the amount of active participants has been the limiting factor in market development although abnormally high freight rates during 2000 and 2001 put new emphasis in this area. Price swaps are much more effective if there is an active futures and OTC derivatives markets. In April 2000, SSY Futures of ship brokers Simpson, Spence, and Young, launched the freight industry’s first on-line trading facility providing a trading platform specifically tailored to the dry bulk derivatives market. Forward freight agreements (FFAs), are principal-to-

148

FREIGHT TRADING: THE EMERGING COMMODITY MARKET

principal OTC derivative swap contracts using as their reference, the daily dry-bulk freight indexes of the Handy, Panamax and Capesize shipping markets. Finally, Imarex, a Norwegian-based group, launched its tanker swaps exchange during the fourth quarter of 2001 backed by guaranteed settling and central clearing. The Oslo Stock Exchange’s derivatives arm “NOS” will act as a central counterpart and Den Norsk Bank, the worlds largest shipping bank, will provide clearing services. Other risk management tools are available to the maritime industry and include bunker swaps, currency swaps, and interest rate swaps. Moreover, a ship owner’s bank can lock in operating margins, similar to the crude oil crack spread. In this case, the ship owner could sell a freight swap and buy a bunker swap to lock in their margin, in effect, hedging the freight versus the fuel price risk. Today, these instruments are used by many banks, ship owners, and charterers. Options offer another risk management technique. For example, a physical freight option, which allows the buyer the right, but not the obligation, to a tanker use in the future at a specific rate for a specific route. These financial tools have had limited success primarily due to the complexity of understanding their application with current personnel in the maritime industry. This situation is changing as more traders enter the freight business. Another example of a freight option would be for a charterer to buy an option to lock in economics on a term deal or spot tender in order to guard against a potentially rising market. On the other side of this equation, an owner could sell the option and collect money for cash flow in a stable market or could buy a bunker swap and lock in the operating margin. Small ship owners could also buy options to supplement their fleet sizes when needed. Options also provide leverage and an ability to take a much larger position with the same amount of capital outlay versus a futures, swaps, or a physical position. Typically, an options strategy will have financial leverage of 3–10:1 times the underlying position versus the futures, swaps, or physical position, which is usually leveraged at 1:1. For example, a charterer could buy several vessel options in a particular location and give the impression to the market that the “market is tight” because the ship owner (s) who sold the options cannot show these vessels to the market, since the buyer may exercise their options and take physical de-

Current Risk Management Initiatives

149

livery of the vessels. As the market rises, the buyer can exercise their options, and take the physical vessel (s) or sell back their options at a profit. It should be noted that the ship owner(s) will probably benefit from the stronger market. Ship owners can do something similar via selling covered calls on their fleet at progressively higher rates. The market will perceive this as an indication of strength and prices will climb higher. If they have options on 4 of 10 ships in their fleet, ship owners may lose on one or two options but gain on the balance of their positions in the fleet through financial leverage. Turning to financial futures, ship brokers Simpson, Spence and Young (SSY) are currently working on developing a futures contract to be cleared through the Baltic Exchange and perhaps other futures exchanges in the future. Both the NYMEX and IPE have studied adding freight futures to their lists of futures products. The use of commodity futures contracts are well documented and simply allow buyers and sellers to lock in a price in the future for a certain commodity at a specific location. Freight futures in conjunction with swaps and options make risk management viable in the shipping industry. Turning to B2B platforms, the past year has produced many ventures bringing Internet applications to the marine industry. ShipIQ, for example, offers end-to-end solutions for the marine industry including links to the oil side of the transaction, such as on-line chartering, pre- and post-fixture reporting, and other related services like agency, inspection, and demurrage. While there will be stiff competition and certain consolidation in this space in the future, the one certainty is that B2B is here to stay, and that it will provide visibility to the marine industry. This visibility will magnify the inefficiencies and current fragmentation of the industry, and encourage freight trading and market making as a means to take advantage of changes underway in the maritime industry. Over time, more sophisticated risk management tools like swaps, options, and futures contracts will develop. These normally trade at 10 times the physical volume. Moreover, voice brokers will survive but their roles will change as they are affected by the B2B shift. One other interesting development resulting from the B2B shift is the alliances being created between ship owners or charterers. Both these groups realize the economies of scale available by pooling resources to create more powerful entitities. Tankers UK is one example of a consortium of

150

FREIGHT TRADING: THE EMERGING COMMODITY MARKET

owners in the VLCC market. Another consortium of ship owners and charterers is One Seas/Levelseas owned by Shell, BP, Cargill, Clarksons and supported by many other ship owners and charterers.

Barriers to Success This is not a question of if but when these changes will occur. The middlemen of the industry will continue to keep the status quo and derail any attempt to change the current business relationships since they are the ultimate losers of this change process. On the other hand, the charterers, until they are compensated for taking risk and are forced to change their skill set will be slow to embrace change. Ship owners, who continue to take huge risks, will follow suit when they are forced to compete or follow the wishes of their clients. While this may sound like a catch-22 situation, changes will occur if we look at oil market evolution. The key to unlocking liquidity in the freight business will be the B2B platforms. Some estimates as high as 50 percent to 75 percent of all chartering within three years will be B2B. The implication for the freight industry is clearly that freight trading will now be visible to a larger population. Secondly, since freight is the largest component of arbitrage and trade economics in oil markets, oil traders need to be enabled to make freight decisions independently, then further market liquidity will develop. Already several shipping companies have put oil traders in senior positions and have crated freight trading departments and freight profit and loss centers. It should be remembered that all freight departments are judged on how they charter shipping versus the spot market, which is usually two weeks out in the future. By definition, this almost insures that they will always be judged favorably. The real question to be asked is “What do they do when the risk was taken?” For example, if a trader wins a spot tender 90 days in the future using a freight department estimate of Worldscale 150 (a standard tanker rate) for the bid, but by the time the cargo loads, it has jumped to Worldscale 300, then the trader is probably going to lose much money for the company. However, the freight department will be judged on how they did their chartering in the spot market (i.e., if Worldscale 300—and maybe lower if they are a major player). Management will be happy with

Current Risk Management Initiatives

151

the freight department and critical of the trading department when in fact the freight department allowed the market price to double without any proactive risk management activity. Market makers will enter into the freight market, and have a similar impact that Wall Street had on the oil markets in the 1980s. They will capture inefficiencies of the bid versus ask. The speed at which freight develops into a liquid commodity market is dependent on the factors explained in this chapter, but in the next few years that change will occur. A dramatic shift will occur and we will look back and wonder why it took so long for freight to emerge as a tradable commodity.

CHAPTER

10

Market Risk in Electric Generation Finance William A. Klun Vice President, DZ Bank, AG*

INTRODUCTION This chapter outlines the key issues that characterize market risk for electric generation suppliers in the competitive electricity markets in the United States, one of the prominent issues being potential overcapacity. These issues form the key market risk considerations for investors and financiers involved in determining the underlying credit quality of electricity generation assets and the cash flows derived from these assets. The market issues reviewed pertain to both a single asset and a portfolio of assets in a single market (or multiple assets in multiple markets). The market issues which effect the competitiveness (hence credit risk) of electric generation assets include ■ ■ ■ ■ ■

The composition and growth of demand The composition and growth of new capacity Supply and demand equilibrium Generation economics and asset displacement Market structure and risk

*This chapter reflects the views of the author and does not represent any opinions, recommendations, or positions of DZ Bank.

153

154

MARKET RISK IN ELECTRIC GENERATION FINANCE

The key conclusions of this analysis are summarized as follows. ■ Electricity demand consists of stable and highly volatile components. The risk criteria for evaluating generation serving stabling demand (such as in the contract market) are different from that of assets serving the commodity market (such as “merchant” power producers). ■ Electricity demand is only partially a function of overall economic growth. It is also a function of the structural change of the U.S. economy to service industries. The applications for the service sector (heating and cooling) is far different from that of the industrial sector (process uses). ■ New generation capacity will be limited by the infrastructure considerations in the natural gas sector. These limitations very clearly show that total capacity cannot feasibly approach the inclusive overbuilt scenarios suggested by capacity announcements. ■ Financing risks resulting from overcapacity vary among the regional U.S. electricity markets. While some markets will probably be in overcapacity, others will remain in undercapacity. ■ The commodity electricity markets operate on the principles of marginal supply and marginal cost. Asset displacement and redundancy in the commodity markets are a function of asset variable costs relative to demand levels. ■ The wholesale electricity markets are roughly split evenly between contract and commodity markets (although this varies among regional markets). ■ The contract markets are relatively immune from market volatility. This is a significant risk in the commodity markets. ■ Generation market risk for assets serving commodity markets (i.e., merchant power producers) is mitigated largely by favorable economics and/or operating flexibility. Local electricity transmission constraints also mitigate risk.

ELECTRICITY IS A UNIQUE COMMODITY Market risks in electricity generation incorporate features that are unique to this industry and that require more specialized analysis approaches and

Market Development

155

tools. These features are attributable to the fact that electricity is a commodity—but with twists: ■ It cannot be stored in inventory (i.e., production and consumption must always be in balance). ■ Supply must be assured (i.e., the lights cannot go out). ■ Demand is highly inelastic in the short run (there are no practical substitutes). These features have significant implications for market risk. For example, the need to guarantee adequate supply at all times means that “overcapacity” is defined differently from that in other commoditized markets. In electricity markets, overcapacity is relative to maximum (peaking) demand, not average demand. In addition, the fact that electricity cannot be stored restricts both buyers and sellers to producing and consuming the marginal megawatt of electricity. There is therefore a consistent relationship between marginal economics and supply/demand balance that is also unique to this industry.

MARKET DEVELOPMENT The creation of the competitive U.S. electricity market is rooted in the passage of the Energy Policy Act of 1992. This Act directed the U.S. Federal Energy Regulatory Commission (FERC) to establish national nondiscriminatory open access to the transportation of electricity. This enabled buyers and sellers of wholesale electricity to directly contract with each other without purchase and resale through an intermediary utility. “Open access” to electricity transmission removed the key structural barrier to electricity wholesale trade and launched a commodity wholesale electricity market. The impact of the Act on the development of a commodity wholesale market is illustrated in Figure 10.1, which shows the growth of commodity “interruptible” power deliveries versus contracted “firm” deliveries from 1990 through 1996. The passage of the 1992 Act has been followed by the gradual liberalization of the retail electricity markets (i.e., the sale of commodity electricity directly to industrial, commercial, and residential users) on a state-by-state basis. As of 2001, more than half of the states in the United States had fully

156

MARKET RISK IN ELECTRIC GENERATION FINANCE

Delivery Volume (MWh)

1,000,000 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0

Firm Interruptible

1990

1996

Growth 1990 –1996

Interruptible + Firm = Total Electricity Delivery

FIGURE 10.1 Growth of Commodity Transactions: Electricity Deliveries, 1990 and 1996 (’000MWh) Source: U.S. Department of Energy

functioning deregulated retail electricity markets or had adopted enabling legislation to put legislation in practice (see Figure 10.2). On the retail level, the successful state models have created competitive electricity markets primarily for industrial and commercial consumers. The residential markets have proven to be much more difficult to open. Industrial and commercial retail competition is likely to accelerate with the expansion of the Regional Transmission Organizations (RTOs), which will take over the incumbent utility control over regional transmission. The deregulation process in the United States has facilitated the growth of competitive wholesale electricity markets through ■ The growth of “trading hubs” throughout the U.S. markets, which serve as central markets to provide liquidity and price discovery. Some of these markets have developed into formal exchanges, such as the New England and New York power exchanges. The ongoing establishment of RTOs will complement the growth of the exchanges by providing commodity transportation and logistical services. ■ The emergence of “merchant” power producers who develop and operate generation facilities specifically for sales into the wholesale and

Market Development

157

1 Arizona, Arkansas, California, Connecticut, Delaware, District of Columbia, Illinois, Maine, Maryland, Massachusetts, Michigan, Montana, Nevada, New Hampshire, New Jersey, New Mexico, Ohio, Oklahoma, Oregon, Pennsylvania, Rhode Island, Texas, Virginia, and West Virginia. 2 New York. 3 None. 4 Alaska, Colorado, Florida, Indiana, Iowa, Kentucky, Louisiana, Minnesota, Mississippi, Missouri, North Carolina, North Dakota, South Carolina, Utah, Vermont, Washington, Wisconsin, and Wyoming. 5 Alabama, Georgia, Hawaii, Idaho, Kansas, Nebraska, South Dakota, and Tennessee.

FIGURE 10.2 Status of State Electric Industry Restructuring Activity as of September 2001 Source: U.S. DOE, Energy Information Administration

liberalized retail markets but have no retail obligation to serve electric load. “Merchant” power producers are generally exempted from federal and state regulation. ■ The emergence of “power marketers” who deal in the purchase and resale of wholesale electric power and competitive retail power. Power marketers assumed the key role of “market makers” in the wholesale markets for both spot and forward transactions.

158

MARKET RISK IN ELECTRIC GENERATION FINANCE

The evolution of the U.S. wholesale markets since the U.S. Energy Policy Act of 1992 has resulted in four basic categories of players in the wholesale electricity markets: ■ Energy service providers: Energy service providers produce and deliver energy to consumers. Energy service providers are integrated into both electricity and natural gas and have interests in or exposure to retail energy distribution. Examples include utility companies such Exelon, Mirant, and NEG. ■ Merchant power producers: Merchant power producers develop and operate generation facilities for electricity sales into the commodity electricity markets essentially acting as electricity traders. Examples include Sithe Energy and Calpine. ■ Independent power producers: Independent power producers develop and operate generation facilities for sales under long-term power purchase contracts to utilities or other long-term off-takers. Examples include Tenaska and AES. ■ Power marketers: Power marketers purchase and sell electricity in both the spot and forward markets. They also will sell power into the industrial and commercial consumer markets. Examples include Enron and Aquila. Electricity deregulation prompted the traditional regulated utilities to fundamentally restructure their business. Several formed unregulated independent subsidiaries to house all generation and compete exclusively in the wholesale market (or they have spun these subsidiaries off). Others shed generation assets outright and evolved into pure retail distribution “pipes and wires” utilities. Of the utilities that did not create independent unregulated “generation companies,” half (30 out of 59) have sold all of their generation assets. This is equal to about 22 percent of total U.S. generation capacity. Preservation of generation assets within the regulated utility structure now resides only with the public and cooperative utilities. The electricity markets themselves have evolved into relatively distinct regional markets distinguished by unique electricity demand and supply profiles (14 total for the continental United States). These markets roughly correspond with the same reliability regions established by the North American Electric Reliability Council (NERC) as a result of the transmission constraints, which delineate intramarket transactions versus import/export transactions. The 14 regional electricity markets are described in Figure 10.3.

159

Composition and Growth of Electricity Demand

LEGEND Market

Symbol

Map

Market

Symbol

Map

New England

NEPOOL

7

East Central Area Reliability Region

ECAR

1

New York

NYPP

6

Mid-America Interconnected Network

MAIN

4

Mid-Atlantic Area Council

MAAC

3

Mid-Continent Area Power Pool

MAPP

5

Southeast Regional Council

SERC

9

Northwest Power Pool

NWPP

11

Florida Regional Coordinating Council

FRCC

8

Rocky Mountains

WSCC-RM

12

Southwest Power Pool

SPP

10

Arizona/New Mexico/Nevada

WSCC-W

11+12

Texas

ERCOT

2

California

WSCC-CA

13

FIGURE 10.3 Regional U.S. Wholesale Electricity Source: U.S. DOE, Energy Information Administration

It should be noted that in the more advanced regional markets, power exchanges and the related RTOs are becoming synonymous with the regional definitions. For example, the Pennsylvania/New Jersey/Maryland power exchange (PJM) is synonymous with MAAC.

COMPOSITION AND GROWTH OF ELECTRICITY DEMAND There are three key observations for electricity demand relative to financial risk analysis: (1) The separation of electricity demand into stable contract and volatile commodity components means different risk criteria are required for assets serving the different markets; (2) electricity demand growth is being increasingly driven by service sector uses. This means that the pace of structural change to a service economy, rather than levels of absolute economic growth, are a key driver of electricity demand; (3) Growth in electricity demand varies among the electricity markets. So, therefore, does the degree of financial risk. Electricity demand can be divided into a relatively constant component, which represents the level of demand required to meet minimal economic

160

MARKET RISK IN ELECTRIC GENERATION FINANCE

needs and a variable component that will be driven by random events, such as weather or unexpected energy needs. The constant component of electricity demand is approximated by average annual demand. Variable demand is modeled by “peaking” demand, which is the highest level of demand in a given hour. Because of the influence of random factors on electricity demand, it can be highly volatile and subject to surprise spikes. Figure 10.4 shows the changes in electricity demand for the New England Power Pool market in Hour 16 (16:00–17:00 hours) during the month of June 2001. Because electricity demand is relatively inelastic and electricity cannot be stored, supply must change in direct relationship with demand. For demand and supply to remain in equilibrium (which, by definition, must occur), the electricity “spot” markets supply the marginal megawatt of power to meet variable demand levels. In turn, the price of the marginal megawatt of power is set by the economics of the generation source delivering this power (since demand is relatively inelastic). On the other hand, the “contract” electricity markets serve stable demand and are not directly vulnerable to wholesale market risk. While electricity demand is volatile, annualized trends show a rela6000

5000

Megawatts

4000

3000

Hour 16 Demand

2000

1000

0

1-

n Ju

3-

n Ju

5-

n Ju

7-

n Ju

9-

n Ju

11

un -J

13

un -J

15

un Jun -Jun un -Jun -Jun -Jun -Jun -J -J 19 23 25 27 29 17 21

Day

FIGURE 10.4 NEPOOL Electricity Demand for Hour 16 (June 2001) Source: New England Independent System Operator

161

Composition and Growth of Electricity Demand

tively consistent pattern of growth through economic cycles. Historical growth in U.S. demand from 1989–1999 is shown in Figure 10.5. Demand growth for 1989–1999 was on average 2.6 percent. This growth tracked economic activity, particularly the recessionary periods of 1987–1988 and 1990–1991, but the linkage between economic growth and electricity demand is not one-for-one. This suggests that growth in electricity demand is only partially linked to economic activity. The decoupling of electricity demand growth to overall economic activity is attributable to the growing influence of service sector electricity demand. Commercial electricity usage has been growing at a rapid pace of 3.2 percent—which is more than twice the growth of industrial usage (see Figure 10.6)—and accounts for an increasing proportion of electricity demand. Commercial electricity usage is characterized by heating and cooling applications, which are tied more to weather than to production processes. The continued expansion of the service sector means that total demand growth will be in between the 1.3 percent growth of the industrial sector and the 3.2 percent growth of the commercial sector. Projected demand growth of 1.7 percent for the U.S. electricity markets 700000

Megawatts

650000

600000

550000

500000

450000 1996 1987 1988 1989 1990 1991 1992

1993 1994 1995 1996

Year

FIGURE 10.5 U.S. Electricity Demand, 1986–2000 Source: U.S. Department of Energy

1997 1998 1999 2000

162

MARKET RISK IN ELECTRIC GENERATION FINANCE

302.5

286.1

Growth Residential

2.5%

Commercial

3.2%

Industrial

1.3%

142.3 FIGURE 10.6 Electricity Demand Growth by Consumer, 1989–2000 (MillionsWh) Source: U.S. Department of Energy

is benchmarked by industrial demand growth (1.3%) and historical growth (2.6%). Based on these two conservative qualifiers, we believe that 1.7 percent is a prudent aggregate indicator of future demand in the forecast period of 2001–2005. Projected demand growth for the individual electricity markets (adopted from Cambridge Energy Research Associates data) is provided in Table 10.1. Both projected and historical demand growth clearly varies among the regional markets. The MAAC market in the U.S. Northeast is projected to grow at 2.5 percent over the 2000–2005 period while the NWPP market shows zero growth. The differences in regional market growth rates illustrate that analysis of demand on a regional basis is much more relevant than single “economic activity-based” measures.

COMPOSITION AND GROWTH OF NEW CAPACITY The risk implications of new capacity for electricity financing relate specifically to the risk of overcapacity as industry participants rush in to seize opportunities in the deregulated markets. While overcapacity concerns are certainly legitimate, market-specific analysis of electric generation shows that overcapacity pertains only to certain assets and markets. It is, by no means, a universal risk. Moreover, overcapacity risk is capped by limitations in the U.S. energy infrastructure (specifically natural gas infrastructure), which limits total

163

Composition and Growth of New Capacity

TABLE 10.1 Projected and Historical Demand Growth by Market Market New England New York Mid-Atlantic Area Council Southeast Regional Council Florida Regional Coordinating Council Southwest Power Pool Texas East Central Area Reliability Region Mid-America Interconnected Network Mid-Continent Area Power Pool Northwest Power Pool Rocky Mountains Arizona/New Mexico/Nevada California Total Regions:

Symbol NEPOOL NYPP MAAC SERC FRCC SPP ERCOT ECAR MAIN MAPP NWPP WSCC-RM WSCC-W WSCC-CA

CAGR CAGR 2001–2005 1986–2000 2.1% 2.1% 2.5% 1.3% 2.3% 1.5% 2.0% 1.8% 1.2% 1.7% 0.0% 1.7% 2.4% 1.9% 1.7%

2.3% 2.3% 2.2% 2.6% 2.2% –1.3% 2.4% 2.4% 2.6% 3.2% N/A N/A N/A 2.6% 2.6%

CAGR: Compound Average Growth Rate Source: Cambridge Energy Research Associates, U.S. Department of Energy

development. Development is highly restricted by gas infrastructure issues. These limitations mean that actual construction of a large part of announced capacity is highly unlikely, if not impossible. Through the decade of the 1990s, there were minimal additions to generating capacity, despite consistent growth in demand. Uncertainty related to the deregulation process coupled with the confusing state-by-state process of deregulation discouraged investment in new capacity. Throughout the decade, industry participants and investors were convinced that the implementation of competitive markets would result in generation overcapacity (the so-called “stranded cost” issue). Competitive market prices would not justify new construction. As a result, there was insufficient investment and the U.S. electricity supply/demand balance actually deteriorated to dangerous levels. Total capacity utilization increased from 74.4 percent in 1989 to 91.8 percent in 1999 leaving very little generation capability to meet certain future growth. The pace of investment picked up rapidly in 2000 when developers realized overcapacity concerns were contradicted by industry fundamentals. Generation investment has accelerated rapidly (Figure 10.7).

164

MARKET RISK IN ELECTRIC GENERATION FINANCE

840000 Capacity (MW)

820000 800000 780000 760000 740000 720000 700000 680000 1996

1997

1998

1999 Year

2000

2001

2002

FIGURE 10.7 Total U.S. Generation Capacity Source: U.S. Department of Energy

Approximately 46,000 megawatts (MW) of new capacity was scheduled to come online in 2001 with an additional 42,000 in construction, which should be operational by the end of 2003. This increases total U.S. generating capacity by about 12 percent over total 2000 capacity (which was about 756,000 MW). These increases do not include approximately 226,000 MW of new capacity, which have been announced for completion by 2005 but are not under construction. The potential that all this new announced capacity would be built has again raised the issue of overcapacity in U.S. electricity markets. If all new capacity announcements were built, total capacity would increase by 314,000 MW, or 41.5 percent. However, raw capacity projections present a very misleading view of future capacity growth. Of the total 314,000 MW in potential new capacity (of which 172,000 MW is announced but not in development) development, 94 percent are fueled by natural gas. If all potential gas-fired generation capacity is built as announced, total incremental natural gas demand will increase 10 to 12 trillion cubic feet by 2005. This is almost 50 percent of total U.S. natural gas consumption in 2000 and corresponds to a 36.2 trillion cubic feet natural gas market by 2005. Figure 10.8 illustrates the impact of this capacity build on natural gas dynamics. Relative to both historical and projected natural gas consumption patterns, increases in delivered natural gas of this magnitude would place huge strains on natural gas production and infrastructure. Total his-

Volume (in trillion cubic feet)

Composition and Growth of New Capacity

165

40 35 30 25

Total Consumption

20

Total Production

15 10 5 0 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001*2002* 2003* 2004*2005*

Year

FIGURE 10.8 Gas Consumption versus Production Source: U.S. Department of Energy; DZ Bank Estimates

torical gas production has averaged about 1.0 percent increases compared to 1.7 percent increases in total demand. This has been the pace at which the gas delivery infrastructure has been able to absorb and respond to new consumption. If all announced new capacity was built, this would require annual delivered gas supply growth of 9.8 percent. Natural gas supply and delivery is simply not adequate to meet these requirements. Despite limitations of the natural gas infrastructure on capacity build, increased generation financing risk through overcapacity could still emerge if developers continue to construct assets that would lie idle or would have limited operation because of fuel issues. However, development of gas-fired generation is sufficiently flexible to avoid investment overshooting. Gasfired generation investment is by nature a multiyear, multistep process. In fact, gas-fired generation development is often modeled as a bundle of sequential options for capital budgeting purposes. This inherent flexibility allows developers to scale back or accelerate investment relative to changing market conditions. Developers are very familiar with this flexibility, which is why they can feel comfortable in making highly optimistic growth announcements for consumption by Wall Street analysts and fund managers. Of the 314,000 MW of potential new capacity, developers can (and will) scale back on the 172,000 MW that is planned—but not in development or construction. In view of gas infrastructure constraints, a more realistic total for new capacity in the 2001 to 2005 period is approximately 141,000 MW, of which 131,000 will be gas-fired. The capacity

166

MARKET RISK IN ELECTRIC GENERATION FINANCE

estimate of 141,000 MW is based on third party consultants’ assessment of projects, which developers will allow to go forward. This capacity is considerably more in line with gas infrastructure considerations (as shown in Table 10.2). A much more reasonable (while still optimistic) 4.7 percent growth rate is implied by 141,000 MW, compared to 1.7 percent historical average growth and a 3.0 percent production growth rate.

SUPPLY AND DEMAND EQUILIBRIUM A careful analysis of the supply and demand balances across the United States reveals that financial risk is relative to the market in which assets are developed or planned. Some markets will clearly be in overbuilt situations and the financial risks stemming from overcapacity will be high. Other markets are very likely to have sustained undercapacity. The financing risk there will be correspondingly low. Supply and demand equilibrium in wholesale electricity markets is measured by reserve margins. Reserve margins indicate the relative degree to which a wholesale electricity market can be considered to be in undercapacity, overcapacity, or approximate balance. Reserve margins are defined as the excess generation capacity in any given market over peaking demand (expressed as a percentage of generation capacity). Reserve margins express the degree to which electricity supplies will be strained to meet potential future demand. They are therefore the key measure of the need (or lack of need) for new capacity. In general: ■ Reserve margins less than 15% indicate a relatively tight market. Since peaking demand is itself volatile, there is insufficient supply to assure that power requirements will be met. TABLE 10.2 New Generation Capacity Impact on Natural Gas Demand Demand and Production*

2000

2005

CAGR

Announced capacity Likely capacity Production

22.7 22.7 19.22

36.2 28.5 22.8

9.80% 4.70% 3.00%

*In trillion cubic feet Source: U.S. Department of Energy, DZ Bank Estimates

Supply and Demand Equilibrium

167

■ Reserve margins of 15 percent to 20 percent indicate a balanced market. Electricity market regulators customarily required reserve margins of 20 percent to assure market reliability and would encourage the necessary investment to reach this benchmark. ■ Reserve margins over 20 percent indicate potential overcapacity. While reserve margins over 20 percent can quickly evaporate if existing capacity exits in the market or is unexpectedly shut down; excess capacity at this level encourages asset displacement, particularly of the less competitive assets. Combining the demand and new capacity analyses discussed in the previous sections provides an assessment of the future undercapacity, overcapacity, or balance for the regional wholesale electricity markets. This assessment is provided in Table 10.3 which shows the reserve margins for regional U.S. electricity markets for the 2000–2005 period. The estimates in Table 10.3 are based on the previously discussed demand and capacity growth forecasts in the previous sections. By 2005, a total of five regional U.S. electricity markets will be in potential overcapacity. Six will still be in potential undercapacity. Three markets will be in approximate balance. Markets, which have a sustained shortage of generation capability, include the economically critical Northeast, Southeast, and California markets that together are 41.7 percent of total U.S. electricity demand. In addition, the Midcontinent Area Power Pool (MAPP) could potentially face a power crisis because of the very limited growth in generation assets in the region. The markets in overcapacity include the New England Power Pool (NEPOOL), the Northwest Power Pool (NWPP) and the markets of Texas and the Southwest Power Pool (SPP). The NEPOOL, Texas, and SPP markets are likely to experience displacement of existing assets and a competitive struggle among low cost assets for available demand. The NWPP situation is, however, a little less clearcut. The NWPP is characterized by heavy dependence on hydroelectric power. Unfavorable weather conditions (specifically rainfall patterns) could significantly reduce availability of hydro assets and cause a capacity surplus to turn quickly into a capacity deficit. The supply and demand equilibrium across the regional markets shows that there will indeed be pockets of overcapacity. However, there are simultaneously significant pockets of undercapacity. The overall fundamentals

168

MARKET RISK IN ELECTRIC GENERATION FINANCE

TABLE 10.3 Supply/Demand Balances by Regional Market 2000 and 2005 (Reserve Margins) Market New England New York Mid-Atlantic Area Council Southeast Regional Council Florida Regional Coordinating Council Southwest Power Pool Texas East Central Area Reliability Region Mid-America Interconnected Network Mid-Continent Area Power Pool Northwest Power Pool Rocky Mountains Arizona/New Mexico/Nevada California

Symbol

2000

2005

Assessment

NEPOOL NYPP MAAC SERC

3.6% 12.5% 15.5% 3.1%

27.2% 10.6% 13.6% 12.8%

Overcapacity Undercapacity Undercapacity Undercapacity

FRCC SPP ERCOT

5.0% 16.7% 9.6%

11.5% 25.5% 22.0%

Undercapacity Overcapacity Overcapacity

ECAR

12.3%

16.2%

Balance

MAIN MAPP NWPP WSCC-RM WSCC-W WSCC-CA

11.8% 0.4% 32.7% 9.5% 0.2% 4.9%

20.7% 0.0% 41.2% 19.1% 17.4% 12.4%

Overcapacity Undercapacity Overcapacity Balance Balance Undercapacity

Source: Cambridge Energy Research Associates, Financial Times Energy (RDI), Williams Capital Group

for the U.S. electricity markets strongly indicate that there will be a compelling need for new generation capacity in the order of at least 140,000 MW over the period 2001–2005. The alternative is periodic electricity blackouts, which are not economically or politically acceptable. While it is highly unlikely that the 172,000 MW announced will be constructed, it is evident that substantial investment in generation will be sustained through the immediate future.

GENERATION ECONOMICS AND ASSET DISPLACEMENT Analysis of generation economics and asset displacement shows that financial risk considerations for generation assets exposed in the wholesale markets are (1) the variable production cost of the asset relative to other assets of its class; (2) the degree of demand volatility in the market and; (3) the

Generation Economics and Asset Displacement

169

operational flexibility of the asset relative to the segment of the market it is designed to serve. Asset displacement in the electricity markets is different from displacement (i.e., asset redundancy) in other industries. Because of the need to ensure adequate capacity, inefficient or uneconomic generation assets may not necessarily be completely retired. They can be idled with the idea of being reintroduced into service if the need arises. The degree to which an inefficient asset will be idled (instead of dismantled) depends on the flexibility of the asset. This, in turn, is a function of the start up/shut down costs. As an illustration, an uneconomic nuclear or large coal generation asset with high start up/shut down costs and slow reaction time will be dismantled. An oil steam generation asset, because of its higher flexibility, may only be idled. The measure of asset displacement in the commodity electricity markets is capacity utilization (known as the “capacity factor”), which is defined as the ratio of total annual generation to total available capacity. “Displacement” in electricity markets means that certain assets are diverting demand from other assets. This diversion is evidenced by changes in the capacity factors. It should be noted that using the capacity factor as a proxy for financial operational risk is valid only for assets with similar operational flexibility. Generation assets specifically designed to meet peaking demand, such as combustion turbines (CTs) will naturally have low utilization rates, but, because of their low set up/shut down costs, do not require frequent utilization to be financially viable. The economic process of asset displacement is based on the relative contribution margins (and associated variable costs) of competing assets. To see why, a simplified example of a commodity electricity market (Figure 10.9) is helpful: Figure 10.9 illustrates a simple electricity market with one consumer and two generators. The consumer has a basic demand of 100 MW in a given hour. This can randomly increase to 200 MW. The two generators consist of a new gas-fired combined cycle facility and an older gas-fired steam facility. The new gas-fired combined cycle facility has an available capacity of 100 MW and variable costs of $20.00 per MWh and capital costs of $20.00 per MWh. The steam facility also has available capacity of 100 MW and variable costs of $40.00 per MWh. On a total cost basis, the cost structures of the two facilities are therefore the same ($40.00 per MWh).

170

MARKET RISK IN ELECTRIC GENERATION FINANCE

Demand $40.00/MW

Market Price Steam

$20.00/MW Combined Cycle

Volume (MW) FIGURE 10.9 Commodity Market Generation Economics At the consumer’s demand level of 100 MW, either of the generators could physically bid for the volume. However, the older steam generator could accept only a minimum price of $40.00. Any less and they would produce at a negative contribution margin which would be financially unsound. The combined cycle facility, on the other hand, could accept a price greater than $20.00 and contribute to its capital costs. If the combined cycle facility were to bid at $25.00 (compared to the steam generator’s $40.00 minimum bid), the consumer would clearly purchase from the combined cycle facility over the steam generator (the steam generator’s capacity is effectively “displaced”). The potential market price is therefore that of the least efficient producer (i.e., $40.00). At the consumer’s demand level of 200 MW, each of the generator’s capacity is needed to meet the volume (there is no ability to draw supply from inventory). Both generators are needed. While the efficient generator could continue to bid at $25.00, the less efficient steam generator will bid over $40.00. The efficient generator is powerfully incented to bid at $40.00 since the consumer will have little choice but to pay $40.00 for the efficient producer’s capacity. In principle, the market price in this scenario could be infinite; the market price will be the bid submitted by the steam generator. In reality, the next highest cost generator competing for the incremental 100 MW will bound the market price. The simplified market just discussed mirrors the actual performance in

171

Generation Economics and Asset Displacement

Average Production Cost ($/MWh)

commodity electricity markets—specifically that the least efficient (most expensive) producer sets the price. This result obtains because electricity cannot be stored (otherwise the consumer’s 200 MW in demand could be met from 100 MW in the combined cycle generator’s inventory at $25.00). Figure 10.10 and 10.11 show the relationship between utilization (capacity factors), marginal production cost, and set up/shut down costs (using capital cost as a proxy for set up/shut down cost) for different generation types in 1999. There is a clear direct relationship between utilization and set up/shut down cost and a clear inverse relationship between utilization and variable cost. This is exactly what would be expected in a market driven by variable cost. 35.00 30.00 25.00 20.00 15.00 10.00 5.00 0.00

Gas

Petroleum

Coal Nuclear

0.0

20.0

40.0

60.0

80.0

100.0

Average Utilization (%)

FIGURE 10.10 Average Generation Economics by Fuel Type (1999 Data)

Average Shut-Down Cost ($/MWh)

Source: PA Consulting, U.S. Department of Energy, DZ Estimates

35000 30000 25000 20000 15000 10000 5000 0

Nuclear Coal Gas Petroleum

0.0

20.0

40.0

60.0

80.0

Average Utilization (%)

FIGURE 10.11 Average Generation Economics by Fuel Type (1999 Data) Source: PA Consulting, U.S. Department of Energy, DZ Estimates

100.0

172

MARKET RISK IN ELECTRIC GENERATION FINANCE

TRANSMISSION CONSTRAINTS AND ASSET DISPLACEMENT The asset competitiveness model outlined in the previous section suggests that the two principle variables for asset competitiveness are (1) operating cost structure and (2) operating flexibility relative to other assets in the same class. This model holds when electricity can freely flow from where it is produced to where it is needed. Using the previous illustrative example, the model works when either the combined cycle generator or the older gas-fired steam generator can serve the consumer demand. It breaks down when transmission bottlenecks prevent electricity production based on economic considerations. Transmission bottlenecks are endemic to wholesale electricity markets. Bottlenecks are caused by the physical rules governing the transport of electric power. In reality, electric power consists of two types: real power and reactive power. Real power is the electricity consumed by electric devices. Reactive power is the electricity level that must flow through an electricity delivery system to keep the system functioning. Imbalances in either of these electricity types can cause system failures (i.e., blackouts or brownouts). Moreover, while the wholesale electricity trade assumes delivery of power directly from one point to another, the actual physical delivery of that power follows a much more convoluted and circuitous route—triggering real and reactive electric power constraints everywhere along the route. Transmission congestion is both frequent and economically significant As shown in Table 10.4, the major wholesale markets in California, MAAC, and New York experienced some degree of congestion ranging from 14.1 percent of total hours in California to 49.0 percent of total hours in New York. The impact of congestion on localized electricity prices is equally noteworthy. Congestion added $775.00 and $915.00 to the price of localized power during the most congested hour in California and New York, respectively. The total cost of congestion can also be sizable. Transmission constraints added $20 million to the NEPOOL power costs in the single month of December 1999. The increases in power costs during congested hours are a result of deviations from the electricity market model. In these conditions, generation assets will run, not because they are the most attractive alternatives on the margin, but because they can provide power where other assets cannot

173

Market Structure and Risk

TABLE 10.4 Frequency and Cost of Congestion (Year 2000) Market California MAAC New York

Frequency (MWH)

Cost ($/MWH)

14.1% 27.7% 49.0%

$775 $136 $915

Source: Beacon Energy

reach. This causes breakdown in the relationship between displacement and generation economics. Noncompetitive units will be able to sustain themselves if their markets are transmission-constrained. Put in a different way, transmission constraints set up tangible barriers to entry. Specific location on the transmission system is therefore an important footnote to generation financial risk.

MARKET STRUCTURE AND RISK The market structure for electricity generation and delivery demonstrates that financial risk is a function of the segment of the market in which generation assets operate. The relative size of the total wholesale markets compared to regulated utility generation show that independent (or unregulated) electricity generation is a fundamentally competitive activity. Within the wholesale market, the size of the commodity segment of the wholesale market varies among regional markets, but is certainly significant (42.7% of the wholesale market). This means that generation assets, which operate in the commodity segment, can be competitive if their operating costs are comparatively low. Because of the volatile nature of electricity demand, the wholesale electricity markets consist of both a bilateral contract market and a commodity market. The contract market consists of long-term power purchase agreements (usually 10 to 15 years in length) which provide firm power delivery commitments at fixed or semi-fixed prices. The commodity markets consist of a spot (i.e., for immediate delivery) and a forward (i.e., for future delivery) market. The spot market is the residual electricity market that brings supply and demand in balance. The forward market is the risk management

174

MARKET RISK IN ELECTRIC GENERATION FINANCE

market that enables participants to trade in risk. The breakdown of the wholesale electricity market is described in Table 10.5. Table 10.5 shows that approximately 42.9 percent of wholesale market volume is attributable to commodity transactions. Long term power purchase agreements consists of 57.1%. Although no national estimates of forward market size exist, transaction volumes in the developed Northeast markets indicate that approximately 25% of total wholesale transactions are spot transactions. This implies that the forward market is approximately 17% of the total market (this proportion is also similar to the breakdown between spot and forward wholesales transactions in the natural gas markets). The commodity markets are a large part of total wholesale electricity transactions, which, given the significant volatility of these markets, makes economic sense. Size of the commodity market does, however, vary between the regional markets. The commodity market in Florida is only 2.6 percent of the wholesale market, while the commodity market in the Southwest Power Pool is about 60.4 percent.

TABLE 10.5 Wholesale Electricity Markets by Region, 1999 (MWh)

ECAR Texas FRCC MAAC MAIN MAPP New Eng. SERC SPP California Other WSCC** New York Total:

Total Net Generation

Total Wholesale Market

Percent Merchant*

Percent of Generation

542,209,400 232,725,818 165,058,236 219,600,956 237,970,592 146,185,966 45,486,669 748,523,214 179,781,704 172,030,549 410,417,752 92,893,261 3,192,884,117

320,009,650 856,427,501 58,375,183 383,476,334 111,169,028 139,218,411 275,932,303 836,908,403 430,142,014 224,240,871 457,557,800 134,667,074 4,228,124,572

22.5% 70.9% 2.6% 45.3% 33.1% 18.1% 41.9% 32.7% 60.4% 13.1% 43.4% 14.5% 42.9%

59.02% 368.00% 35.37% 174.62% 46.72% 95.23% 606.62% 111.81% 239.26% 130.35% 111.49% 144.97% 132.42%

*Includes spot and forward markets **Other WSCC includes the Northwest Power Pool (“NWP”), the Rocky Mountain Area and Arizona/New Mexico/Nevada Source: U.S. Department of Energy

175

Market Structure and Risk

It is also important to note the size of the wholesale markets relative to incumbent utility production. On average, utilities are 1.3 times more on the wholesale markets than they generate. This implies that the wholesale electricity markets are very necessary means for utilities to serve the ultimate customers and that regulated utilities are constantly in the wholesale market either to buy or to sell. Because both the spot and forward markets serve the volatile portion of electricity demand, they respond to the price signals determined by the marginal production costs discussed in the previous section. The number of formal exchanges is relatively limited, consisting of the California Power Exchange, the New York ISO, the New England ISO and the PJM ISO. However, these exchanges provide transparency for the commodity markets, in general. The primary trading forum in these exchanges is the “day ahead” market, in which the exchange managers will schedule supply for estimated next day demand based on the total price/supply offers given by producers to meet demand. Price behavior in the “day ahead” market should follow

$600

$500

July 25 (HE 16)

Bid Price ($ per MWh)

August 9 (HE15) $400

July 25 (HE 16) Supply available below $75/Mwh : 22,020 MW Total available supply : 25,222 MW

$300

August 9 (HE 15) Supply available below $75/Mwh : 21,868 MW Total available supply : 25,417 MW

$200

$100

$0

5,000

10,000

$(100)

15,000

20,000

25,000

30,000

Capacity (MW)

FIGURE 10.12 July 25, 2001 Peak versus August 9, 2001 Peak Internal Resources Supply Curve (Capped at $500/MWh) Source: New England System Operator

176

MARKET RISK IN ELECTRIC GENERATION FINANCE

marginal cost structures. An illustration of the relationship between marginal cost and market price is shown in Figure 10.12, which shows market prices relative to market volume in the NEPOOL market during 2001. As can be seen in the analysis, market prices increased directly with required blocks of production and increased sharply at production levels needed to meet peaking demand. This is exactly the behavior, which would be anticipated following the illustrative example outlined in the previous section.

CHAPTER

11

Convergent Systems P. Kumar President

Shiva Gowrinathan Vice President Client Services Nirvanasoft Inc.

INTRODUCTION The convergence of energy and related commodities and markets has created new challenges for the back offices of the participants. Complex contracts and trading products, both financial and physical, are being developed, based on mathematical models and algorithms. And, as in any other business, it is the job of the software systems to perform the operational activities from deal capture to settlement on the one hand, and to provide analytic results to determine the profitability of these contracts on the other. The traditional model has followed a compartmentalized architecture: First of all, software systems that can handle multiple commodities (other than very simple contracts) are few and far between. Even for a particular commodity, the different functional components (such as generation and demand forecasts vs. price forecasts and settlements) are separated into their own worlds, with expensive and often unreliable interfaces. To compound the complexity there is also a need for the integration between the physical and financial transactions, too. The emergence of retail energy deregulation has significantly increased the criticality of such integrated solutions by introducing a lot more moving pieces into the equation, including noncommodity products as an integral part of a service contract. The absence of integration and reliable interfaces has introduced a couple of new buzzwords into the energy

177

178

CONVERGENT SYSTEMS

marketer’s vocabulary—“revenue leakage” and “inadvertent hedging”—as well as the problem of individual departments working with data that are not synchronized with each other, resulting in lost revenue, unanticipated expenses, and incomplete management of volumetric and financial risk. This paper is an attempt to outline the basics of a software system that cuts across multiple commodities and functions that need to be performed in a consistent, integrated, and synchronized fashion. This is by no means an attempt to define or create a standard or list of requirements; the fast-moving energy industry does not have the clear definition necessary to specify standards. In the emerging markets driven by regulatory and international market forces, the only certainty is that whatever you build today has to be able to be adapted tomorrow. In other words, the only specification is God Only Knows (GOK). We present here a high-level view of some of the ideas that we are implementing at NirvanaSoft Inc. in our attempt to build the GOK solution.

OVERVIEW We start with a review of the current world in terms of the organizational entities involved in the energy service companies’ (ESCO’s) business process. The participants in the deregulated energy market consist of ■ Energy service companies, who primarily deal with retail energy contracts. ■ Wholesalers and traders whose focus is mainly on the financial contracts. ■ Independent power producers who also have to participate in this market, and usually their contracts are a mixture of physical and financial contracts. These organizational pieces work mostly with stove-piped systems, in the sense that they are different systems each of which supports a business function, and are usually limited to transactions that involve one commodity. Next, we look at the brave new world where transactions involve multiple commodities and, very often, noncommodity products. Our definition of a convergent system has several components. ■ It must be able to accommodate transactions that involve multiple commodities and types of products. It should certainly have the flexibility to accommodate many complex contract and product structures.

The Current World

179

Preferably, the job of setting up new contracts and products should not involve custom programming. It should be a matter of filling in forms on the screen. ■ The underlying data structures should not distinguish among the different commodities; otherwise, any analysis in terms of forecasting, pricing, costing, and risk management cannot compare apples to apples. Under this requirement, using stove-piped systems to do individual contracts and electronically stapling the results does not qualify as a convergent system. ■ The system must be able to accommodate data from varied sources in multiple formats (such as flat files, EDI, FIX, XML, etc.) following multiple transmission protocols (such as SMTP, HTTP, FTP, file posting, etc.). ■ The system must perform “straight-through processing.” This means that when a transaction is posted into the system, it should simultaneously update all the structures that are affected by it. It should have the intelligence to route itself through the many steps of the business process following a pre-written script.

THE CURRENT WORLD In the current world, both in utilities and marketing organizations, each commodity or service group operates as an island. In the utilities, the functional components were vertically integrated. Until deregulation, there was little perceived need for crossbreeding across the commodities or functional areas. The result has been a hodgepodge of software from many vendors, plus some homegrown software; the interfaces are custom made for each environment. In the organizations, one typically encounters such islands supported by their own software: ■ The structuring desk, which creates new products and pricing schemes. The tool of choice in this world is spreadsheets and homegrown modeling systems. They are typically linked to databases with real-time market statistics or in the case of retail markets, historical information from a variety of sources including the Independent System Operator (ISO) and the utilities. ■ The originators (in the wholesale world) or sales and marketing professionals (in the retail world), deal with the physical transactions. A

180

■

■

■

■

CONVERGENT SYSTEMS

variety of sales force automation and contract administration software rule this world. The traders dealing exclusively with financial transactions. In the energy market, there are a few electronic exchanges, along with proprietary software that run OTC trades between trading parties. The operations staff, responsible for fulfillment of the physical contracts in terms of delivery. Their tools are volumetric forecasting systems. These provide schedules, which are then posted to the appropriate entity, such as the ISO. The accountants, responsible for billing, settlements, and collections. Many billing systems are in use. However, due to the nature of these complex contracts, many of them are used as glorified word processors; the calculations are done offline, and the final piece is boiled down to a set of data that the billing system can print the bill from and manage the accounts receivable activity. The reconciliation of settlements for physical transactions with the utilities and the ISO are primarily manual processes, if done at all, which, in our experience, is not very often. The finance staff, responsible for risk management, profitability analysis, and revenue and earnings forecasts. There is a lot of activity on the part of the software industry to create robust risk management systems, and there are a few dominant players in the market. In the absence of integrated systems, profitability analysis traditionally was based on simplifications, such as average costs. Again, the traditional utility model demanded no more. After all, rates were set based on average costs and return on asset formulae.

Since the different commodities have traditionally been the turf of different departments, it is not uncommon to see multiple systems being used by the same organization, one for electricity, one for natural gas, and yet another for other services. The two basic problems that have forced this traditional stovepipe architecture on systems are: 1. Different commodities require different contract determinants with different data types. 2. The rules of interaction between the determinants are different for different commodities.

The Case for Integration

181

To start with, the simplest of financial contracts for electricity specifies, over and above the basic information about the counterparty, the points of delivery, the duration (perhaps monthly), megawatts by time blocks, and a price per megawatt. A similar contract for natural gas specifies the same kind of parameters, and would be expressed in MMBtus, day-blocks, and a price per MMBtu. Staying purely within the world of financial contracts, the simplest of weather hedges involves a similar set of determinants, that is, a location, number of units by time blocks, price per unit, and the degree differential. If all life were that simple, you wouldn’t need any systems, and if you insist on having one, almost anything can handle these different commodity contracts. But once you get into products that go beyond these simple ones, the divergence starts. For example, if you consider an energy exchange transaction based either on a time period or delivery point,1 the point of differential is a location in one case and a date and time in the other. Life becomes a lot more complicated when you look at the physical side of a transaction. Assume for a minute that you are using a number of contracts to manage a generation portfolio. What else do you need to capture to assess the financial risk of the simple contract above? The underlying generation and transmission introduce risk determinants that can number in the hundreds, from fuel price risk to plant outage and, potentially, the emission credits associated with the fuel type. Thanks to the fragmented nature of state-by-state and utility-by-utility energy deregulation in the United States, a contract for retail delivery of energy has become one of the more complex things to manage. The number of determinants that allow you to schedule, deliver, and settle the contract run into the dozens, and are different depending on the physical location of your client, let alone commodity. The service contracts that are being signed today involve not just the energy, but also a whole host of secondary market products, such as installed capacity, ancillary services, and schedule coordination for large industrial and commercial customers, and energy audits, lease-back programs for equipment, and building services for smaller customers (See Figure 11.1.).

THE CASE FOR INTEGRATION Two levels of integration are necessary: first, across the multiple products and commodities, and second, across the functional areas we describe previously. The market participants in the energy industry are already dealing

182

CONVERGENT SYSTEMS

Marketing

Contracts

Products

Offers

Pricing

Customer Acquisition Campaign

Sales Support Customer

Customer Profitability Models

REMNet Credit Data

Enrollment

Usage Data

Orders & Usage

A/R and Revenues

Provisioning

DISCO

ISO/PX

TRANSCO

Order Processing and Fullfillment

Logistics & Distribution

GENCO

Price Risk Books

Billing, Payment & Collections

A/P & Settlements

Forecasting and Inventory Control

Commodity Trading

Customer Data Warehouse

Usage/Buying Profiles

Chum Analysis Source Analysis

FIGURE 11.1 The ESCO Functional Diagram with multiple commodities in contracts and trades on exchanges. Therefore, a complete picture of the risk profile of any book requires merging the risk from multiple commodity books. As is being done today, this can be achieved by doing the value-at-risk (VaR) calculations independently and then, consolidating them. The problem is that the transactions are increasingly becoming interdependent As a matter of fact, they have always been interdependent in the physical world. It is just that most systems chose to ignore this because of the underlying complexity. In one of the pioneering transactions, Enron did an electricity-for-gas swap in the Northeast to move the natural gas to Chicago as early as 1996, motivated by the record high gas prices. Any hedging strategy for a power producer involves transactions for electricity and fuel choice of natural gas, coal, oil, or nuclear.2 Any price move in the price of fuel is going to impact the price of electricity, and given the mix of the fuel portfolio, the impact is going to be plant specific in terms of changing the risk profile of the simplest of electricity

The Case for Integration

183

contracts. In an ideal world, once the fuel price volatility is entered into the system, it should reassess the statistics for the entire portfolio. The case for integration across the different functional areas is stronger and more obvious. In one sentence, it is just that with so many moving pieces, you need everybody on the same page. The fact that multiple commodities are involved, in combination with the need to bridge retail and physical transactions with the financial transactions make this even more critical for smooth operations. As we mention earlier, the first problem that needs to be solved is the multiplicity of contract determinants. Object-oriented design is our starting point. However, the traditional object programming needs to be taken to one higher level of abstraction, that is, almost a meta-object design. The idea is to create object templates to represent counterparties, commodities, or services and their determinants. Remember that these templates should make no assumptions about the format and content of the determinants. In effect, the determinants are self-specifying in terms of their formats and any restrictions on their content. For example, if the delivery point is considered a determinant, its value should be restricted to a list of delivery points we are licensed to work in; and some of the contracts need a delivery point while others do not. The second step is to build the ability to create and manipulate these contract objects on the fly. Therefore, the programming behind these objects reflects not the traditional concept of application design, even in the object oriented model; it follows a compiler design. This is very familiar to systems engineers who have been building tools for other systems builders. An example will make the difference clear: Consider a spreadsheet program, such as Microsoft Excel. In the sixties and the seventies, a significant amount of computer programmers’ time and energy were devoted to creating custom financial models on mainframes, using languages such as APL and FORTRAN. Then came the spreadsheets which, in effect, presented the end user with a set of tools for creating these financial models. A spreadsheet is nothing but a collection of rows, columns, and cells. Depending on what the user put into the cells, the underlying software interprets it either as text, numbers, or formulae; if they are formulae, then the rules of interpretation for the cell take over, and perform the necessary operations. The person who created and sold the spreadsheet software made very minimal assumptions about the content the user was going to put into the cells. The two promises that spreadsheets delivered on were the open architecture and ease of use, making it the “killer app” to such an extent

184

CONVERGENT SYSTEMS

that Lotus 1-2-3 broke even within one month after shipping the first copy of its software in 1982. Today, spreadsheets are used to keep track of everything from databases of personal music collections to complex financial models to price weather derivatives. Excel is now a staple on any computer used in the business world today. You can consider the spreadsheet as a “convergent” system that allows manipulation of completely unrelated objects within the same framework. That brings us to the next problem to solve: the underlying data storage model. Unfortunately, the spreadsheets are rather weak on this score. The unrestricted open data content makes the organization of information behind the scenes hard. It is this problem that limits the utility of spreadsheets when it comes to operating in a collaborative environment in an organization. Each spreadsheet exists as an island on its own, making sharing of data and logic cumbersome. So, in building a new system environment, we are forced to limit the degree of openness as a trade-off to allow the data and logic component organization in the background. Although there are software packages that can deliver true object storage, their performance is far from ready for prime time in a mission critical application. So, we are forced to operate in the relational database environment: Oracle, SQL Server, DB2 and so on. The key goals of the data architecture are: ■ Assume as little as possible about the data. By that, we mean minimizing column definitions that are specific to a given contract or product determinant. ■ Build the rules of validation and interpretation of the data elements into an encapsulated object structure. The final piece of the solution is the “mediation”: the software component that acts as the intermediary between the relational data structures and the creation of the business objects. Embedded within the mediation piece is the logic that specifies the relationships between the determinants and the transformations of one set of determinants into another, and performs the calculations to output a standard set of values, whether they be prices for billing or statistics for measuring the risk and variability. Bundling two or more of such products together, you can now create complex cross-commodity offerings. The underlying statistical models are mea-

The Case for Integration

185

suring the variability in the net effect of price or volumetric forecasts, thereby giving you a true measure of the risk involved in this complex contract. This methodology is opposed to the traditional mechanisms, which compute the statistics one at a time, and then electronically staple the results together. If all of the above sounds like abstract computer theory nonsense to you, that is because it is. Let us try to clarify, using an example. To convey the power of this approach, we take a simplified example: the timeperiod and delivery point differential electricity products we just described. Start off with the simple underlying product, namely the simplest of contracts—time block electricity commodity. This has the following determinants: ■ ■ ■ ■

Term: start date, end date Time blocks: start times, end times (e.g., 24 by 7 or 16 by 5) Megawatts delivered Price per megawatt hour

Using the forward price curves, it is easy enough to measure the statistics associated with this product. To move on to the time differential contract, in our system, the recipe is straightforward: You create two instances of this product. The only difference is the term, where there is a relative shift for the time differential. Combine these two products into a bundle, with the rule for computing the bundle price specified as the difference between the two components. Feed the same raw historical data that you used for projecting forward prices into this product, take the output, and feed it back into the algorithms that you used to create the simple forward curves3 and you now have a forward curve for the differential product! You can use the same methods as you have for the simple block contracts to measure the variability or to mark this contract to market. What is the difference between this approach and examining the two components independently? As it turns out, a lot. ■ From a pure statistical point of view, if the two components are highly correlated, which is the case more often than not. Examining the summary statistics for the two products and then deriving the statistics for

186

CONVERGENT SYSTEMS

the combined product usually results in higher variability coefficients, thereby inflating the risk associated with the contract. If you are using multivariate regressions or time-series processing techniques such as ARIMA, then the high correlation among the variables makes the resulting model extremely unstable. ■ From an operational point of view, notice that our recipe asks you to do nothing out of the ordinary. The effort involved in managing this contract is no more and no less than managing the simple block contract: you feed the historical data in, take the output, and feed it into your forecasting mechanism. This is obviously a lot easier than constructing custom distribution curves based on the time differentials, setting up the forward curves with time period offsets, and running the models. And here is where it gets even more interesting: To convert this product into a delivery point differential, you add on one more determinant called the “delivery point.” The rule for determining which forward curve to use will now depend on the delivery point. The rest of the recipe is the same. We are sure you’ve already thought of the next variation on the above product: a delivery point and time differential. Obviously, from the above example, you’ve probably figured out how to do it, too. Let us take this example one step further, and create a cross-commodity derivative product. To mitigate its risk, an aluminum factory would love to buy a retail electricity contract pegged to the spot price for aluminum on the NYMEX. We picked the example of aluminum since energy is a significant component of the cost of production in that industry. Once again, we can create a similar bundle, measure the variability of the resultant product (although the aluminum prices follow a different distribution), and come up with a pricing for this contract. The final piece of the puzzle in all the above contracts is the settlements and determining the profitability after the fact. The meta-object construction we describe above makes this part really easy. The product was set up in the system for analytic purposes before the contract and therefore, when the actual values are available, no further work is necessary in terms of programming or setups. Instead of forecasts, feed in the actual values, and print the results out on a bill. It should be and is that simple.

Integration

187

INTEGRATION The next problem we tackle is the integration of the system across functional areas. The first challenge is that the multiple commodity environments make it critical to integrate the data and logic components across the different functional parts of an organization. The second challenge is that the incoming data, whether their source is from utilities, real-time markets, or ISOs, for different commodities and services follow completely different formats. Unlike the financial industry, the energy industry data exchanges have almost no standards, and those standards that do exist are respected more in violation than in compliance. Fortunately, over the past year, a new standard called Extensible Markup Language or XML is gaining ground across many different industries and software and hardware platforms. In the traditional application software design, XML is used primarily as a labor saving device. It eliminates the necessity to create custom parsing routines for each data format; it does not eliminate the requirement for the two trading partners to agree on the list of data elements to be sent, and what they should contain before the data exchange can take place. The receiving system has to know how to map the incoming data into its structures. In our meta-object model, XML becomes much more powerful. As a data specification technique, it shares a lot of the desirable qualities of our meta-object model: For one thing, the definition of the data in XML is itself specified in XML. Thus, the specification as well as the data itself can be made part of one single stream, making it self-specifying. Combine this with the fact that our own product and contract objects are created on the fly and you have a really powerful mechanism to use the incoming data to create the determinants you want. Indeed, you can accommodate determinants that you did not know about when you started out! The other benefit from XML is the ability to communicate structurally different data within the same framework. This feature is crucial to any system that hopes to handle multiple commodity contracts. To illustrate by an example, consider the data needed to settle a nonfirm or interruptible contract. Over and above the standard commodity determinants, there is an array of data corresponding to interruption events, which in a given time period, may or may not be present. The XML specifications can accommodate such nested structures, and allow for the fact that they may be present.

188

CONVERGENT SYSTEMS

The final piece of integration is the need to receive one data stream and make it follow the path of all the business processes. This is commonly called “straight-through processing.” Depending on the systems environment, an incoming data stream may have to be routed to many different systems, and updated concurrently. Over the past year, several software vendors have introduced a technology called “transaction hubs” to address this problem. Excelergy’s eXact and Lodestar’s Transaction Hub are examples of such software (see Figure 11.2). In the spring of 2001, Microsoft introduced a set of software servers under the .NET framework, which delivers a complete transaction hub technology, and typical of Microsoft, is priced at a fraction of the price the other vendors have been charging. Microsoft’s BizTalk and Commerce servers deliver a transaction hub solution based almost completely on XML and COM+ objects, using Internet and Active Server pages as the transport mechanisms. Perhaps borrowing a page from FedEx, the software has security, encryption, tracking, and document audits built in every step of the way. Going one step beyond the other vendors, BizTalk has a built-in integration layer whose function is to provide systems integration with existing software via message queues and application integration components. With the use of COM+ objects and active server pages, the integration into existing systems can be accomplished without any need for specialists—once again, translating into significant cost savings.

In-House Systems

FTP

SMTP

Data Streams

EDI X.12

Flat Files

Custom Formats

XML

Integration Layer

Trading Partner

HTTP

Data Translation Layer

Trading Partner

Contract Structuring

XML

Transport Layer

Trading Partner

FIGURE 11.2 Integrated Transaction Management

Forecasting & Scheduling

Trading & Risk Mgmt Billing Settlements

Conclusion

189

CONCLUSION As we mentioned earlier, these are some of the ideas we have been working with at NirvanaSoft Inc. over the past year. As we look into the nearterm future, two things are apparent: (1) the commodities and services covered by contracts are converging; (2) as a result of this convergence, the need for ever more complex contract structuring and settlement tools is imminent. Fortunately, the software technologists have been busy. The current application-oriented software systems are going to give way, much like the FORTRAN models of the 1970s, to systems that deliver tools to the end user via compiler-based designs and open architectures. The dominance of Internet and related data transfer protocols such as XML are facilitating the move toward this convergent world. The business model of the industry is undergoing dramatic changes due to globalization and deregulation. It is time for the tool builders also to rethink their fundamental beliefs in how systems should be designed and implemented.

CHAPTER

12

Energy Risk Management in the Merger Context Howard L. Margulis, Esq.1 Partner, Squire, Sanders & Dempsey LLP

INTRODUCTION The recent confluence of new accounting rules governing the recognition and treatment of hedging and derivatives positions and a re-examination of traditional merger concepts of “material adverse change” or “materially adverse effect” are certain to increase the due diligence aspect of energy sector mergers and acquisitions for the indefinite future. This chapter analyzes the background of these trends, the current methods of managing those issues, and offers some suggested techniques for future consideration.

FASB: A SHORT STORY The Financial Accounting Standards Board (FASB), based in Norwalk, Connecticut, is responsible for promulgating rules for the accounting profession, rules that have global impact as the final authority on the financial reporting for U.S. companies, and in particular those publicly held companies whose financial reports have the ability to cause shocks and tremors in the stock market. Generally, FASB pronouncements have all the appeal (and none of the palliative effect) of milquetoast. However, in 1998, FASB turned the energy industry on its head.

191

192

ENERGY RISK MANAGEMENT IN THE MERGER CONTEXT

FASB RULE 133: A CAUTIONARY TALE In June 1998 FASB issued a preliminary report entitled “Accounting for Derivative Instruments and Hedging Activities.” Simply stated, Rule 133 is intended to require companies to include on their balance sheets every derivative instrument they hold, including certain instruments embedded in other contracts. By design, Rule 133 was promulgated to diminish volatility in financial markets, ostensibly in the wake of certain high profile financial problems, such as the Orange County, California, bond default, the failure of London’s Barings Bank (now part of the Dutch ING Group), the crash of certain power marketing firms in the Midwest, and other spectacular financial bombshells. Before Rule 133, hedging and derivative positions were treated for accounting purposes as off-balance sheet items, and therefore not subject to specific disclosure or valuation analysis. All of this has changed, and so dramatically, that FASB delayed the implementation of Rule 133 for more than a year. Indeed, Rule 133 did not take effect until January 2001. And it was not without a struggle that Rule 133 ever became effective—more than six years of hearings, innumerable consultations with the U.S. Securities and Exchange Commission (SEC), and industry lobbying from the financial, energy, and transportation sectors have contributed to making Rule 133 one of the most controversial accounting rules in years. Ironically, but perhaps not surprisingly, the near panic caused by the introduction of Rule 133 has resulted in a final version of the rule, which runs more than 300 pages and has been the subject of more than 125 written interpretations from the Derivatives Implementation Group (DIG), a consortium of Rule 133 specialists from the Big Five accounting firms, major investment banks, and large industrial corporations.

Rule 133 in a Nutshell The key aspects of Rule 133 are: ■ All derivatives must be carried at fair value and recorded on the balance sheet. ■ Gains and losses from hedging activities are not assets or liabilities, and therefore, should not be deferred from recognition.

The Wide Net

193

■ The definition of a derivative is now broader, and includes futures, swaps, caps, floors, collars, options, swap options and some forward contracts. ■ Special and different accounting treatment has been established for three kinds of hedges: (1) hedges of changes in the fair value of assets, liabilities, or firm commitments (fair value hedges); (2) hedges of variable cash flows of forecasted transactions (cash flow hedges); and (3) hedges of foreign currency exposure of net investments in foreign operations. ■ Changes in fair value of derivatives that do not meet the above criteria in these categories are included in income. ■ The ineffective portion of a hedge is recognized in income and not deferred, thus creating potential volatility in income and earnings. ■ Cash flow hedges are recognized in other comprehensive income, thus creating potential volatility in equity.

THE WIDE NET FASB Rule 133’s stated intention of creating more complete disclosure about corporative activities and reconciling inconsistencies in the accounting treatment of derivatives has the salutary (depending upon one’s views) of casting a wide net upon energy companies. Some commentators has reached the conclusion that if you are buying or selling energy in the wholesale commodities markets, whether or not you hedge, assume it is a derivative unless proven otherwise. Others have called FASB Rule 133 the “Y2K of Accounting for Energy Derivatives”; that is, non-compliance (or minimal compliance) is not an option. Certainly, companies will have to implement significant recordkeeping, at least quarterly, as to each trading position in order to be able to deliver financial reporting within required deadlines. And companies will be compelled to record information on a mark-to-market basis for both derivatives and physical commodities. While it is certainly true that the same contract could be a derivative subject to Rule 133 for one company but not a derivative for the counterparty, the definition of a derivative has also been expanded. Now, a “derivative” includes such items as convertible debt instruments held for investment, certain commodity purchase agreements (especially to the extent based upon market-price terms), some structured instrument notes, and certain insurance contracts.

194

ENERGY RISK MANAGEMENT IN THE MERGER CONTEXT

A WIDE IMPACT To meet the challenge of Rule 133 companies must focus on the strength and integrity of their data, because hedging (in)efficiencies will directly impact reported earnings. For years it was routinely understood that the cost (time, recordkeeping, analysis, monitoring, and capital) of the “perfect hedge” was too high (and dilutive). Accordingly, the standard hedging strategy was to disregard the ineffective elements of the derivatives portfolio. Under Rule 133 that ineffectiveness is recorded to the income statement, and is reflected in quarterly results. Moreover, any change in the derivative position must also be reflected in the financial reporting, and not deferred. Meeting these requirements will be a substantial data management task in and of itself, without regard to the obvious impact on results, which this new reporting will spotlight. And in order to offset a derivative’s gains or losses against the recognition of the hedged item in the income statement, special accounting for such “qualifying hedges” requires that a company formally document, designate, and assess the effectiveness of such transactions and make full disclosure of such programs.

EARNINGS: A SPECIAL IMPACT Rule 133 can change a company’s earnings picture substantially. Indeed, Rule 133 can dramatically increase volatility on the income statement, making the balance sheet look significantly larger or alter the debt-to-equity ratios by an order of magnitude. For instance, in the quarter ended September 30, 2000, Barrett Resources Corporation reported its net income pershare for nine months at $1.37, excluding mark-to-market recognition for derivatives positions, but only $.76 per-share when the derivatives valuations were included.2 Such disparities, while perhaps not uniform in the energy sector, would not be altogether uncommon, either.

DERIVATIVES IN THE MERGER CONTEXT: A NEW PARADIGM Ironically, until recently some commentators have suggested that derivatives could be a boon to the energy M&A frenzy. It was thought, and not without substantial empirical data, that utilities could point to energy de-

Merger “Outs”

195

rivatives and could ameliorate certain market power concerns. For instance, the availability of commodity swaps and other instruments would have the effect of lessening the traditional Herfindahl-Hirschman Index (HHI) measure of market concentration around transmission-constrained areas. Much of those ostensibly positive impacts, however, would appear to be overrun by Rule 133’s emphasis on current accounting for derivatives positions. Accordingly, rather than relying upon energy derivatives to support merger approval, energy derivatives are now at the center of management’s concerns in planning and executing merger transactions. Rule 133, has, in essence, turned energy derivatives from a supporting tool for energy transactions to yet another (albeit significant) item in the due diligence examination.

MERGER “OUTS” At the time of this writing3 the world, and the United States in particular, is reeling from the impact of the terrorist destruction of the World Trade Center Towers in New York City and the Pentagon, in Arlington, Virginia. At the same time, legal and financial specialists are being called upon to face the challenge posed by announced and future mergers for which the impact of these events may have a substantial economic impact not previously anticipated.4 For years, merger agreements have contained rather uniform language allowing one party or the other to withdraw from a transaction on the basis of “material adverse changes” or “materials adverse effects,” called MAC provisions. A typical MAC clause reads as follows: “Material adverse change” or “material adverse effect” means, when used in connection with this Plan and Agreement of Merger, ABC, DEF, and XYZ Merger Subsidiary/Acquisition Corporation, any change, effect, event, occurrence, or state of facts that is, or would reasonably be expected to have a material adverse effect on, the business, properties, financial condition or results of operations of such person and its subsidiaries, taken as a whole (other than effects relating to the oil, gas and/or power industry(ies) in general), or prevent such person from performing its obligations under this Plan and Agreement of Merger or prevent consummation of the transactions contemplated hereby, provided, however, that the term Material Adverse Effect or

196

ENERGY RISK MANAGEMENT IN THE MERGER CONTEXT

Material Adverse Change shall not include any effects on the business, properties, financial condition or results of operations relating to or caused by generalized events or occurrences in the energy industry, including but not limited to, [specified events].

A COMPOUND PROBLEM Reconciling the impact of Rule 133 on a company’s operating cash flows and earnings is difficult enough in the financial reporting area. Assessing the resulting disclosures from the perspective of merger analysis is likely to strain even the most well-prepared coterie of transaction advisors. For instance, in a typical utility merger, due diligence consumes enormous time and resources, focusing on such complex matters as ERISA integration, workforce rules, and collective bargaining agreements, environmental externalities and a host of other socioeconomic considerations even to the level of maintaining charitable support in the host communities of both entities. With Rule 133, the task becomes harder still: how to reconcile diverse commodity positions while simultaneously monitoring the energy pricing volatilities that might give one regretful party or other the right to invoke the MAC clause and withdraw from the transaction. A great body of scholarship has testified to the resilience (and occasional inflexibility) of the MAC provisions in current use. Most notably, in the recent Delaware Chancery Court decision in In re IBP Shareholders Litigation, Consolidated Dkt. No. 18373 (June 18, 2001, as corrected), Vice Chancellor Stine wrote: [A MAC] is best read as a backstop protecting the acquirer from the occurrence of unknown events that substantially threaten the overall earnings potential of the target in a durationally-significant manner. A short term hiccup in earnings should not suffice; rather the Material Adverse Effect should be material when viewed from the longer-term perspective of a reasonable acquirer. While certainly reasonable in its conclusion, Vice Chancellor Stine’s opinion also adds some complexity to the Rule 133 problem, for it focuses the analysis on temporal considerations, which admit of no easy answers

A Practical Problem

197

when one is valuing and accounting for derivative positions. For although it is precisely because of the varied timing constraints attendant to derivative transactions that Rule 133 sought to establish a quantification requirement based upon current values, the problem with hedging scenarios is that the current value of a position may be entirely different the day after the financial reporting date. Hence, the invocation of the MAC to dissolve a Plan and Agreement of Merger will require nearly superhuman financial modeling capabilities, with the data inputs becoming even more critical than ever before.

A PRACTICAL PROBLEM As energy mergers have continued to accelerate despite downturns in other industries, including efforts (some successful, others not) by foreign companies to invest in the U.S. utility and energy sector, a number of problems remain to be addressed, including the following. ■ Power marketing affiliates of United States investor-owned utilities are frequently involved in substantial energy trading activities, leaving their ultimate parent corporations with potential liabilities in the event of a dramatic change in the energy commodity pricing markets. ■ Industrial companies who have outsourced energy procurement to third-party specialists, including marketing firms, will need to pay close attention to contractual terms, which shift risk to them and, by implication, compel Rule 133 disclosure and accounting treatment. ■ Energy companies with consolidated accounting will need to manage affiliate energy trading schemes, particularly those that impose market pricing regimes on fuels, capacity, and transportation on energy production facilities, even certain facilities that are “special purpose” companies for project finance purposes. Each of these presents a new risk for management in the energy industry. Affiliate power marketers are likely to be consolidated for tax and accounting treatment at the parent company level, presenting financial professionals with the enormous task of reconciling trading positions and accounting for them in the financial presentation to investors and analysts. Industrial firms may have felt safer with outside professionals handling

198

ENERGY RISK MANAGEMENT IN THE MERGER CONTEXT

energy procurement, but many of those arrangements leave the risk of commodity markets on the ultimate customer. And for those projectfinanced entities, there may be accounting and tax consolidation through intermediate affiliates of the reporting company, particularly if ownership and/or earnings distribution channels leave the ultimate parent with the unenviable task of reporting the good, the bad, and the ugly about fuel supply contracts.

SOME SUGGESTED SOLUTIONS There are already existing solutions for managing energy derivatives, many of which are software based. Some of these programs currently implement present value forecasting, risk assessment levels, and transaction monitoring techniques. Nevertheless, it is critical that senior management be vigilant in guarding against complacency and insist on fundamental controls over the trading portfolios in all operations. To be sure, accounting, tax, and legal professionals are (or should be) aware of the impact of Rule 133 on potential transactions. But it is equally important that day-to-day oversight be in place to prevent catastrophic results in financial reporting and the concomitant impact on enterprise-wide planning. Some firms have put forth a senior officer as the Risk Management Czar. Likewise, other firms have put energy trading specialists on active retainer, with the right to both oversee trading operations and report directly to senior management on an as-needed basis. Undoubtedly, while software solutions will play a large role in this effort, given the high costs of negligence, it is imperative that empowered human capital be involved as well.

CONCLUSION We have seen that the interplay between new accounting rules and merger analysis creates a climate of heightened scrutiny for energy firms. What may not be so obvious, however, is that this confluence of issues provides an opportunity for firms that exercise proper planning methods in pursuing new acquisitions. That is, those companies that have undertaken the effort to have their own risk management portfolios in order will have the luxury of utilizing that knowledge base in analyzing other

Conclusion

199

firms in the sector with a critical eye. And to those firms that have not yet implemented a comprehensive management of their own trading positions, there is now even greater incentive to do so. For, in the end, those firms that fail to do so will either be left out of the merger mania sweeping the energy field, or be acquired at a less than optimal price for their shares. Neither prospect is one for which senior management wants to be remembered in the board room.

CHAPTER

13

Managing Energy Risk for Industrial and Large Commercial Entities Kelly Douvlis

INTRODUCTION In order to define energy risk, we need to understand what led the industry to deregulate, its current landscape compared to the past, the underlying factors of the continuous change toward competitive markets, and possible remedies to achieve fairness and level the playing field for all market participants. After defining all factors contributing to energy risk, we should be able to suggest strategic alternatives to possibly minimize it or effectively manage it.

THE ANNOUNCED PURPOSE OF ELECTRICITY DEREGULATION The power industry has shifted from a regulated monopoly to competition driven by the free market forces created by the Energy Policy Act (EPAct) passed in 1992. FERC 888, the later order of the Federal Energy Regulatory Commission, established the requirements for utilities to open the wholesale power markets to competition. The goals of the above legislation were to achieve lower price of the electricity commodity, ensure reliability, and provide open and fair transmission access to all market participants.

201

202 MANAGING ENERGY RISK FOR INDUSTRIAL AND LARGE COMMERCIAL ENTITIES

BRIEF OVERVIEW OF THE PAST 10 YEARS OF DEREGULATION During the past 10 years since electricity deregulation began in 1992, the power industry has undergone enormous changes as it faces new business practices. First, the rules set for competitive markets and the dynamics influencing them are very different from the rigid regulatory regime of the past. Second, the lessons learned from this transformation process from the past monopolistic business environment to free markets is continuous. There is one certainty, however, and that is given the history of the events of the past decade, nothing will ever be the same. There will be more changes for the years to come until the industry matures to enter a competitive environment of business practices.

OUTCOMES Deregulation of the power industry, as with deregulation of other industries, has brought fundamental changes to both the supply and demand side of the electricity commodity. This has impacted electricity consumers. Their main concern has been at what price will they pay and how reliable will the product be delivered. Therefore, the focus of this chapter is first to clearly understand the factors contributing to price and reliability risks, and second to suggest possible alternatives to best manage these new risks. Furthermore, this chapter is addressed to industrial and large commercial energy consumers.

ELECTRICITY PRICES IN DEREGULATED MARKETS Before deregulation, electricity commodity prices were predictable with a slow but sure upward movement (cost-based pricing) and with known reliability on the responsibility of one supplier (the local utility). Nine years into deregulation we have learned very well that the price of electricity as well as gas is not predictable. The power markets are indicating high price volatility as they slowly move toward better liquidity. The past five years of market activity is not enough yet to give us a long-term knowledge of the electricity forward curve. There are several reasons for lacking a long-term forward curve, the foremost being that there has been constant change in the industry landscape from the supply side (generation, transmission) and an incom-

What Is Really Affecting Price?

203

plete deregulation on the demand side. It is safe to say that it will take more than the next five years to have a better understanding regarding forward curves, which will result in more predictable prices. Even then, the market’s forces will always influence pricing as we continue to see with oil prices, which have been deregulated in the United States for more than 20 years.

WHAT IS REALLY AFFECTING PRICE? There are several factors that influence electricity pricing. These include rapid demand growth and short supply, reliability, an inadequate and aging transmission infrastructure, and the market power of electricity generators.

Rapid Demand Growth Coupled with Short Supply Margins between available generation, transmission, and load are thin. Nationwide, electricity consumption grew 31 percent in the decade between 1988 and 1998. Consumption grew about 10.6 percent between 1994 and 1998 alone. According to the North American Reliability Council data, generating capacity in the United States will need to expand by about 12 percent by 2007, an increase of 91,000MW of installed generation, in order to maintain adequate reserve margins.1 The United States DOE now projects a need for more than 390,000MW of new capacity by 2020.2 Several studies project the need for 38,000 miles of interstate gas pipelines, 255,000 miles of gas distribution pipelines, and 8,800 miles of electric transmission lines. Total future investment of more than $250 billion is needed to support increased electric and gas consumption through supply-side measures.3

Inadequate and Aging Transmission Lines Transmission lines are inadequate in certain high growth population areas, which result in transmission constraints especially during seasons of high demand. These events will cause shortages and increased demand, which will result in higher prices.

Market Power of Generators Electricity’s unique characteristics are that it must be generated and consumed simultaneously, across an interconnected network where all users

204 MANAGING ENERGY RISK FOR INDUSTRIAL AND LARGE COMMERCIAL ENTITIES are affected by the consumption patterns of all other users. When the margin between generation and load is thin, that gives a great deal of market power to generators, which are able to command very high prices on the spot markets, resulting in the spikes that we often see occurring. This change was also enhanced by the rapid pace of consolidation among the owners of generation facilities in the United States, yielding fewer and fewer competitors in regional power markets. In 1935, 13 companies controlled more than 50 percent of all investor-owned utility (IOU) generation. After the breakup of the major holding companies following passage of the Public Utility Holding Company Act (PUHCA) in 1935, and for the decades between 1955 and 1995, one had to combine the assets of the 200 largest companies to cross the 50 percent threshold. Today, 10 companies own 51 percent of IOU generation.4

Reliability In order to access the reliability part in delivering the commodity, we simply need to understand the power delivery chain and its geographical regional characteristics. The United States is divided into pockets of generation pools controlled by Independent System Operators (ISO). These ISOs coordinate the operation of central plants in a defined region, ensure reliability of the grid, and typically oversee financial transactions of market participants. A simple supply/demand imbalance or an inefficiently operated ISO will have a significant impact on both reliability and cost. We are also faced with aging transmission lines and with inadequate capacity, which creates transmission constraints contributing defaults in reliability. From the distribution level of the commodity delivery, we might be faced with line bottlenecks from pockets of high demand, faulty lines due to aging systems, or insufficient maintenance on overhead lines, and the inability to meet reliability requirements above three nines (the measure of reliability).

DEVELOPMENTS TO IMPROVE PRICE AND RELIABILITY As the recently released Bush Administration’s National Energy Policy report states, “A fundamental imbalance between supply and demand defines our nation’s energy crisis.” This translates into a shortage of generating supply, inadequate and aging transmission lines, and demand

Understanding Energy Risk

205

growth. The past four years have seen most of the activity concentrated in building new generation in high demand locations. Also existing generation assets, changing ownership and developments to increase their efficiency, are underway. There is not much done to build new transmission lines due to their high initial capital and complicated regulation. However, there are plans in the future to improve transmission capacity in the most congested regions. The current movement to establish regional transmission organization (RTOs) would also improve the electricity flow and create a fair playing field for market participants. Energy conservation and applications of currently available technology to reduce demand have not been embraced by consumers since the 1980s. The demand side management programs, which were supported by the utilities, government, and consumers before deregulation, are now almost gone. Utility interests now are to invest in their transformation of businesses in order to keep alive. Government funding and programs also faded away due to loss of infrastructure for such programs. On the other hand, consumers during the 1990s were less concerned about energy pricing due to the economy’s strength and also the expectation that deregulation would lower prices. However, reducing demand could easily contribute to price reductions or stability and also reliability. Currently, the technology has improved tremendously for energy conservation, load management, metering devices, cogeneration technologies, distributed generation, and control systems. Investing in applications of current technologies would impact pricing and reliability for the future.

UNDERSTANDING ENERGY RISK Deregulation of the gas industry and the evolving deregulation of the electricity industry might leave the end users with the notion that the cost of both commodities will go down. However, they have not really come to the point in understanding that the open markets have created exposure to both supply (reliability) and price risks. While moving toward full-scale deregulation of the energy industry, the following risks are ahead for large industrial and commercial end users. ■ Price Risk Price risk is associated with fuel price changes, supplydemand imbalances that might result from extreme weather conditions

206 MANAGING ENERGY RISK FOR INDUSTRIAL AND LARGE COMMERCIAL ENTITIES

■

■

■

■

or poor capacity planning and development or unforeseen generation units outages. Financial Supply Risk Financial supply risk implies that different suppliers default on delivery to end users suppliers. That might not result in a physical delivery interruption but end users could be hit with monetary penalties for consuming energy, should their suppliers default on delivery to the local distribution company (LDC). This problem exists today more in the electricity industry than the natural gas industry in the United States. The risk difference between the two industries is that the natural gas industry deregulation is almost mature, which means that the industry has reached consolidation to fewer suppliers, and problems only arise by the pipeline’s capacity limitations during peak demand periods. On the other hand, electricity deregulation has a long way to go yet. As the electric power industry continues to deregulate on a state-by-state basis, reliability would be a function determined primarily by transmission limitations rather than by suppliers’ default. That would be resolved as all sellers gain nondiscriminatory access to transmission wires (transmission limitations), and procedures for long-term capacity planning and development are in place. Financial Demand Risk Financial demand risk refers to volumetric risk. In other words, it’s the cash penalties that could arise, should an end user fail to consume all the energy for which it has contracted or came short of the energy volume it had originally planned in consuming. Risk tolerance for price will determine the degree of financial firmness. Physical Supply Risk Physical supply risk is defined from the distribution end of the energy delivery system. The risks of such disruptions in electricity could rise in short-term planning as utilities try to mitigate this underperformance-based rate making. This could lead to cost cutting and to their market-driven based decisions for capacity investments. Most importantly as we have seen during the past few years, physical risk could come due to the new, less experienced players who enter and exit the electricity markets. Physical Demand Risk Physical demand risk is defined from the demand side where end user operations could cause price spikes or jeopardize reliability by stressing the supply line due to equipment or system malfunction.

Managing Energy Risk

207

MANAGING ENERGY RISK There are several methods and techniques available for combating potential risk exposures in the procurement of energy. The following are the major categories for minimizing supply and demand risk.

Managing Supply Risk ■ Financial Instruments: Futures and options, swaps, and insurance policies are all options for hedging energy price risks. However, it is expected that few end users will use such instruments directly. End users will price management risk through marketers. The reason is that financial instruments work efficiently in large books rather than in individual cases. Some end users will opt for straight fixed prices or will take on contract terms with marketers with some limited exposure to market price swings using indexed pricing in conjunction with floors, caps, and collars. The degree of energy price risk that an end user may have tolerance for will depend on its understanding of the new markets, the proportion of its total cost of energy usage, and the market in which it competes. At the end of the day end users have to work with marketers to reshape pricing terms based on the risk they can tolerate. ■ Market Intelligence: Knowledge of the markets and rigorous supplier qualification by the end user will minimize potential supply risks. An end user that deems energy cost as a strategic component of its operating cost in order to compete in today’s global economy needs to acquire greater market intelligence by its own and by its relationship with a legitimate energy marketer. There is a simple procedure to evaluate the prospective energy marketer by inquiries regarding its markets activities and share, credit, corporate guarantees, contracts (short and long term), agreements for liquidate damages for nonperformance, business, ownership of generating or gas production assets, offer of risk management services, offer of full-service provider including outsourcing, and energy conservation. ■ Contracts and Conditions for Supplying the Energy Commodity: Risk adverse end users are likely to seek contract terms that shift full responsibility to energy suppliers for default penalty payments. The contracts will also seek terms that limit marketers’ abilities to invoke force majeure. This will be determined at the lower end of the market

208 MANAGING ENERGY RISK FOR INDUSTRIAL AND LARGE COMMERCIAL ENTITIES price advantage. In general, price would be set according to risk shift within the terms and conditions of the contract. ■ Qualifying Energy Suppliers: Suppliers would access their supply by following a two-step procurement process consisting of a request for qualifications followed by a request of proposal. Besides this basic process it’s useful to check suppliers’ market share, ownership of generating assets, offering of corporate guarantees, and good standing credit reports.

Managing Demand Risk ■ Demand Side Management: The ability to react physically to market price signals by knowing how energy is used would necessarily enhance an organization’s willingness or ability to take on price risk. A physical investment by the end user such as backup supply, demand management capabilities, or distributed generation contributes to a position of being able to manage better price risk with suppliers. A great alternative to managing energy risk is to develop an alternative fuel capability and switch when prices are favorable. Self-generation would contribute to substantial savings and flexibility to assess lower-priced supplies. ■ Benchmarking Energy: “Load profiles” projected to cost data could establish standardized best practice activities. Benchmarking quantifies a variety of energy-related cost factors. These include energy usage per unit cost of production, historical usage and cost comparisons between operations and facilities, and costs associated with power quality issues including outages and comparisons with industry standards. It can also help identify inequities, problems, or outright errors in utility bills. Benchmarked data can be used as an analysis tools, which could more confidently evaluate the impacts of competing supplier options on their business. ■ Energy Outsourcing: Outsourcing energy needs will increase a company’s competitive position by improving its financial position. Energy infrastructure equipment is capital intensive to acquire, upgrade, and maintain, and they can be removed from the balance sheet as assets. For example, equipment upgrades can be in central power plants, boilers, chillers, or back-up generation. Outsourcing would enable industrials and commercials entities to invest capital in other core business issues. ■ Technology for Energy Conservation: Reducing energy costs involves

Managing Energy Risk

209

more than competitive pricing, consuming less energy and more efficiently by implementing a wide range of technologies available to reach that goal. Energy conservation includes all areas of operations from energy infrastructure equipment to manufacturing processes, to lighting and controls, and building envelope upgrades. ■ Technology Contributing to More Efficient Power Pricing: There are energy technologies that provide an advantage to the power purchaser from suppliers’ artificial limits by allowing the applications of creative available technologies. Fuel switching technologies contribute to selecting the least expensive power supply at any given time. Load-shaping technologies affect power costs. Manipulating or shaping energy use could result in substantial savings by converting from a full requirements supply to a recallable or interruptible one. New advanced metering technologies represent the most costeffective first step toward long-term energy planning. Industrial or commercial entities with multiple sites could converge monitoring and telecommunications, thus creating new ways to manage loads, simplify billing, and aggregate loads. The benefits of this technology are to provide realtime purchasing of the commodity. A more sophisticated metering technology pinpoints electrical costs and potential power-quality problems. Energy software technologies is for systematic tracking, analysis, and reporting of energy use and cost. It quantifies where the energy is used, calculates and forecasts costs, and verifies performance of energyconsuming building and systems. It encompasses the energy accounting software critical to relate energy cost to budgeting decisions. Hybrid fuel plants are building systems and generators that can be powered by multiple fuel supplies. The key benefits of this technology is the ability to switch interruptible fuel supplies and avoid high peak demand charges. There are new generators, chillers, and boilers on the market that can burn different fossil fuels. Cogeneration, the ultimate hybrid-fuel technology offers tremendous benefits by generating power on site, thus providing independence and flexibility from single supply sources. It also ensures reliability of power at power supply constraints and significant savings at peak demand pricing. Besides cogeneration technology fired by oil or gas supplies there are other technologies slowly evolving in the markets, like microturbines, fuel cells, photovoltaics, and thermal storage.

210 MANAGING ENERGY RISK FOR INDUSTRIAL AND LARGE COMMERCIAL ENTITIES Lastly, there are technologies available to deliver power beyond three nines probably reaching and exceeding six nines in terms of reliability. These technologies are based on proper integration of existing systems coupled with creative solutions on DC link technologies and other system architecture configurations to integrate multiple independent on-site power generators.

HOW SHOULD INDUSTRIAL AND LARGE COMMERCIAL CUSTOMERS MANAGE ENERGY RISK? The key answer to the question of managing risk for industrial and large commercial customers is to develop a comprehensive energy strategy to address supply and demand risk by employing internal and external resources. In order to develop such a strategy industrials and commercials need to evaluate their internal capabilities resources. The following are some major categories to consider in order to establish an effective energy strategy to manage energy risk. ■ ■ ■ ■ ■ ■ ■

Energy market expertise within company Risk management capabilities Experience in financial instruments for energy products Energy conservation expertise Energy contracts evaluation expertise Available capital or financing resources for energy projects Energy related technologies expertise

In most cases, industrial and commercial customers lack, to a different degree, the above resources because their focus is concentrated on running their core business. Therefore, they are seeking external resources in order to manage their energy risk. External resources are energy supplier services companies and energy consulting firms. Some companies choose to employ a consultant to assist them in selecting an energy supplier and advise them in developing a suitable to their needs energy strategy. Others rely directly on their good relationship with the energy supplier. The most important issue for managing energy risk is to measure it against doing business and its effects on future growth. Therefore, industrial and commercial customers need to develop an energy strategy employing minimum internal resources coupled with the best available external resources qualified to accomplish their long-term energy needs.

How Should Industrial and Large Commercial Customers Manage Energy Risk?

211

Competitive markets bring new risk for companies to manage. Unfortunately, industrial and large commercial customers do not have this expertise inhouse and need to be more proactive in anticipating large price changes for gas and electricity that impact their bottom. The emergence of energy service providers (Escos) has been one response to creating a onestop shop to perform energy audits, financing and implementation of new measures, and to provide the engineering skill set for process upgrades. Energy has now become a critical component on company costs and profitability once again. High energy prices coupled with unprecedented price volatility have made the job of energy manager at industrial and commercial companies that much more complex. The logical extension of energy risk management would not be to create trading entities for this universe of companies but to acquire the solutions from outside energy service providers. This change will create more market liquidity as corporate boards change their perceptions of energy risk management into a fiduciary responsibility and a necessary part of doing business.

CHAPTER

14

Green Finance: The Emerging Financial Markets for Protecting the Environment Peter C. Fusaro President, Global Change Associates Inc.

INTRODUCTION The energy business is already globalized and multinational as large energy companies operate in more than 100 countries. This globalized business coupled with the spread of information across borders through media such as the Internet, CNN, and television have significantly changed public perception about the environment. In effect, pollution can not be exported across borders anymore as a new, globally conscious environmentalism has been created over the past decade. This global environmentalism is even more true of greenhouse gas emissions, which affect the entire planet. With carbon content increasing in the atmosphere at 4 ppm per year, the fear is that inaction will only lead to ecological disaster. Thus, the potential for web-based emissions trading is beginning as the web is borderless and international trading platforms are global. However, before this changes, we need to review where we are today and the emissions trading experience that has evolved so far. Environmental protection in many countries in the past has followed the heavyhanded command and control approach that has proved to be expensive and cumbersome. Instead, more cost effective market-based incentives using tradable permits have been gathering momentum over the past decade. The initial successes to date have been the trading of chlorofluorocarbons

213

214

GREEN FINANCE

under the Montreal Protocol of 1987 to save the ozone layer, and the United States emissions trading scheme for sulfur dioxide (SO2) for acid rain abatement, which began in 1995. The key has been the introduction of tradable permits combined with sanctions for non-compliance.

KYOTO PROTOCOL SETS THE STAGE The Kyoto Protocol of December 1997 obliged the following greenhouse gas (GHG) reductions over the 1990 period baseline: United States at 7 percent, Japan and Canada at 6 percent, EU, Switzerland, and most of Central and Eastern Europe at 8 percent. Each country is setting its own program to deal with emissions reductions. Commitments made by the Annex 1 countries can be fulfilled by the purchase of emissions rights from other countries. The agreement needs approval by 55 countries accounting for the 55 percent of Annex 1 (developed countries) emissions in order to be implemented. The Kyoto Protocol will come into force 90 days after the date on which it receives the required number of ratifications. The goal is to get the agreement activated in 2002 due to the Conference of the Parties (COP6) consensus developed at the Bonn meetings of July 2001 and COP7 meetings in Marrakech in November 2001. The resumed sessions of COP 6 in Bonn resulted in a flawed agreement that is often referred to as “Kyoto Lite.” COP7 further weakened Kyoto targets. The agreement over carbon sinks as a means of achieving the Kyoto targets has inevitably watered down the agreement further. The Kyoto Protocol of 1997 had sought greenhouse gas emission reductions from developed countries to 5 percent below their 1990 levels by 2012. The Bonn agreement cuts emissions only 1 percent to 3 percent. It also remains unresolved how carbon sinks will be recognized and how credits will be calculated. Besides the lowered emissions goals and lack of United States participation (which emits 25% of greenhouse gases) as well as that of developing countries, makes this an agreement without teeth. It is essentially a fig leaf to cover the lack of real progress on greenhouse gas emissions. Tables 14.1 and 14.2 show the percentage of CO2 emissions for each country in Annex 1.

215

Kyoto Protocol Sets the Stage

TABLE 14.1 Percentage of ANNEX 1 1990 CO2 Emissions

United States European Union Russian Federation Japan Canada Poland Australia Czech Republic Romania Bulgaria Hungary Slovakia Estonia Norway Switzerland Latvia New Zealand

Individual

Cumulative (without United States)

36.1 24.2 17.4 8.5 3.3 3.0 2.1 1.2 1.2 0.6 0.5 0.4 0.3 0.3 0.3 0.2 0.2

0 24.2 41.6 50.1 53.4 56.4 58.5 59.7 60.9 61.5 62.0 62.4 62.7 63.0 63.3 63.5 63.7

TABLE 14.2 European Union Breakdown of CO2 Emissions Germany United Kingdom Italy France Spain Netherlands Greece Austria Denmark Finland Sweden Portugal Ireland Luxembourg Total

7.4 4.3 3.1 2.7 1.9 1.2 0.6 0.4 0.4 0.4 0.4 0.3 0.2 0.1 24.2

216

GREEN FINANCE

MOVING BEYOND KYOTO The private sector will take the lead on the development of emissions trading markets since they have a vested commercial interest in emissions reductions. Compliance responsibility, however, will rest with government. There is also the strong belief that markets will form first and that government should not inhibit their growth. Since European, Japanese, and U.S.-based companies are now moving ahead to develop pilot programs, there exists a first mover advantage in this field since waiting for regulatory approval may prove more costly in the future. Emissions rights may be traded through bilateral transaction, listing on exchanges or through brokerage houses. In the Kyoto Protocol, it was envisioned that three international mechanisms would enable Annex 1 to reduce emissions to reach Kyoto targets beginning in 2008 through 2012. These mechanisms included emissions trading, joint implementation (JI) and the clean development mechanism (CDM). All three modes are currently being used. It is thought that bilateral trade between countries would be the most effective means to initially trade emissions. The emissions unit to be traded is one ton of carbon dioxide equivalent for the six greenhouse gases. NOX and CH4 (methane) emissions, two other greenhouse gases are more difficult to quantify in many countries. The United States has already established an Over-the-Counter market for both NOX emissions which began in 1999 and CO2 emissions. It has also completed cross-border trades with Canada. Since trading mechanisms will be part of any long-term approach to limiting GHG emissions, the emissions market is going forward on many fronts without Kyoto approval or U.S. participation in Kyoto. It is thought that actions taken today will most likely be grandfathered into the future revised treaty. Kyoto was meant to be flexible and allow market-based solutions to trading greenhouse gases as a carbon reduction strategy and as a means to influence the spread of energy efficiency technologies for industry. Governments also expect industry to make the largest greenhouse gas reductions and this falls heavily on electric and gas utilities, manufacturing, and automakers. Japan has been slow to establish emissions trading although many projects have been proposed by NEDO, a semi-governmental organization under the auspices of the METI and established in 1980 with the objective then of establishing alternatives to Japan’s oil dependency. Later, its mis-

The U.S. Emissions Trading Experience

217

sion was expanded to include energy conservation and research and development of industrial technology. This is the natural progression toward developing and implementing sustainable developments projects throughout the world. It is currently identifying and promoting potential private projects, which will reduce greenhouse gases through the introduction of energy efficiency and alternative energy technologies. NEDO projects are evaluated on the basis of energy savings, greenhouse gas reductions, and effect of technology diffusion. It has proposed projects in Russia, Poland, Indonesia, and Bangladesh to show the breadth of its global mission. It has been estimated that Japan will have the highest cost of compliance in an emissions trading market of more than $500 per ton of carbon. The numbers are even higher for a market without allowances and have been estimated to reach $1,075 per ton. These are very onerous costs to industry and should accelerate the moving to adoption of a domestic emissions trading scheme in Japan.

THE U.S. EMISSIONS TRADING EXPERIENCE Despite the fact that many countries continue to propose emissions trading schemes in the form of green certificates, the reality is that the United States is the only country that has successfully developed an emissions trading market that has worked well for the past seven years. As initially proposed by the Environmental Defense Fund (a U.S. environmental organization now called Environmental Defense) to the first Bush Administration for the trading of sulfur dioxide (SO2) credits, the emissions trading market has been successful beyond what its architects envisioned. Basically, the U.S. Environmental Protection Agency runs an emissions auction during March of each year that is supervised by the Chicago Board of Trade. Under Phase I which began on January 1, 1995, the 110 highest emitting utility plants were mandated to reduce their annual sulfur dioxide emissions by 3.5 million tons. This process was begun in 1995 for sulfur dioxide and extended to nitrous oxides (NOX) in 1999. The Over-the-Counter (OTC) forward markets trade these vintaged credits through the year 2030. Several OTC energy brokers are involved in brokering these credits including Evolution Markets, Natsource, Prebon, and Cantor Fitzgerald, and over one million trades per year occur. Thus, the market is liquid and has created emissions credits that are a fungible financial product. It has also saved $1 billion per year over command and control strategies. Under Phase II which began on

218

GREEN FINANCE

January 1, 2000, a more stringent standard calling for an additional annual reduction of 5 million tons of sulfur dioxide was required, and the program was expanded to another 700 utility plants throughout the U.S. Under the SO2 program, utilities are given one allowance for each metric ton of sulfur dioxide emitted. The utilities are given flexibility on how they meet the mandated targets, and can switch to fuels with lower sulfur content, install pollution control equipment, or buy allowances in order to comply with the law. In order to buy allowances, other utilities must reduce their emissions below their emissions limit. These emissions allowances are fully marketable once they are allocated through an EPA auction. The allowances therefore can be bought, sold, and banked. The allowances are allocated in phases. The later phases tighten the limits on previously impacted sources of pollution and are also imposed on smaller cleaner units. Compliance is assured through continuous emissions monitoring at plants and regular reports to the EPA. Fines are assessed if companies don’t comply with the law. The system has an allowance trading system. All transfers are recorded and posted on the Internet. Serial numbers allow the tracking of each allowance’s trading history, and an inventory for all accounts is provided. The most interesting phenomenon from this market-based solution to pollution has been that from 1995 through 1999 the market not only met its emissions reduction targets but was 30 percent under compliance. This approach has exceeded expectations by lowering emissions below the announced targets because some companies demonstrated unexpected behavior such as banking rather than selling emissions credits. Companies such as Minneapolis-based 3M Company did not sell their sulfur dioxide emissions credits as part of their corporate philosophy to be perceived as an environmentally benign company. Other companies followed this example of corporate environmental stewardship.

CREATION OF THE MARKETPLACE Because of the ability to establish exchanges quickly on the Internet, it is thought that this may be a desired outcome for emissions trading. Internetbased emissions trading would allow immediate disclosure for market players and has low costs of operation. The concept behind the allowances was to foster the implementation of demand side efficiencies or use of renewable energy. These concepts are tailored to the developing CO2 market development and the use of the Internet as the means to implement change.

Corporate Responses to Kyoto

219

The thought is that the creation of a marketplace for emissions trading will motivate firms with surplus emissions rights to supply them to the market. In effect, there are merits to move forward early despite the risk of uncertainty on future rules. It seems evident that industry-driven schemes will be grandfathered in the future as the rules are more clearly defined. Thus, industry can create its own domestic and international portfolio of emissions allowances or credits. The argument today is that to do it early will probably be less costly than in the future. Using GHG emissions allowances now is a form of insurance for industry participants. Moreover, emissions trading delivers significant environmental reductions as reduced compliance costs as well as promotes environmental technologies. There are several similar characteristics of emissions trading schemes, and in many countries the dual process of electric power industry liberalization. Emissions trading and electric power deregulation intersect since the power industry contributes to the greenhouse gas emissions. The impetus will be there to move the process forward.

CORPORATE RESPONSES TO KYOTO The Kyoto Protocol is unfortunately a market failure in its present form without the participation of the United States which emits 25 percent of greenhouse gases. Moreover, the present form of Kyoto under COP7 has significantly lowered the goals of Kyoto. Greenhouse gas emission reduction will take decades to achieve rather than the limited goal of 2012 envisioned under Kyoto. In effect, Kyoto is a very modest effort to contain emissions. It is only a first step. The need to create market liquidity is the primary challenge for CO2 emissions trading to succeed. With electric load growth and economic activity increasing each year, there is need to create incentives for new technologies to penetrate new markets due to liberalization. One obstacle to change has been the fossil fuel subsidies in many countries. These must end since they create the wrong economic incentives. These incentives must have the flexibility to develop market-based solutions but should not be overly onerous as not to work. Many private companies are moving forward under their own initiatives. They are, in effect, creating a global emissions portfolio that will develop as a result provided that energy companies can assume the risk. The

220

GREEN FINANCE

BP and Shell internal emissions trading systems are leading the way for energy companies to reduce greenhouse gas emissions. BP has about 150 of its business units operating in more than 100 countries involved in a cap and trade scheme to reduce its greenhouse gas emissions. It began the program in January 2000. Both CO2 and methane are traded in the BP system. The concept is to aggregate reductions from all business units. At the end of 2000, BP has traded 2.7 million tons of CO2 at an average price of $7.60. Shell has pledged to reduce its GHG emissions by 10 percent by 2002 compared to the 1990 baseline levels. Shell’s upstream oil, downstream refining and chemicals businesses are trading emissions. Estimates are that Shell’s carbon reductions range in value from $5 to $40 per ton. The program is reconciled internally on a yearly basis. Both Shell and BP plan to extend their programs externally as they develop expertise and further success. These companies and others should be encouraging companies to trade their emissions permits internally between countries as a means to accelerate technology transfer and reduce greenhouse gas emissions. In essence, we need to create global emission permit allocations, and essentially have a market-based solution for global pollution. They have the added benefit that it is cheaper to buy credits today as an insurance play.

CREATING THE GLOBAL CO 2 EMISSIONS PORTFOLIO The goal is a gradual reduction in emissions driven by measurable targets using market-based incentives. These can include outright purchase of emissions reductions. The aim is to encourage better technologies, better fuel choices, and better results and accelerated technology transfer. Multinational companies in North America, Europe, and Asia are developing emissions reducing schemes that can be transferred to their affiliates in developing countries. Any market needs trading liquidity in order to ensure fungibility. Presently, the CO2 emissions trading market has completed only 50 trades including one North American/Europe carbon trade and one European/ Australian trade. Other factors that influence trading are caps. The reality is that the greenhouse gas emissions market is in its infancy and trading caps can either be adopted by government or left open-ended for the markets to decide. There is competition to create global environmental exchanges. They

Green Finance: Project Finance as the Way Forward

221

need not be mutually exclusive as today’s Internet technology creates a borderless trading environment. In effect, we can have world greenhouse trade through the Internet. Today, exchanges getting into the act include the Sydney Futures Exchange, the International Petroleum Exchange, the Paris Bourse/UNIPEDE, and the Chicago Climate Exchange. Over-the-Counter brokers active in GHG emissions trade include Evolution Markets, Natsource, Prebon, and Cantor Fitzgerald.

GREEN FINANCE: PROJECT FINANCE AS THE WAY FORWARD But the key breakthrough for CO2 trading will be the use of the project finance mechanism to create “clean development mechanism” credits. In this way, a stream of emissions credits for 30 to 40 years (the life of the project) can be banked upfront. Investment and commercial banks can later create environment checklist for banks so that further streams of credits can be created. Finally, the creation of a global CO2 market will be traded on the Internet as the Internet will accelerate trading, is cross border, and can bring the most players to the global marketplace. Green Finance is thus born as the solution for global pollution and greenhouse gas mitigation strategies through the use of financial engineering at its best.

CHAPTER

15

Energy Convergence: What’s on the Horizon? Peter C. Fusaro President, Global Change Associates Inc.

his book has introduced the reader to a variety of emerging commodity market trading developments that are being driven by global changes in energy industry regulatory structures, privatization efforts, and competitive market forces. The good news is that as world energy markets are open to competitive and new risks, there are now more risk management mitigation techniques for managing those risks. These include more sophisticated options strategies and models, and more flexible (now web-based) risk management software solutions that are available or under development. Weather, telecommunications bandwidth, and emissions markets emergence in trading and risk management arbitrage will continue to be followed by even newer commodity markets such as LNG (liquefied natural gas) and web-based trading for all these commodities. As trading liquidity grows in the global market place, the Internet-based trading platforms will absorb much of this trading liquidity and in fact enhance its growth. Today, we are seeing the beginning of convergence to the multicommodity market founded on over more than decades of oil trading and risk management. The extension of energy commodity trading expertise for natural gas began in North American markets in 1990 and in Europe since the late 1990s. The beginning of physical and financial markets for electric power in North America, Europe, and Australia surfaced in the mid-1990s. An active emissions trading market began in 1995 in North America. And finally, the ultimate convergence of telecommunications

T

223

224

ENERGY CONVERGENCE: WHAT’S ON THE HORIZON?

bandwidth and electric power started in late 1999 with the advent of bandwidth trading. These markets are in various states of emergence and maturity. But the key fact is that they are growing, and extending the energy risk management platform. Moreover, Internet trading of these commodities, which began as proprietary electronic trading in 1993, is now moving forward globally, and is the Internet-based with its lower transaction costs and global access. The oil and gas industry has been gradually using energy risk management tools to manage its financial risks over the past two decades. The unprecedented price volatility in oil, gas, and electric power markets experienced since late 1998 are accelerating industry adoption of both financial instruments and Internet energy trading although the Enron debacle may slow the process down. While liquidity on the Internet today remains low, the movement to Internet trading will create its own liquidity and force the OTC paper markets for oil, gas, and power to migrate to the Web. Moreover, other industry trends of consolidation, market liberalization, and privatization are creating greater risk that must managed more proactively. The Internet will be the medium for that trade. Commoditization has been accelerating into many new markets through the innovative efforts of companies such as Enron, Morgan Stanley, Dynegy, Goldman Sachs, Aquila, El Paso, and others who act as market makers. The good news is that commodity markets need more players to develop fungible financial products and provide liquidity.

LNG TRADING LNG trading and hedging is following the development of a global market similar to the track that oil trading followed more than 20 years ago. Growing spot trading in LNG since 1999 and the development of global gas markets have brought forth the opportunity for LNG hedging. The development of the LNG spot market has been stimulated by the expiration of some 20-year supply contracts, new market entrants on the producing side, availability of LNG tankers, and LNG capacity creep (excess capacity over nameplate). The first significant hedging of LNG was for cargoes delivered into Taiwan by Mobil Oil in 1999. Since those initial market developments, El Paso, Enron, Sempra Energy Trading, CMS, and Aquila are setting up or already have structuring desks to hedge LNG cargoes. Basically, this involves using oil OTC contracts as surrogates for the

Electronic Trading: Set to Take Off

225

natural gas, and either deconstructing the crude cocktail on which East Asian LNG prices are based or using the NYMEX Henry Hub natural gas delivery point as a price marker. In effect, the optionality of the contracts is extracted. The new suppliers of LNG have made the market much more competitive. While East Asia still takes 75 percent of the LNG supply, many new countries such as Spain, Turkey, France, Belgium, and Germany are adding more diversity on the consuming side. The same is true on the supply side with more suppliers including Qatar, Egypt, and Trinidad. Moreover, the reemergence of the United States as an LNG consumer, with the reactivation and possible expansion of the Lake Charles, Louisiana facility particularly being important, have brought forth more cargo movements and liquidity to the market. Particular interest has been seen of a North American West Coast receiving terminal that now brings the United States into both Atlantic and Pacific basin LNG trade. The hedging of LNG cargoes is giving suppliers and end users flexibility and greater efficiency in a very capital intensive industry. It is also another application of innovative financial engineering to commoditize a new market.

ELECTRONIC TRADING: SET TO TAKE OFF While the number of electronic energy trading platforms continues to ramp up, some of these platforms are in dire financial straits and have not built trading liquidity. Ultimately, most of these trading platforms will fail, but what appears inevitable is that several will dominate the Internet trading space while others will perform well in niche market functions. This pattern is true of major changes in any new market. Something has radically changed in energy trading markets, and that was the age of electronic trading coupled with the OTC market flexibility, which usurped exchange-traded futures contracts. The futures exchanges have been slow to react to this Internet phenomenon. Other critical changes have occurred over the past 20 years including that of price assessment panels and index trading, which had failed in the late 1980s but succeeded in the 1990s. Electronic index construction such as Dow Jones electricity or bandwidth indexes coupled with screen trading has now catalyzed the global energy industry into the Internet age. The proliferation of electronic brokering and trading platforms that are emerg-

226

ENERGY CONVERGENCE: WHAT’S ON THE HORIZON?

ing today will continue to change the face of energy trading. While energy markets have traditionally followed financial markets in terms of application of trading methodology and instruments, the commoditization of the energy complex is now in full swing and the pace is quickening. These platforms are all focused on the paper trading of oil, gas, and power, and will extend into coal, bandwidth, and emissions trading. They are dominated by the OTC markets. Energy trading has changed forever.

The Intercontinental Exchange Today, one clear electronic trading winner has emerged and that is the Intercontinental Exchange also known as ICE, backed by the major energy companies, gas and electric utilities and investment banks. ICE launched all its energy verticals by October 2000. These included crude oil, petroleum products, natural gas, and electricity. The major players behind this electronic trade are very active in oil, gas, and power markets on both the physical and financial sides. The backers include American Electric Power, BP, Shell, Duke, Aquila Energy, Morgan Stanley, Goldman Sachs, El Paso Energy, Reliant Energy, Southern Energy, Deutsche Bank, Totalfina Elf, and SG Investment Banking. The only piece missing for total oil, gas, and power dominance by ICE are the participation of the major European gas and electric utilities, such as Electricite de France, Germany’s RWE, Italy’s ENEL, and Spain’s Endesa. Market liquidity is growing rapidly even in niche markets like jet fuel hedging. Its acquisition of the International Petroleum Exchange in London in June 2001 further solidified its hold on electronic trading dominance as it will commingle both energy futures and OTC products as well as shift the IPE platform to total electronic trading over time. ICE will extend its energy trading platform to the emerging commodity markets of coal, weather, emissions, and bandwidth in the future.

NYMEX NYMEX’s attempt at electronic trading with an Internet-based platform seems destined for failure due to its announced plans to launch an OTC energy derivatives market for oil and gas where it has no expertise. While NYMEX has had an after-hours trading system for its energy futures contracts for the past eight years called ACCESS, the exchange has now

Electronic Trading: Set to Take Off

227

launched a web-enabled ACCESS in September 2001 (which has become more important due to the World Trade Center disaster and the need to trade energy futures electronically). This now seems to be their strategy. The reason NYMEX had not extended electronic trading to floor trading hours (except due to events since the September 11 emergency) is that the exchange members do not want to cannibalize their existing trading floor business and its highly liquid and successful oil and natural gas futures contracts, which continue to grow each year. Thus, they have decided to launch their ill-conceived OTC trading platform, which puts it in direct competition with ICE and other OTC Internet trading platforms. A more sanguine strategy would have been for NYMEX to become the global Internet oil and gas futures exchange, and for NYMEX to clear for all Internet energy platforms with its very successful clearinghouse function. Ultimately, e-NYMEX will fail but not for a lack of resources since eNYMEX will not be allowed to penetrate NYMEX’s core oil and gas futures floor business and will suffer liquidity problems accordingly.

EnronOnline Led the Way EnronOnline was not a true exchange, but it was the most liquid Internet energy platform, trading almost $900 billion. It is now dead. Enron acted as market maker to all transactions on both the buy and sell side. It made markets in oil, gas, and power but extended its platform into bandwidth, emissions credits, pulp and paper, steel, petrochemicals, plastics, commercial credit, weather, and metals. EnronOnline facilitated the trading of more than 1,800 different commodities. EnronOnline went live on November 29, 1999 and wildly exceeded all expectations. These offerings moved Enron’s business model into the commoditization of all markets, and also reduced its marketing costs as it shifted more trading activity from the trading floor to the Internet. EnronOnline was active throughout both Asia and Europe as well as North America. The premise behind EnronOnline was that commodity trading would appeal to a much broader market if a more accessible trading mechanism could be designed. Among Enron’s boldest ideas was its use of a single, powerful software-based trading platform, which enabled it to insert itself as a principal buyer or seller in hundreds of different markets. Registered users could trade a range of commodities from coal to emissions. One thing that differentiated EnronOnline from most other Internet

228

ENERGY CONVERGENCE: WHAT’S ON THE HORIZON?

marketplaces was that it was free. There were no transaction fees or catalog, inventory, or subscription costs for users. Even more unusual was Enron’s role as a principal in each transaction. Enron was either the buyer or seller in every trade, unlike market makers that simply act as matchmakers between buyers and sellers. As either the seller or buyer in all of its trades, Enron guaranteed delivery or payment on all trades that were executed on EnronOnline. The question became how profitable was it. Unlike other B2B exchanges that act as impartial intermediaries for buyers and sellers, Enron’s marketplaces created a benchmark product at a set price. The company acted as counterparty in each trade, guaranteeing that commodities purchased on its site were delivered at the agreed-upon price and terms. Enron’s profit lay in the spread between what it paid for a commodity and what it sold. It seems likely that ICE will capture the lion’s share of EnronOnline business.

DynegyDirect Enron’s highly successful business model has been emulated by Dynegy’s dynegydirect which is offering more than 150 products in North American gas, power, and natural gas liquids markets. Dynegy wants to aggregate multiple transactions of its site and then execute them on TradeSpark. Dynegyconnect has been launched as a means to offer increased connectivity to expand networks and add Internet access to their service portfolios.

Altra The oldest of all electronic trading platforms with the launch of Chalkboard for oil trading more than nine years ago, Altra Energy Technologies’ acquisition mode of TransEnergy Management, Quicktrade, and Energy Imperium during the past few years added significant overhead and systems integration problems. While still a very liquid gas trading system, its effort to create more verticals has been suspect. The company seems likely to be absorbed by other B2B platforms.

TradeSpark TradeSpark has a different mix of partners than ICE. Williams, Dynegy, Coral Energy, Dominion, Koch, TXU, Entergy, and broker Cantor Fitzger-

Electronic Trading: Set to Take Off

229

ald’s eSpeed may have a chance at some liquidity. These well-capitalized partners may offer another alternative to the dominance of the Intercontinental Exchange.

European Markets Presently, European Internet platforms are just as poorly defined as those in the United States. The need for separate power exchanges for almost each country in the EU does not demonstrate a willingness to move away from nationalism. The Internet is borderless. The many platforms are highly parochial. There are the Amsterdam Power Exchange, the Leipzig Power Exchange, the European Energy Exchange (which is actually German), the Polish Power Exchange, the Austrian Power Exchange, and up to four electricity verticals in the United Kingdom (The Leipzig Power Exchange and European Energy Exchange have merged in early 2002.) The only successful platform has been Nord Pool in Scandinavia which has worked successfully since 1993 due to the nature of the hydropower/nuclear power markets in Norway, Sweden, and Finland. It is a small, successful niche market.

Asian Markets The Asian Internet energy markets are evolving more slowly. The Singapore Exchange will now only launch electronic energy contracts in the future as its oil futures contracts for Brent mutual offset and fuel oil have fallen by the wayside. While Platts Global 190 trades oil electronically, it seems to be by default and by a lack of trade interest. There is an interesting play in open spec naphtha (e-OSN.com) in the East Asian markets that will probably work since it has had an active and liquid forward OTC market for more than a decade. E-OSN.com acts as broker to all market participants; however, the site is a niche market play. It does however, demonstrate the potential viability of niche market segments for energy trading on the Web. The Tokyo Commodity Exchange energy contracts for gasoline is gaining some liquidity but the contract volume in kiloliters is quite small. ICE may become the dominant player in Asia as well as Europe and North America. One of the obstacles in Asian market development has been that Asian Internet growth lags worldwide usage but should accelerate over the next several years. China and India will undergo rapid growth as the Internet currently is under 1 percent of their populations. India has the world’s

230

ENERGY CONVERGENCE: WHAT’S ON THE HORIZON?

largest electronic exchange (for equity trading). Both countries may embrace Internet energy trading in a big way by leapfrogging technology.

What’s on the Horizon? The death of the Internet has been greatly exaggerated just as its entrance was. The key barriers today are not technological but changing human behavior to use the net and break away from human relationships which are still a strong part of the energy trading complex. No one doubts there is a need to attract more liquidity, create methods for more price discovery, create adequate controls and systems to handle credit risk, clearing, and settlement. The bogus issues of secure encryption are not an active barrier to entry. Interbank transfers of funds have high degrees of security and have been going on for more than two decades. B2B exchanges will not only survive but thrive in the coming two years as the need of digital markets to centralize information, create communities, and trade energy accelerate. The fragmented markets of today will lead to the consolidated trading markets of tomorrow with several dominant global players, and regional and niche players relegated to the sidelines. Parsing and weaving the best of other platforms for other commodities related to energy such as emissions and bandwidth will create a multicommodity warehouse that will have clearing and settlement taken care of by the major banks that are already part of a global consortium. The platforms will have better functionality as the technologies improve. Buying customers and technology, the front, middle, and back office will be web-enabled so that the exchanges can provide seamless trading opportunities on a 24/7, borderless basis. Pricing will become more transparent and liquidity will grow exponentially. The energy industry is the nexus for the application of the Internet for not only energy trading, but procurement and etailing. A global business packed with information would seem to be the place to catalyze Internet applications.1

EMERGING MARKETS Bandwidth, emissions, LNG, and electronic energy trading are coming on strong in global financial markets because these commodities are global and driven by global market drivers. These markets are becoming estab-

Emerging Markets

231

lished on the Internet where time compression increases and the maturation process decreases. They are real world applications of financial engineering techniques. One key change will be the use of energy project finance. We will see some of this application in the use of creating clean development mechanism credits in the emissions markers and that will bring liquidity to the Greenhouse Gas emissions market. Such green finance will be embraced by investment banks and traders to hedge environmental risk and securitize energy projects. The future holds endless applications of financial engineering, commoditization, and risk management for the global energy market, and energy risk management has now become a fiduciary responsibility of energy companies. Energy hedging is still in its infancy, and there will be greater growth over the next decade even in the mature oil trading markets.

glossary of energy risk management terms

American option An option that can be exercised by the buyer at any time during its life. arbitrage The simultaneous purchase and sale of similar or identical commodities in two markets. Asian option An option that can be exercised only at expiration, based on the difference between the strike price and the average of daily spot prices over the life of the option. Also called average price options. backwardation Market situation in which prices are progressively lower in distant delivery months. Opposite of contango. basis The differential that exists at any time between the futures or forward price for a given commodity and the comparable cash or spot price for the commodity. Basis can reflect different time periods, product qualities, or locations. basis risk The uncertainty as to whether the cash–futures spread will widen to narrow between the time a hedge position is implemented and liquidated. bid/ask A measure of market liquidity. The bid is the price level at which buyers are willing to buy and the ask is the price level at which sellers are willing to sell. book The total of all physical, futures, and OTC derivatives positions held by a trader or company (includes documentation). broker In futures, the person who executes the buy and sell orders of a customer in return for a commission or fee. In the OTC markets, the person who introduces counterparties and arranges the transaction charging a fee for this service. Brokers never take a position in the market. calendar spread An option position created by selling one call and buying another with a longer expiration at the same strike price. Also called time spread.

233

234

GLOSSARY OF ENERGY RISK MANAGEMENT TERMS

call option An option that gives the buyer the right (but not the obligation) to enter into a long futures position at a predetermined strike price, and obligates the seller to enter into a short futures position at that price, should the option be exercised. See put option. cap A contract between a buyer and seller in which the buyer is assured he will have to pay more than a maximum price. See Floor. carrying charge The total of storing a physical commodity including storage, insurance, interest, and opportunity costs. CFD A contract for difference, which is basically a price swap and is usually applied to the short term Brent crude oil swaps market. CIF Cost, insurance, and freight. clearing member Members of an exchange who accept responsibility for all trades cleared through them and share secondary responsibility for the liquidity of the exchange’s clearing operation. Clearing members must meet minimum capital requirements. clearinghouse An exchange-associated body charged with insuring (guaranteeing) the financial integrity of each trade. Orders are cleared by the clearinghouse acting as a buyer to all sellers and seller to all buyers. The clearinghouse stands behind all trades made on the exchange. close The period at the end of a trading session; also called the settlement price of that commodity. collar Options strategy designed to minimize upfront costs of a cap or floor though the sale of a cap or floor. contango Market situation in which prices are progressively higher in succeeding months than in the nearest delivery month. Opposite of backwardation. counterparty The person or institution standing on the opposite side of a transaction. credit risk The risk of default by either counterparty in a transaction. crack spread An intermarket spread where futures are bought and sold to mimic the refining of crude oil into petroleum products. delivery month The month in a given contract for delivery of the physical commodity. derivatives Financial instrument derived from the underlying commodity including forward, futures, swaps, and options. European option An option that can be exercised by the buyer only at expiration.

Glossary of Energy Risk Management Terms

235

exotics A term used to refer to more structured over-the-counter instruments, such as swaps with a range of maturities, volumes, or fixed prices; transactions that incorporate different commodities or complex combinations of options. floor A contract between a buyer and a seller in which the seller is assured he will receive at least a minimum price. See cap. FOB (free on board) A transaction in which the seller provides a commodity at an agreed upon price, at a specific loading point within a specific period of time. The buyer must arrange for transportation and insurance for delivery. forward Standardized contract for the purchase or sale of a commodity, which is traded for future delivery not under the provisions of an established exchange. fungibility Characteristic of products or instruments that can be commingled for trading, shipment, storage, or consumption. Only fungible goods can be traded as commodities. futures Standardized contract for the purchase or sale of a commodity for future delivery under the provisions of exchange regulations. hedging In the futures or OTC markets, a simultaneous initiation of equal and opposite positions in the cash and futures markets. Hedging is employed as a form of financial protection against adverse price moves in the cash market. Opposite of speculation. in-the-money An option that has intrinsic value at expiration. Opposite of out-of-the-money. initial margin Margin posted when a futures position is initiated. intrinsic value The value of an option if it were to expire immediately. Cannot be less than zero. ISDA International Swaps Dealers Association. A trade association, primarily of banks, that promotes the use of all derivative instruments. ISDA has its own master swaps agreement that is heavily used by the energy trade. liquidity A characteristic of a market where there is a high level of trading activity. local A commodity or options principal and exchange member who buys and sells for his own account on the floor of the exchange. long position In the futures market, the position of a contract buyer whose purchase obligates him to accept delivery unless he liquidates

236

GLOSSARY OF ENERGY RISK MANAGEMENT TERMS

his contract with an offsetting sale. In the forward market, a long position obligates a buyer to accept delivery unless a book-out agreement is subsequently made. margin Funds or good faith deposits posted during the trading life of a futures contract to guarantee fulfillment of contract obligations. mark-to-market Daily adjustment of open positions to reflect profits and losses resulting from price movements that occurred during the last trading session. market maker A dealer who consistently quotes bid and offer prices for a commodity. Can be an energy company. moving average Technical analysis term, which clearly signals any change in the trendline. net position The difference between an entity’s open long contracts and open short positions in any one commodity. notional The underlying principal of either an exchange traded or OTC transaction. offer A motion to sell a futures, forward, physical, or options contract at a specific time. offset A transaction that liquidates or closes out an open contract position. open The period at the beginning of the trading session on a commodities exchange. open interest The number of futures contracts on an exchange that remain to be settled. option The instrument that gives the holder the right to buy or sell the underlying commodity at a given price or at a specific date. out-of-the-money An option that has no intrinsic value at expiration. See in-the-money. OTC (over-the-counter) Purchase and sale of financial instruments not conducted on an organized exchange. paper barrels Term used to designate nonphysical oil markets including futures, forward, swaps, and options. position taking The action of commercial participants who use the futures market as an alternative cash market rather than as a hedging vehicle. premium The price paid by the option buyer to the option seller. price signaling Advanced publication of prices prior to their effective date for the purpose of encouraging competitors to make similar price changes.

Glossary of Energy Risk Management Terms

237

put option An option that gives the buyers the right (but not the obligation) to enter into a short futures position at a predetermined strike price and obligates the seller to assume along a futures position, should the option be exercised. See call option. short position In the futures market, the position of a contract seller whose sale obligates him to deliver the commodity unless he liquidates his contract by an offsetting purchase. See long position. speculation The opposite of hedging in which the speculator holds no offsetting cash market position and deliberately incurs price risk in order to reap potential rewards. spot A one-time open market transaction, where a commodity is purchased on the spot at a current market price. spot month The futures contract held closest to maturity. spread The simultaneous purchase of one futures or forward contract and the sale of a different futures or forward contract. Also refers to futures/forward contract purchase in one market and a simultaneous sale of the same commodity in some other market. straddle The purchase or sale of both a put or call having the same strike and expiration date. strike price The predetermined price level at which the exercise of an option takes place. swap Customized contractual agreement between two parties to exchange interest payments, typically a fixed rate payment for floating rate payment. No physical commodity exchange takes place. swaption An option of a swap. technical analysis Examination of patterns of futures price changes, rates of change, and changes in trading volume and open interest, often by charting in order to predict and profit from such trends. time value Part of the option premium that reflects the excess over intrinsic value. trend Price activity in markets in a particular direction; characterized by higher highs and higher lows. variation margin Margin paid or collected in order to maintain a minimum level based on daily fluctuations in contract value. volume The number of transaction occurring on an exchange during a specified period of time.

endnotes

CHAPTER 2 1. “The Federal Reserve Board Remarks,” by Chairman Alan Greenspan before the Futures Industry Association, Boca Raton, Florida, March 19, 1999. 2. Bank of International Settlements. Central Bank Survey of Foreign Exchange and Derivatives Market Activity, 1998. Basel, 1999.

CHAPTER 3 1. Snowfall-linked coupons have been attached to bonds issued by ski resorts, for example. See John C. Hull, Options, Futures and Other Derivatives (Prentice-Hall/Simon & Schuster, 2000), page 1. 2. Beware the diversity of daily averages: These can range from [absolute maximum + absolute minimum]/2 to the average of the highest and lowest readings recorded within a day. These may be recorded hourly or on a more erratic basis. Indeed, the definition of a day itself can vary significantly from reporting station to reporting station. 3. Note that this result may be based on official data, which may take the reporting agency weeks or months to produce. To ameliorate this problem, a rough value may be used to determine the payment at maturity, with some minor adjustment possible when the official data are finally published. 4. The Over-the-Counter (OTC) market is where counterparties to a trade assume direct payment liability (and so credit risk) with each other. For the purpose of this text, no distinction is made between an OTC trade that was struck directly between the two counterparties or one that has been intermediated, either by a voice broker or online. 5. If the data are taken from a nearby world meteorological organization from an approved station, rather than the client’s own site, that is. Naturally, the hedger’s own measurements can be used, but this will

239

240

ENDNOTES

increase pricing difficulties and cost to the hedger, and will add some obvious agency problems to the mix. 6. With the possible exception of cyclical forecasting, such as El Niño/La Niña. 7. Goldman Sachs International, March 2000; estimates for activity up to end 1999: Weather $4,550 million versus Catastrophe $3,173 million.

CHAPTER 5 1. This section is a modification of Brooks (1998). For more information on building C++ applications related to various approaches to valuation, see Brooks (2000a). 2. For more details on the LSC model, see Brooks (2000b). 3. We could easily define ∆fj.t as the percentage change in the m-maturity forward contract price. Therefore, depending on empirical analysis, ∆f j,t ≡ f j,t +1 − f j,t or ∆f j,t ≡

f j,t +1 − f j,t f j ,t

 f j,t +1  or ∆f j,t ≡ ln .  f j ,t 

However, fj,t is not the investment; hence the second and third definitions are not the rate of return for forward contracts.

CHAPTER 7 1. Mostly as a result of Soviet perestrojka and glasnost; cf. Kolodko (1999). 2. Cf. David (1985), Erdmann (1993), Fulda et al. (1997). 3. Of course such a view seems to be too simplistic, as the experiences in China and Vietnam show; cf. Kolodko (1999). 4. Cf. Arthur (1989), Reichel (1998) concerning the historical meaning of lock-ins. 5. Cf. Balmann/Reichel (2001), Arthur (1983), Witt (1997). 6. Cf. Arthur (1983), Lehmann-Waffenschmidt/Reichel (2000), Reichel (1998). 7. That is, development path. 8. There are some widely discussed and also doubted examples of such lock-ins, for example, the dominance of the QWERTY-claviature, the development of nuclear technologies, the dominance of the VHS-

Endnotes

241

System compared to Betamax, and so on; Arthur (1989), Cowan (1990), David (1985), Friedrich (2000). 9. Cf. Riesner (1999).

CHAPTER 11 1. These are transactions based on either time period or delivery point differentials. A time period differential is a transaction in which power is delivered in the present in exchange for power to be delivered at some time in the future. A delivery point differential is a transaction in which power is delivered at a certain point in exchange for power to be delivered at another point. 2. In our purist terminology, we use the word “hedging” for a combination that hedges both sides of a physical requirement. We call a onesided transaction a “bet,” not a hedge. 3. We are cheating a little in this example. The distributions associated with linear combinations of statistical variables are easier to manipulate than more complex relationships; but then we told you this was a simplified example. Even in the case of complex interactions, the methodology is the same. It is just that the resulting variables follow more complex statistical distributions.

CHAPTER 12 1. Howard L. Margulis is a Partner in the international law firm of Squire, Sanders & Dempsey L.L.P., and leads the firm’s New York energy transactions and project finance practice. He is a graduate of Northwestern University (1984) and Illinois Institute of Technology Chicago-Kent College of Law (1987, summa cum laude). He regularly counsels energy firms and financial institutions in large-scale energy mergers and acquisitions and structured finance transactions, international energy project financings, and energy regulatory and advanced technology matters. He can be reached at hmargulis @ssd.com. The views expressed herein are solely those of the author, and are not the stated or official opinions or positions of Squire, Sanders & Dempsey L.L.P. The author wishes to express his thanks to Peter C. Fusaro for his expert advice and guidance in preparing this material.

242

ENDNOTES

2. Apparently, this impact of Rule 133 has not prevented Barrett Resources from consideration as a suitable merger candidate, as witnessed by the recent efforts by affiliates of Royal Dutch Shell Petroleum to acquire the company via an unsolicited acquisition overture. 3. October 2001. 4. One newly formulated provision dealing with terrorism’s impacts on the economic climate for mergers reads as follows: For purposes of this Agreement, the term “MATERIAL ADVERSE EFFECT” shall mean any change or event or effect . . . except to the extent caused by a material worsening of current conditions caused by acts of terrorism or war (whether or not declared) occurring after the date of this Agreement which materially impair the Company’s ability to conduct its operations except on a temporary basis. . . . Other suggested provisions have focused, instead, on whether or not financial markets are open and available, giving either party a right to withdraw if, for example, major markets are closed for ten (10) days or more.

CHAPTER 13 1. Electricity Daily 6/4/99. 2. “National Energy Policy” (May 2001) at pages 1–4. 3. “Power Generation Opportunities in a Restructured Environment.” Reported in Deregulation Watch, July 15, 1999, page 7. 4. Richard H. Cowart. Efficient Reliability, June 2001, page 10.

CHAPTER 15 1. Part of the discussion of electronic energy trading platforms was written in an article by Peter Fusaro (the author) in the March 26, 2001 issue of the Oil & Gas Journal. This has been updated and edited substantially for this book.

index

ACCESS, 226–227 Accountants, role of, 180 Adriatic Sea, 134 AES, 158 Aframax, 145 Agents, in shipping industry, 144 Algeria, as supplier, 134 Alternative Risk Transfer (ART), 31 Alternative stochastic processes LSC model, energy option valuation, 77–79 Altra, 228 Amerex, 36 American Electric Power, 226 Amount, in weather derivatives, 22–23 Amsterdam Power Exchange, 229 Ancillary services, 181 Application design, 183 Aquila Energy, 36, 158, 226 Arbitrage: defined, 145 freight industry, 150–151 geographical, 42–45 transportation costs and, 143 Arbitrageurs, role of, 55 ARIMA, 186 Asian markets, electronic trading, 229–230 Asset displacement, wholesale electricity markets: generation economics, 168–171 transmission constraints, 172–173 AT&T, 34 Attachment point, 18 Austrian Power Exchange, 229 Average annual demand, electricity market, 160 Average option, 139

Bachelier, Louis, 51 Baltic, 145 Bandwidth Desk (Scudder Publishing Group), 35 Bandwidth market liquidity: Enron/Global Crossing deal, 33 forward prices analysis, 38–40 geographical arbitrage, 42–45 market data, 36–37 market development, 34–36 statistics and metrics, 37–38 time series, percentage price decline, 40–42, 48 tracking, analytical model, 45–48 Bandwidth Market Report (BMR) (McGraw-Hill), 35, 37 Bandwidth Trade Organization (BTO), 35, 48 Barings Bank, 192 Barrett Resources Corporation, 194 Bartering, Ukrainian energy sector, 121 Basket option, 139 Belarus, 104 Belief system, significance of, 85 Benchmarks/benchmarking, 7, 10, 162, 208 Beta, 452 Bid/ask spread, bandwidth market liquidity: geographical arbitrage, 38–42 tracking, analytical model, 45–49 Bid/offer, bandwidth market, 36–37 Binomial tree, 61 Black-Scholes model, 139 Blugas, 137 Body risk, 30 Bottlenecks, wholesale electricity markets, 172 BP, 220, 226

243

244 British Gas, 134 Brokers: in bandwidth market, 35–36, 48 electronic trading, 225–226 emissions trading, 217, 221 in shipping industry, 144–145, 147 Brooks, Dr. Robert, 2 Bundling of services, 138, 184–186 Bunker swaps, 148 Bush Administration: Environmental Defense Fund, 217 National Energy Policy report, 204–205 Business to business (B2B) exchanges: in electronic trading, 228, 230 maritime industry, 143–145, 149–150 Buyout pricing, 30 California: energy crisis in, 95 transmission congestion in, 172 California Power Exchange, 175 Call options: characteristics of, 11–12 energy option valuation approaches, 54–55, 61–63, 67–70 weather derivatives, 24 Calpine, 158 Cantor Fitzgerald, 217, 221, 228–229 Cap, in weather derivatives, 23. See also Call options Capacity development: Poland, 112–118 Ukraine, 119–122 Capacity factor, defined, 169–170 Capacity utilization, electricity market, 162–163 Capesize, 148 Capital asset pricing model (CAPM), 52 Capital investment, 7 Capital markets, overlap of, 31 Caribbs–USGC/USEC, 145 Cash flow adjusted approach (CFAA), energy option valuation: characteristics of, 52–53, 57–63, 66 discount factor adjusted approach (DFAA) compared with, 65–67 Cash flow hedges, 93, 95, 193 Cash flow table, 55–56

INDEX

Central and Eastern Europe (CEE): convergence process, 101–102, 108–109, 123–126 divergence process, 101–102, 123–126 energy and economic tendencies, 102–105 international cooperation, 107–108 path dependencies, 105–107 Poland, 110–118, 125 Ukraine, 118–127 Central European Power Index, 7 CENTREL/UCTE, 122–123, 125 CH4 (methane) emissions, 216 Charterers, role in shipping industry, 145, 147 Chicago Climate Exchange, 221 Chlorofluorocarbons, emissions trading, 213–214 Clean development mechanism (CDM), 216 CO2 emissions: challenges of, 219–220 global portfolio, 220–221 Kyoto Protocol, 214–216 market development, 218–219 COB Index, see Dow Jones California/Oregon border electricity price index Collar, as weather derivative, 24–25 Combined cycle gas turbine (CCGT) technology, 132, 169–171 Combustion turbines (Cts), 169 Commercial electricity usage, 161 Commodity, defined, 145 Commodity market generation economics, 168–171 Common stock, valuation methods, 64–65 Competition, generally: electricity market, 156, 163 energy risk and, 211 market indexes and, 7–8 Congestion, wholesale electricity markets, 172–173 Conservation, government funding for, 205, 217–218 Consolidated accounting, in mergers, 197

Index

Consultant, role of, 210 Contract electricity markets, 160 Contract terms, 207–208 Convergence process, 101–102 Convergent systems: current world, 179–181 integration, 181–188 overview, 178–179 Convertible debt instruments, 193 Cooling degree days (CDDs), 23, 26 Coral Energy, 228 Correlation, in weather derivatives, 22 Crametz, J.P., 2 Cumulative distribution function (CDF), 75–77 Currency exposure, hedges, 93 Currency swaps, 148 Czech Republic, energy and economic tendencies in, 103–105 Data architecture, 184, 189 Data processing, in weather derivatives, 22 Data storage, 184 Data stream, 188 Day ahead market, 175–176 DB2, 184 Default penalty payments, 207 Degree day, significance of, 27. See also Cooling degree days (CDDs); Heating degree days (HDDs) Delivery point, defined, 186 Demand development: Poland, 111–112, 125 Ukraine, 119–127 Demand in a given hour, electricity market, 160 Demand management, 208 Demand risk, electricity market, 208–210 Demand/supply, in convergence process, 109. See also Supply and demand Den Norsk Bank, 148 Deregulation: electricity markets, 156–157 electricity prices, 202–205 energy risk and, 205–211 historical perspective, 202 impact of, generally, 7–8, 158

245 outcomes of, 202 retail delivery, impact on, 181 Derivative securities, development of, 51. See also Weather derivatives Deutsche Bank, 226 Digitals, as weather derivatives, 25 Discount factor adjusted approach (DFAA), energy option valuation, 63–67 Divergence process, 101–102 Dividend discount model, 64 DJ Bandwidth Intelligence Alert, 35, 37 Dominion, 228 Douvlis, Kelly, 3 Dow Jones California/Oregon border electricity price index (COB Index), 10–11 Dow Jones Newswire, 35, 37–38 Dry bulk derivatives market, 147–148 DS3 capacity, 42–44 Due diligence, 191 Duke Energy, 36, 226 Dynegy, 36 DynegyDirect, 228 Earnings, significance of, 194 E-commerce, 138 Egypt, as supplier, 134 El Paso Energy, 36, 226 Electric generation finance, market risk in: demand, composition and growth of, 154, 159–162 electricity as unique commodity, 154–155 generation economics and asset displacement, 168–172 market finance, 155–159 market structure and risk, 173–176 new capacity, composition and growth of, 154, 162–166 supply and demand equilibrium, 166–168 transmission constraints and asset displacement, 172–173 U.S. generation capacity, 164 Electricite de France, 226

246 Electricity: generation finance, see Electric generation finance, market risk in markets, generally, 7–8 price risk, 141 prices, see Electricity prices sector turmoil, in Italy, 131–133 trade, Central and Eastern Europe, 126–127 Electricity-for-gas swap, 182 Electricity prices: deregulated markets, generally, 202–203 improvement strategies, 204–205 influential factors, 203–204 Electronic trading: Altra, 228 Asian markets, 229 DynegyDirect, 228 EnronOnline, 227–228 European markets, 229 future directions for, generally, 225–226, 230 Intercontinental Exchange (ICE), 226 NYMEX, 226–227 platforms, 34–36, 48 TradeSpark, 228–229 Elettrogen, 132 Embedded derivatives, 99 Emerging industries, market indexes in, 5 Emerging markets, 230–231 Emissions trading, see Kyoto Protocol (1997) development of, 216–217 Japan, 217 marketplace development, 218–219 United States, 217–218 Endesa (Spain), 132, 226 End-of-day prices, 6 ENEL (Italy), 132–133, 135, 226 Energia, 135 Energorynok, 119, 121 Energy audits, 181 Energy companies, bandwidth market, 36 Energy conservation, 205, 208–209. See also Global environmentalism

INDEX

Energy Derivatives: Trading Emerging Markets (Fusaro), 1 Energy Imperium, 228 Energy option valuation framework, 67–70 Energy risk: demand risk, 208–210 industrial and commercial customers, 210–211 management of, 207–210 supply risk, 207–208 types of, 205–206 Energy Risk Management (Fusaro), 1 Energy service companies (ESCOs): business process, 178–179 energy risk and, 211 functional diagram, 182 Energy service providers, defined, 158 Energy software technologies, 209 Eni, 133–136 Enipower, 133 Enron, 33, 36, 158, 182, 224, 227–228 EnronOnline, 227–228 Entergy, 228 Environmental Defense Fund, 217 Environmental factors, price volatility and, 146 Eon, 126 E-procurement systems, 138 Equities markets, 5 Equivalent-martingale valuation, 52, 57, 60–61 EU Directives, 104 Eurogen, 132 European Bank for Reconstruction and Development, 107 European Energy Exchange, 229 European markets, electronic trading, 229 European Union (EU), 101–102, 125–127, 131–133 Europe-Asia market, bandwidth market liquidity, 43–45, 48 Eustache, Dr. Antoine, 1–2, 36 Evolution Markets, 217, 221 eXact (Excelergy), 188 Excess allocation, 18–20 Exelon, 158 Exercise price, 11

Index

Exotic option, 139 Extensible Markup Language (XML), 187–189 Facultative reinsurance, 17–18 Fair value, measurement of, 91–92, 95, 139 Fair value hedges, 193 FAS 52, 91 FAS 133 (“Accounting for Derivative Instruments and Hedging Activities”): components of, 192–194 derivative hedges, generally, 29, 89–95, 99–100 embedded derivatives, 99 inventory hedge, 96–97 production hedge, 97–100 Federal Energy Regulatory Commission (FERC), 201 FERC 888, 201 Finance staff, role of, 180 Financial Accounting Standards (FAS), see FAS 52; FAS 133 Financial Accounting Standards Board (FASB), 90–92, 94, 98–99, 191–193 Financial demand risk, 206 Financial Engineering, 104 Financial statements, reporting requirements, 29, 193, 198 Financial supply risk, 206 Fixed price swap, 12–13 Fixed rate, defined, 23 Floating rate, defined, 23 Floor, defined, 24 Florida, wholesale electricity market, 174 Foreign currency exposure, hedges of, 193 Foreign exchange market, 90 FORTRAN, 189 Forward contracts: bandwidth market, 39–40 energy options valuation, 54–55 Forward freight agreements (FFAs), 147–148 Forward price(s): analysis, 38–40 defined, 71

247 relative, cash flow adjusted valuation, 60 Freight futures, 149 Freight trading: barriers to success, 150–151 development of, generally, 143–144 maritime industry, 144–146 price volatility, 146–147 risk management, 147–150 Fuel cells, 209 Fundamentals, significance of, 6 Fusaro, Peter, 3–4 Future directions: electronic trading, 225–230 liquefied natural gas (LNG), 224–225 Futures, generally: supply risk and, 207 as weather derivatives, 25 Futures contract, defined, 145 Gains/losses, derivatives, 93 Gaming laws, 29 Gas distribution, Italy: local distribution companies (LDCs), 132, 135–137 multi-utility companies, overview, 137–138 Gas IT, 137 Gas liberalization, Italy, 134–136 Gas production, historical, 164–165 Gas reserves, Italian, 134 Gaussian distribution, 20 Gencos, 132–133 Generation companies, electricity market, 158 Generation market risk, 154 Generators, market power of, 203–204 Geographical arbitrage, bandwidth market: Europe-Asia and Intra-Asia, 43–45, 48 Intra-Asia market, 43–45, 48–49 Intra-European market, 43–44, 48 United States–Europe, 42–43, 48 Global Crossing, 33 Global environmentalism, 213 God Only Knows (GOK), 178 Goldman Sachs, 226

248 Gowrinathan, Shiva, 3 Green finance: benefit of, 221 emissions trading markets, 216–217 exchanges involved in, 221 future directions for, 231 global CO2 emissions portfolio, 220–221 Kyoto Protocol, 214–215, 219–220 marketplace, creation of, 218–219 U.S. emissions trading, 217–218 Greenhouse gas (GHG) emissions, 214–216, 219–220, 231 Handy, 148 Heating degree days (HDDs), 23, 25–26 Hedge accounting, 89–93, 100 Hedging, see specific types of hedges in freight industry, 148 temperature, 22 weather derivatives, 22–32 Herfindahl-Hirschman Index (HHI), 195 Hungary, energy and economic tendencies in, 103–104, 122, 126 Hybrid fuel plants, 209 Imarex, 148 Inadvertent hedging, 178 Independent power producers, defined, 158 Independent System Operator (ISO), 179–180, 187, 204 Industrial electricity demand, 162 Installed capacity, 181 Institutional framework, in convergence process, 108–109 Insurance, supply risk and, 207. See also Reinsurance Integration: implementation of, 187–188 need for, generally, 177–178 types of, 181–186 Intercontinental Exchange (ICE), 226, 229 Interest rate swaps, 148

INDEX

International Accounting Standards Committee (IASC), 91 International cooperation, in convergence process, 107–108 International Petroleum Exchange (IPE), 144–145, 149, 221, 226 International Swaps and Derivatives Association (ISDA), 30 Internet: emissions trading, 221 liquefied natural gas (LNG), 224 privatization of, 34 trading, impact of, 9–10. See also Electronic trading Interpower, 132 Intra-Asia market, bandwidth market liquidity, 43–45, 48–49 Intra-European market, bandwidth market liquidity, 43–44, 48 Inventory hedge, 96–97 Italian Edison, 134–135 Italy: electricity sector turmoil, 131–133 gas liberalization, 134–136 multi-utility model, 136–138 risk management products, 138–141 Japan, emissions trading, 214–217 Joint implementation (JI), 216 Jumps, energy option valuation, 84–85 Klun, William A., 3 Koch, 36, 228 Kumar, P., 3 Kyoto Protocol (1997): components of, 214–215 COP7, 219 corporate responses to, 219–220 Lease-back programs, 181 Least squares regression, 73 Leipzig Power Exchange, 229 Liberalization: in European Union, 131–132 Italy, 134–136 Liquefied natural gas (LNG): future directions for, generally, 223 regasification terminals, 134 trading, 224–225

Index

Liquidity: bandwidth market, 33–49 defined, 145 significance of, 8 Load profiles, 208 Load-shaping technologies, 209 Local distribution companies (LDCs): financial supply risk and, 206 Italy, 132, 135–137 Location, in weather derivatives, 22 Lock-in/lock-out breaks, 105 Long-term power purchase agreements (LTPPA), 115–116, 124, 126, 158, 173 Lotus 1-2-3, 184 LSC (level, slope, and curvatures) market model, energy option valuation: alternative stochastic processes, 77–79 characteristics of, 70–77, 86 MAAC market, 159, 162, 172 Mallory, Jones, Lynch, and Flynn, 147 Marginal economics, 155 Margulis, Howard, 3 Maritime industry, 144–146 Mark-to-market, 71, 91, 93 Market analysis, 7 Market comparables approach (MCA), energy option valuation, 53–57, 66 Market environment, in convergence process, 109 Market fragmentation, 230 Market indexes: applications, 10–13 defined, 5 importance of, 5–9 methodological considerations, 9–10 Market intelligence, 207 Market makers: bandwidth market, 35, 48 electricity market, 157 freight market, 151 Over-the-Counter derivatives, 90 Market niches, overlap of, 30–31 Market value, importance of, 8

249 Market volatility, electric generation finance and, 154 Material adverse change (MAC)/materially adverse effect, 191, 195–197 Mauro, Alessandro, 3 Mean-reverting, multivariate normal process, energy option valuation, 82–84 Mean-reverting, multivariate normal with Poisson jump process, energy option valuation, 84–85 Mediation, 184–185 Merchant power producers: defined, 158 emergence of, 156–157 Mergers: derivatives and, 194–195 earnings, impact of, 194 “outs,” 195–196 problems with, 196–198 suggested solutions, 198 Metering technologies, 209 Methods-to-market energy, 104 METI, 216 Microsoft: BizTalk, 188 Commerce, 188 Excel, 183–184 Microturbines, 209 Midcontinent Area Power Pool (MAPP), 167 Middle East, as supplier, 134 Middlemen, in shipping industry, 144 Miller, Nedia, 2 Mirant, 158 Modelling, 20 Monte Carlo simulation, 139–140 Montreal Protocol (1987), 214 Morgan Stanley, 226 Multicommodity market, 223 Multiple commodity contracts, 187 Multi-utility model, Italy, 136–138 Multivariate normal model, energy option valuation: characteristics of, 79–82 mean-reverting, 82–84 mean-reverting with Poisson jump, 84–85

250 National Electricity Regulatory Commission (NERC), 121 National Energy Policy, 204–205 Natsource, 217, 221 Natural gas: forward prices, 71–72 importance of, generally, 8 infrastructure, 163–165 liquefied, see Liquefied natural gas (LNG) option contract, valuation approach, 52 NEDO, 216–217 NEG, 158 New England ISO, 175 New England power exchange, 156 New England Power Pool (NEPOOL), 160, 167, 172, 176 New Markets at Spectron, 15 New York ISO, 175 New York Mercantile Exchange (NYMEX), 144–145, 149, 186 New York power exchange, 156 Nigeria, as supplier, 134 NirvanaSoft Inc., 178, 189 Nontemperature products, 27 Nord Pool, 229 Normal distribution, 20 North American Electric Reliability Council (NERC), 158, 203 NOX (nitrous oxide) emissions, 216–217 Nuclear power, Italian ban on, 139 Nuclear power plants, 124 NWPP market, 162 NYMEX, 96–98, 225–227 Oil economics, price volatility and, 146 Oil-indexed pricing, 135, 139 Oil industry, generally, 144 One Seas/Levelseas, 150 OPEC, 144 Operations staff, role of, 180 Option hedge, effectiveness of, 94–95 Option pricing model, 70 Options, see specific types of options applications, generally, 10 defined, 145–146 freight, 148–149

INDEX

plain vanilla, 11–12 supply risk and, 207 Oracle, 184 Originators, defined, 179–180 Oslo Stock Exchange, 148 Other Comprehensive Income (OCI), 92–93 Outsourcing, mergers and, 197, 208 Over-the-Counter (OTC): derivatives market, 90 energy markets, 144 swaps markets, 147 weather derivatives, 24–25 Overcapacity risk, electricity market, 162 Overhead lines, maintenance of, 204 Panamax, 145, 148 Pan-European energy market, 105 Paris Bourse/UNIPEDE, 221 Path dependencies, 105–107 Payout distribution, reinsurance, 19–20 Payout profile, reinsurance, 18–19 Pennsylvania/New Jersey/Maryland power exchange (PJM), 159 Period, in weather derivatives, 22 Photovoltaics, 209 Physical demand risk, 206 Physical supply risk, 206 Pipelines, Italy, 134–135 Pipes and wires utilities, 158 PJM ISO, 175 Plain vanilla, generally: options, 11–12 swaps, 12–13 Platts Global, 229 Plurigas, 137 Poisson probability density function, 84 Poland: capacity development, 112–118 competition, 115–117 demand development, 110–118 electricity market structure, 111 electricity trade prospects, 126 energy and economic tendencies in, 103–105, 110–111 international energy exchange, 114 liberalization, 115–116

Index

price development, 117–118 Ukrainian exports to, 123 Warsaw Power Exchange (PolPX), 116–117, 126 Polish Energy Act, 110 Polish Power Exchange, 229 Political tension, impact of, 146 Polskie SiECi Elektroenergetyczne S.A. (PSE S.A.), 115, 126 Pooling points/hubs, 34–35, 48 Power marketers: affiliated, 197 defined, 158 emergence of, 157 Prebon Yamane, 36, 217, 221 Preferential pricing, 23 Price decline, percentage, 40–42 Price erosion, 39 Price risk, 138–141, 205–206 Price volatility: freight trading, 146–147 significance of, 11 Pricing: preferential, 23 strategies, generally, 138–139 technological advances and, 209 Primary market, 31 Production hedge, 97–100 Profit and loss (P&L): inventory hedge, 96 production hedge, 97–98 Proportional allocation, 18–19 Public Utility Holding Company Act (PUHCA), 204 Put-call parity, energy option valuation, 54–56 Put options: characteristics of, 11–12 energy option valuation, 54–55 weather derivatives, 24 Q-88 data, 145 Quicktrade, 228 RateXchange Trading System (RTS), 2, 36–38, 46 RateXlabs, 36 Real-time price indexes, 6 Real-time purchasing, 209

251 Real-time switching/trading, 39 Recession, impact on electricity market, 161 Recordkeeping requirements, 193 Regional Transmission Organizations (RTOs), 156, 159, 205 Regulatory requirements, 7–8 Reichel, Dr. Markus, 2–3 Reinsurance: defined, 16–17 facultative, 17–18 risk allocation, 18–20 treaty, 17–18 weather derivatives and, 29–31 Relational database, 184 Reliability, significance of, 204 Reliant Energy, 226 Reporting requirements, 29, 193, 198 Reserve margins, wholesale electricity markets, 166–167 Revenue leakage, 178 Risk adjustment, 52 Risk allocation, reinsurance, 18–20 Risk management: defined, 146 in freight trading, 147–150 Italian energy market, 138–141 risk allocation, 18–20 weather, historical, 15–16 Risk-neutral valuation, 52 RWE (Germany), 226 Rybnik, 126 Schedule coordination, 181 Seasonality, 71 Secondary market, 31, 181 Securities and Exchange Commission (SEC), 192 Securitization, 31 Self-financing, dynamic replicating strategy, 63 Sensitivity coefficients, 80 Service providers, shipping industry and, 144 SG Investment Banking, 226 Shell, 220, 226 Ship owners, role in freight industry, 145–146, 148–149 Shipping banks, 148

252 Short method, 91 Short-selling, 59, 67 Simple block contracts, 185–186 Simpson, Spence, and Young (SSY), 147–149 Sithe Energy, 158 Slovakia, energy and economic tendencies in, 103–104, 122 Southern Energy, 226 Southwest Power Pool (SPP), 167, 174 Speculation, 93 Speculators, bandwidth market, 45 Spot markets/transactions, wholesale electricity markets, 160, 173–174 Spreadsheet programs, 183–184 SQL Server, 184 Standard deviation, 82 State-claims, energy option valuation, 52, 58 State-space, energy option valuation, 59 Statistics and metrics, bandwidth market liquidity, 37–38 Stine, Vice Chancellor, 196 Stochastic discount factor valuation, 52 Stochastic processes, energy option valuation: alternative LSC model, 77–79 significance of, 67, 70 Stovepipe architecture, 178, 180–181 Straight-through processing, 179 Stranded cost, electricity market, 163 Stranded Costs Compensation System (SOK), 115–116 Strike price, 11, 54 Structuring desk, 179 Subsidiaries, in electricity market, 158 Suezmax, 145 Sulfur dioxide (SO2), emissions trading, 214, 217–218 Supplier(s): qualifications of, 208 relationship with, 7–8, 210 Supply and demand: bandwidth market, 39, 41 electricity, 155, 166–168

INDEX

electricity prices, impact on, 203 price volatility and, 146 Supply risk, electricity market, 207–208 Swap(s): applications, 10 defined, 146 Over-the-Counter, 147–148 plain vanilla, 12–13 supply risk and, 207 as weather derivative, 23–25 Swiss Electricity Price Index, 7 Sydney Futures Exchange, 221 Tankers UK, 149–150 Tariffs, Ukrainian energy sector, 121–122 Telecommunications Act of 1996, 34 Telecommunications bandwidth, 223–224 Telecommunications era, 34, 36 Temperature: cooling degree days (CDDs), 23, 26 heating degree days (HDDs), 23, 25–26 as weather derivative, 27 winter, historical average, 21 Tenaska, 158 Thermal storage, 209 Time chartering, 146 Time series analysis, bandwidth market liquidity, 40–42, 48 Time-series processing, 186 Timing, in weather derivatives, 22 Tokyo Commodity Exchange, 229 Totalfina Elf, 226 Tracking methods, bandwidth market liquidity, 45–48 Traders, role of, 180 TradeSpark, 228–229 Trading desks, bandwidth market, 35–36, 48 Trading hubs, 156 Tradition Financial Services (TFS), 36 Transaction-based price indexes, 9 Transaction Hub (Lodestar), 188

Index

Transaction hubs, 188 Transaction management, integrated, 188 TransEnergy Management, 228 Transfer protocols, 188–189 Transformer companies, 30 Transmission, wholesale electricity markets, 172–173 Transmission lines/wires, 203–204, 206 Treaty reinsurance, 17–18 TXU, 228 Ukraine: capacity development, 119–127 competition, 119, 121 demand development, 118 energy and economic tendencies in, 105 energy market structure, 119–121 liberalization, 119–120 nuclear power plants, 124 price development, 122 privatization, 122, 124 Underlying assets, 31, 67, 69–70 Underwriter, reinsurance, 18–19, 30 U.S. Department of Energy, 203 U.S. Energy Policy Act of 1992 (EPAct), 155, 158, 201 U.S. Environmental Protection Agency (EPA), 217–218 U.S. Federal Energy Regulatory Commission (FERC), 155 United States–Europe, bandwidth market liquidity, 42–43, 48 Valuation: alternative stochastic processes LSC model, 77–79 bandwidth market, 46 cash flow adjusted approach (CFAA), 57–63, 65–67 discount factor adjusted approach (DFAA), 63–67 importance of, 5 LSC market model, 70–77, 86 market comparables approach (MCA), 53–57

253 mean-reverting, multivariate normal process, 82–84 mean-reverting, multivariate normal with Poisson jump process, 84–85 multivariate normal model, 79–82 option framework, 67–70 pricing strategies and, 139–140 Value-at-risk (VaR), 75, 77, 140, 182 Vann, Kirk, 3 Vattenfall, 126 VLCC market, 145, 150 Volatility, see Price volatility defined, 146 energy options valuation, impact on, 70–71 generally, 11 WAF–USGC/USEC, 145 Ward, Nicholas, 2, 15 Warsaw Power Exchange (PolPX), 116–117, 126 Weather: derivatives, see Weather derivatives energy risk and, 205 forecasts, 28 price volatility and, 146 variables, types of, 21–22 Weather derivatives: common structures, 23–25 cooling degree days (CDDs), 23, 26 defined, 20–21 future directions of, 32 heating degree days (HDDs), 23, 25–26 key elements of, 21–23 market development, 28 price action dynamics, 28 progress to date, 27 reinsurance and, 29 risk management, see Weather risk management temperature, as dominant product, 27 Weather risk management: historical, 15–16 policy implementation, 26–27

254 Wholesale electricity markets: composition of, 159–160 growth of, 156 influential factors on, generally, 154–155 necessity of, 175 by region, 174 supply and demand equilibrium, 166–167

INDEX

Wholesale Market, Ukraine, 119, 121 Williams, 36, 228 Williams, John Burr, 64 Wing risk, 30 World Bank, 107 Worldscale (tanker rate), 150 World Trade Organization (WTO), 34