Chronicle of a deregulation disaster foretold, by Maria Kielmas.
When governments worldwide began to liberalise their electricity markets in the 1980s, and to deregulate them in the 1990s, all the interested parties - policymakers, utility company executives, financiers and re/insurers - were persuaded to believe that this would be a long, though admittedly untried, process of risk transfer. The market, financing, credit and environmental risks associated with power generation, would be shifted from the captive customers of a centralised, usually monopolistic parastatal power generator to the shareholders, financiers and re/insurers of private sector companies which had bought formerly state-owned assets.
The process coincided with the development of new power generation technology using gas turbines, which in turn provided a market for significant discoveries of natural gas. Until then, oil and gas companies had deemed such gas reserves to be of limited value. In the industrialised world this gave governments the opportunity to promote policies enabling the substitution of natural gas for coal in power generation, and to tackle truculent coal mining unions. Governments in the developing world hoped to attract foreign investment and new technology into privatised or part privatised energy sectors during an era when the Washington Consensus - that mix of fiscal discipline and free markets - was the accepted mantra.
This was also a time when the insurance and banking sectors were converging. In the energy industry the gas and electricity sectors were converging. But what few foresaw a decade-and-a-half ago was that the energy and finance industries would not only converge, but result in electricity shortages, market manipulation, fraudulent trades, the Enron bankruptcy and 'mark to just about anything' accounting. Such is the scale of power utility debt which needs refinancing that numerous bankruptcies are still expected and a massive sector restructuring will be required over the next few years in North America and eventually in Europe. Energy trading, the great innovation of the 1990s, has all but expired in the US while it remains only hesitantly alive in Europe. The fundamental issue underlying all of these problems has been the inability of policymakers, market makers, financiers and re/insurers to appreciate the physical nature of the energy industries in general, and the electricity sector in particular.
The first lesson
The re/insurance industry's first lesson in how the reformed electricity sector would behave came with a spate of business interruption claims. Under the former centrally-controlled, state-owned system, the power generator did not buy business interruption insurance; instead the captive customers - the general public - suffered power cuts. In the deregulated energy market, power generators which failed to deliver as contracted would not just lose earnings but would also have to pay a penalty.
New gas turbine technology has also been a major problem, explains Aon executive director Rob Woods, who is currently setting up a power and utilities team in London. Utility companies would try to take advantage of power price volatility and ratchet up power generation capacity during price spikes. In the US electricity market these could typically be of the order of $1,800 per megawatt-hour (MWh) when the average price was around $30 per MWh. Re/insurers had been accustomed to dealing with power companies who operated their turbines with a large margin of safety. But in the new market the independent power producers (IPPs), also known as merchant plants, were sometimes using untested gas turbines. As a result, machinery breakdown claims surged from about 20%-25% of total claims prior to energy market deregulation to about 75% after deregulation.
Some of the largest price spikes in the US electricity market during the period 1999-2000 were caused by natural catastrophes in the Mid West. But many others were the product of now notorious 'wash trades', whereby companies simultaneously bought and sold electricity to the same counterparty in the hope of boosting profits by reporting ever-higher trading volumes. In addition, the costs of repairing gas turbines had been underestimated leaving re/insurers to pick up the slack. The soft reinsurance market of the time meant ever-lower deductibles, so re/insurers were getting a higher frequency of lower level losses, notes Woods.
Eventually, major players such as Cox Power and Brockbank ceased providing this type of cover. Re/insurers have wisened up as their own market has hardened. Machinery breakdown premiums have risen by 300% in some cases while business interruption insurance is offered only up to a particular power price cap so that the insurer will not be liable for spot price spikes.
Lack of understanding
Although the worst of the business interruption and machinery breakdown losses occurred in the late 1990s, during an unusually soft period in the re/insurance market, Woods acknowledges that re/insurers at the time did not appreciate what was happening in the energy industry and how technology was developing. But a worse problem has been that the players in the energy market itself did not understand what was going on. So, as competitive electricity markets have evolved, their consequences have been largely unforeseen and the mitigation and management of associated risks remains at a conceptual stage.
While re/insurers invest much manpower in estimating potential losses from natural disasters, disaster preparedness in the energy sector comes at a price which may not be acceptable in a deregulated or part-deregulated market. Efficient energy markets require a minimum level of redundant (spare generation) capacity. Recommended disaster preparedness in, say, California, states that power generators should hold at least 20% redundant generating capacity. But California's botched electricity sector deregulation and subsequent crisis have been such that power companies are unable to plan properly for standard - let alone redundant - supply. Policy makers and power companies in California are not interested today in hearing about the possible economic consequences of another 1994 Northridge-type earthquake. The economic losses associated with this earthquake were given as $25bn at the time but the losses due to the state's energy crisis could be in the order of $100bn or more. Electricity companies take the view that redundant capacity equals a loss of earnings, especially if power prices peak at $1,000 per MWh.
Nevertheless, most of the economic losses following the 1999 Chi Chi earthquake in Taiwan were due to a lack of redundancy in the power transmission system. There was only one main transmission line from generating centres, situated in the south of the island, to the large industrial users, which were situated in the north. And the resultant disruption of computer chip manufacture hit California's IT sector just ahead of the state's energy crisis.
No-one could foresee what would happen when a competitive electricity market was introduced in the UK in March 2001. The previous arrangement was an Electricity Pool operated by the National Grid Company (NGK). Power generators would bid to supply the pool, with the highest bidder setting the next day price. But this was subject to considerable price manipulation, so the New Electricity Trading Arrangements (Neta) were introduced to avoid this, with the political motive of providing some element of protection for the near-defunct British coal industry. But one unforeseen result of this has been that wholesale power prices in the UK have slumped by 30%, bringing the country's largest generator, British Energy, to the brink of bankruptcy.
Power companies on both sides of the Atlantic rushed to invest in new generating capacity which, they believed, would take advantage of a growing, competitive market. But electricity demand did not reach the companies' unrealistically high, often spurious, forecasts. Under a US accounting rule brought in during the late-1990s, and partly rescinded in December last year, energy companies were able to book profits from long-term electricity supply contracts on a mark-to-market (MtM) basis. This enabled them to account for profits from multi-year contracts within the first year and was widely abused. But the recent change in rules still does not include derivatives trading, where the price exposure is the same. Critics say the inadequacy of this rule change illustrates a lack of clear thinking on the part of the authorities and their overall knee-jerk reaction to the US power sector crisis, which could have potentially disastrous consequences.
US companies such as Enron, Dynergy and Aquila were instrumental in creating electricity trading mechanisms and introducing these in Europe. The idea here was that the deregulated electricity market could adopt trading and hedging instruments and price modelling developed from the financial markets. But seasonality, extreme volatility and mean reversion (where price spikes fall back to the prevailing average) of power prices is not seen in the fixed income markets. More importantly, electricity is not just another commodity. It cannot be stored like wheat, sugar or oil, and cannot be held, like equity. It is a continuous process which functions within a circuit to provide heat and light and operate appliances in real time. So arbitrage - where a trader can take advantage of price spreads between national markets - is not really applicable.
Nevertheless energy traders resorted to off-the-shelf methods developed in the financial markets to price instruments such as options, futures and derivatives. Famous among these was the Black-Scholes method, a framework for pricing options developed in the early 1970s by economists Fisher Black and Myron Scholes, the shortcomings of which were already apparent to the financial markets when it was adopted by energy traders. When these models couldn't cope with power price volatility the modellers attempted to compensate by evolving ever more complex solutions, and the supply of analytical solutions and software aimed at modelling future prices became a growth industry. Graduate mathematicians with a grounding in stochastic differential equations, but no understanding of the idiosyncracies of the electricity markets, such as regulatory or transmission constraints, became the kingmakers of power trading. But they might as well have been throwing dice. There was little liquidity in the new energy trading market; power price data was limited, expensive, manipulated and unreliable. So, elaborate mathematical models of future power prices were meaningless. As a result neither policymakers nor company executives had a clue about the risks they were taking and regulators never made it clear what was a measure of risk, says Rene Carmona, Paul Wythes professor at Princeton University's department of operational Research and Financial Engineering.
Overstated trading earnings - in Enron's case this was about 95% of its business in the year 2000 - contributed to the disintegration of the power market. Enron hid much of its trading losses in now notorious off balance sheet entities. But power companies who used such special purpose entities (SPEs) for legitimate project financing of joint ventures are now coming under extra scrutiny from politicians, credit rating agencies and insurers. Despite the claims of power companies in North America and Europe to have developed risk management strategies to protect themselves from counterparty and credit exposure, the credit risk issue has hardly been addressed, says Carmona. Credit risk received much attention in the academic literature, he explains, "but this does not matter if these elucubrations remain confined to academic textbooks." He add: "There is an enormous gap between what is understood at the research level and the everyday practice of risk management. This is especially true of the energy markets."
With a huge number of power projects seeking refinancing, financial guarantee insurers such as XL Capital Assurance are hoping to cherry pick the best deals. These companies provide financial guarantee and triple-A credit enhancement for investors in the power and utilities sectors. Power companies' debt refinancing faces many problems associated with forward power price projections and continuously changing regulatory and environmental rules. And there is still no agreement in the power trading sector on what a real measure of risk is. The long-preferred measure of risk in a portfolio, one-day value-at-risk (Var), is increasingly viewed as a convenient alibi for those executives who do not understand their risks at all. None of the elaborate price forecast models could predict the California crisis and the slump in UK wholesale power prices. One credit analyst suggests that if anyone in the UK had predicted that in early 2002 power prices could fall by 30%, they would never have been taken seriously.
After the upheavals of the last two years, power companies have taken on board the need to disclose the proportion of their earnings that comes from generating activities, from structured deals and from trading activities. The next step is to encourage meaningful dialogue between the engineers dealing with the wires, pipes and oil/gas wells, financiers, re/insurers, market makers and policymakers. Rene Carmona is trying to promote just such a dialogue at Princeton University. He believes that academics are just as guilty in the energy market's disasters as everyone else. Such recognition of the nature of the problems in the energy sector could be the first step in understanding and managing its risks.
By Maria Kielmas
Maria Kielmas is a freelance journalist and consultant.