The deregulation of electricity generation around the world was thought to be good news for customers. In place of government agencies compelling them to pay for capacity that far exceeded any imaginable need, markets would more closely and cost-effectively match supply to demand. If any party suffered, it would be the generators, forced to surrender the guaranteed profits that carried them through the industry’s cycles.

At first, deregulated electricity markets followed that script: wholesale prices went down by 20 to 50 percent, service improved, and the generators’ profits fell. However, recent events suggest that electricity markets don’t necessarily behave like other commodity markets. In California, for example, electricity wholesale prices in 2000–01 were five times higher than they were in 1998 despite the advent of a competitive generation market. Expecting wholesale prices to fall further, industry analysts that year advised utilities to sell their generation assets.

Meanwhile, in the Netherlands, wholesale prices have on average been 30 percent higher than expected over the past two years notwithstanding significant oversupply. In Sweden—the universal model of successful deregulation—the peak price in 2001 actually reached almost ten times the normal one. And in Brazil, a drought that reduced hydropower output has resulted in skyrocketing prices and widespread rationing. There too, most analysts were preoccupied, only a few years ago, with the risk of oversupply.

Two problems

These unexpected and extreme events raise the possibility that a California-style crisis could erupt almost anywhere (Exhibit 1). Temporary regulatory flaws, which are certainly an important element of the problem in many cases, usually get most of the blame. For example, high and volatile prices in the Netherlands—where a capacity oversupply of 20 to 30 percent should logically have forced prices down to the operating costs of the country’s generating plants—are partly the result of a failure, in the aftermath of deregulation, to introduce more competition among the four generators. A drought in Brazil had dramatic repercussions because supply was already falling short as a result of a slow and uncertain deregulation process that discouraged investment in new capacity.

In California, the flaws include a bureaucracy encompassing many regulatory bodies rather than a single authority responsible for overseeing the whole system. One result of this proliferation is a long and complicated process for gaining permission to build new plants. Another regulatory flaw was the deregulation of the wholesale price but not the retail price. An increase in the former and an inability to raise the latter led to bankruptcies among suppliers.¹ (For a solution, see "Power by the minute.")

Although such flaws made things much worse, the root of the problem in California and other liberalized markets was the intrinsic nature of electricity as a commodity:

1. Demand for electricity, whether to run air conditioners or electric heaters, varies daily and seasonally. Since no economically viable way to store it in large quantities exists, substantial reserve generation capacity is required.
2. For most applications, electricity has no substitute, and the cost of shortages—to the economy, to local and national security, and to the health of the public—is very high. As a result, short-term demand is inelastic.
3. Like demand, supply varies with weather conditions: in systems that depend on hydropower, droughts can dramatically reduce it.
4. Power generation is usually subject to tight environmental regulation.

These characteristics produce two unexpected and unwanted effects in deregulated markets: supplies can become excessively tight, and, partly as a result, prices are highly volatile. When supplies run short, there is almost no limit on what can be charged.

Supply becomes too tight

Every electricity system needs to have excess generating capacity for times when demand surges. Because such occasions are by definition sporadic, however, even the high prices charged then don’t fully defray the cost of developing this capacity.

In the old, regulated system, the authorities controlled prices, passing on the cost of unprofitable capacity to the customer. The result was an unnecessary amount of reserve capacity, as well as high prices. In many markets, deregulation was meant to secure lower prices for end users by eliminating excessive reserve margins. And it has indeed eliminated them—only too effectively, it now turns out—because investors, unlike regulators, are not willing to keep, much less build, peak capacity until they are absolutely certain that the market is heading into a longish period of tightness and higher prices. When, in such circumstances, the decision is made to add new capacity, it tends to come too late. Environmental regulation often engenders further delays. The result has been a steady decline in real reserve margins, which are now, in many markets, lower than the required 10 to 15 percent of total capacity (Exhibit 2).

Low reserve margins in deregulated markets lie at the root of an ominous phenomenon: the rapid shift from excess supply to shortages, especially in systems that depend heavily on the weather. California proves the point. Almost a third of the electricity in the western United States comes from hydropower, and air-conditioning is responsible for a significant share of peak demand. The unusually warm and dry weather in 2000 not only boosted the use of air-conditioning but also, at the same time, cut hydro capacity. In a matter of months, a 10 percent reserve margin had completely disappeared.

Volatile prices with almost no ceiling

In electricity markets where prices have been deregulated, wholesale prices "fly up" the moment demand surges, though there may still be significant spare capacity in the system. We believe that two features of electricity are mainly responsible for this phenomenon. First, the inability to store electricity means that supplies are never assured. Unscheduled plant maintenance work, for example, could render suddenly inadequate even supplies that would ordinarily be sufficient to handle unanticipated surges in demand. Prices are bound to reflect this uncertainty, as well as the second special feature of electricity: the lack of any substitute for it.

As a result, electricity prices tend to start rising at lower levels of capacity utilization than do prices for other commodities. A rule of thumb in the chemical industry, for instance, is that prices more than cover operating costs when the utilization of available capacity exceeds 90 percent. But in the electricity industry, our data indicate, utilization rates of only 80 to 85 percent, and sometimes less, will drive price increases (Exhibit 3).

Higher prices in most commodity industries are attributable to the oligopolies that dominate them, but not necessarily in the electricity industry. The generation market in California has more than 200 participants, the largest of them controlling less than 10 percent of total capacity. Yet such fragmentation can do little to mitigate the effect on price of society’s unavoidable dependence upon electricity.

Predictions of electricity price trends must take into account these idiosyncrasies of liberalized electricity markets. We applied our understanding of liberalized markets to Sweden, a part of the Nordic market. Our analysis indicates that the whole of that market, and Sweden in particular, could be the next place to undergo energy shortages and skyrocketing prices.

High electricity prices and energy crises should be completely alien to the Nordic region, which has enormous, varied energy resources, including very significant hydropower supplies and massive oil and gas reserves in the North Sea. In addition, Sweden gets almost half of its electricity from nuclear power. By the time the Nordic region began to deregulate, in the mid-1990s, it had significant overcapacity.

At first glance, everything has gone according to script in Sweden. Prices for both business and residential customers have fallen, and supply has been reliable. Producers, despite reducing costs significantly, have seen their profitability decline by 30 to 50 percent. So all is well? Hardly. From a consumer perspective, the next act could prove to be as profoundly gloomy as an Ingmar Bergman film.

Because of longtime oversupply and low prices, virtually no new capacity has been built over the past decade, and very little is on the drawing board. Meanwhile, demand has grown steadily, cutting the reserve margin from 14 percent to 6 percent—almost the level seen in California in 1999, shortly before the first signs of crisis appeared.

In Sweden, though prices are up, they are not high enough to make large-scale investment in new generating capacity profitable

Prices have moved higher as a result but are still 30 to 50 percent below the levels that would make large-scale investment in new generating capacity profitable. In fact, low profitability is causing the industry to close many existing peak-capacity generation stations. This development has prompted Sweden’s state-run Independent System Operator, which is responsible for the system’s technical operation, to acquire some of them.

Hydro’s share of Sweden’s total capacity is about double that of California’s, making Sweden even more dependent on the weather. Three consecutive years of heavy rainfall and snow boosted the country’s hydropower capacity, thus masking the underlying trend of tighter reserve margins. But last year the weather returned to normal. A dry and cold year—unfortunately, the two go hand in hand—could lead to a situation like the one California endured during the first part of 2001.

Current reserve margins suggest that new capacity is needed in Sweden. The generation industry estimates that new plants could be ready in four to five years if planning and licensing were to go smoothly, which probably wouldn’t happen. Any large-scale development of gas-fired capacity—the low-cost option—would involve disregarding treaties intended to combat climate change. Sweden isn’t likely to do so, since it is a strong supporter of the Kyoto Protocol on Climate Change, which mandated reductions in carbon dioxide emissions. Renewable technologies such as biomass are at least 50 percent more expensive than gas. Sweden’s decision to phase out nuclear power gradually won’t make things easier.

The potential crunch is starting to have an impact on peak wholesale prices, which surged from the equivalent of $50 to $60 per megawatt-hour in 1998 to $200 to $400 per megawatt-hour in 2001. Average wholesale prices doubled in 2001, to $23 to $25 per megawatt-hour, and could rise much higher. In a country where temperatures at peak demand typically range from –25 to –40 centigrade, it is obvious that consumers cannot live without electricity and that generators, though numerous and highly competitive, would be in a strong position to raise prices.

Having experienced five years of oversupply and low, stable prices, Sweden could now face at least five years of the opposite because it would be four or five years before new gas-fired capacity came on line after a decision to increase it significantly.

A strategy for electricity

Naturally, understanding how liberalized electricity markets work can better equip their participants to operate in them. The implications are different for generators, for retailers and industrial customers, and for regulators.

Generators

Some liberalized markets offer generators both high risk and potentially high rewards. Other markets, with more stable prices, offer lower risk and more modest but predictable rewards. Prices are cyclical, for instance, in markets with a large share of hydro and nuclear capacity, because hydro and nuclear plants are cheap to operate but extremely expensive to build. Such a system usually offers very low prices in times of oversupply, because low operating costs allow generators to compete on price. But generators view periods of high capacity utilization as opportunities to recover their substantial investment costs, and these periods tend to come late in the business cycle.

Electricity generators will have to decide very carefully the types of markets in which they want to play. The factors to consider include the financial position of the generators, the profile of their current business, their corporate culture and risk-aversion level, and the expectations of their shareholders. Those choosing to play in high-risk, high-reward markets will need skills and tools—including market dynamic analysis, asset trading, risk management, and the ability to predict prices—that are quite different from those of traditional utilities.

In such an environment, the timing of capacity investments and asset sales is crucial: new capacity must be brought on line just as the market reaches utilization levels of 80 to 85 percent—the point at which prices pick up. That means making investment decisions three to five years in advance, depending on licensing and construction lead times. Many of the generators now planning to move into the US market could already be too late. Examples of volatile markets with high risk and potentially high rewards are Canada, most of South America, the Nordic market, Portugal, Spain, and the western and northeastern parts of the United States.

Prices for electricity generated by thermal systems (gas, coal, and oil-fueled ones) are likely to be steadier. Thermal plants are expensive to operate because of the fuel’s cost, but the relatively low investment required to build them supports prices at lower levels of capacity utilization; the economics of the business do not depend on tight supply. Australia, Germany, the southern and midwestern parts of the United States, and the United Kingdom are examples of low-risk, low-reward markets.

Electricity retailers and industrial customers

An understanding of future price trends will help determine the mix of fixed and variable price components that customers should negotiate, as well as the duration of any ensuing agreement. Since prices are likely to rise in the Nordic market, for example, Swedish industrial customers that have signed long-term fixed-price contracts during the past two years would seem to have a good deal. In view of the close correlation between price volatility and tight market conditions, the historical analysis of price movements—the basis for assigning option values—must be complemented by fundamental analysis of supply-and-demand trends.

Many Swedish retailers, having based the terms of their contracts on the belief that wholesale prices would decline by winter’s end, as had happened for years, got burned in the spring of 2001: the winter of 2000–01 was dry, and prices rose. The lesson is that electricity buyers must cover their supply needs in advance unless they are certain that they can meet those needs later at a lower price.

Regulators

The role of the regulator, to put it simply, is to ensure that the system offers a reliable supply of electricity at competitive prices. But to succeed, regulators need to recognize the fatal weakness of liberalized electricity wholesale markets: their tendency to suffer from overly tight supply, sending prices skyward, from time to time. How can regulators guarantee, without heavy-handedly intervening in markets, that the necessary reserve capacity is added before it is too late?

One possible move would be to pay generators to keep or add reserve capacity. (Of course, the benefits of this approach must be weighed against the risk that these payments could stimulate excessive oversupply and lead to high prices.) Another would be for regulators to impose capacity requirements: generators would have to acquire sufficient available capacity—including a real reserve margin of 10 to 15 percent—to allow them to honor their contracts at times of peak demand. These generators could buy that reserve from players with surpluses, and its cost would be reflected in higher prices.

The problems of liberalization don’t mean that it isn’t a good thing. Liberalized electricity markets introduce competition and therefore generally lead to lower prices. The trick is to find the optimal balance between lower prices and adequate reserve margins. Regulators may be able to help on the supply side. A complementary approach would be to give customers strong incentives to use less electricity at peak periods. It would surely be a shame if the good name of deregulation were to be impaled on a price spike.