A Simple Guide to Making Money in the



A Simple Guide to Making Money in the

New York Capacity Market

A recent article in Newsday by Tom McGinty (“Consumers hit by spike in electricity fees”, 7/26/06) describes how the “capacity payments” made to generators in the New York City region are going to be much higher this summer even though the potential supply of installed generating capacity has increased substantially. Simple economic theory suggests that more supply should lower prices. Why didn’t this happen? The answer is that, given the existing pattern of ownership of generating capacity and the current design of the capacity market, it is easy for generators to exploit this market for their own gains. This type of exploitation is probably legal, but it also implies that customers will end up paying higher rates in the future and will get little in return.

Even though generators will be paid over $400 million from the capacity auctions this summer (plus payments to an additional one third of the generating capacity through existing bilateral contracts), there is no guarantee that these payments will lead to improvements in the reliability of supply or reduce the likelihood of blackouts in the future. After the recent blackouts in Queens and Staten Island, most people would agree that it would be preferable if Con Ed had spent this money on upgrading the distribution system. Instead, the capacity market designed by our regulators in Albany hands this money over to generators for being “available” to generate electricity --- something they probably would have done anyway. The purpose of this article is to show how easy it is for the generators to manipulate the capacity market and earn more money for their shareholders.

The structure of ownership of generating capacity in New York City (NYC) is shown below in Table 1. Three of the companies, NRG Power, KeySpan and Astoria Generation, are merchant generators that own enough capacity and have the economic incentive to manipulate prices in the capacity market by withholding capacity from the auctions. Two of the companies, Calpine and SCS Energy, are too small to manipulate prices effectively, the Power Authority is, hopefully, operating in the public interest and Con Ed is the main distribution company that is buying most of the capacity, and therefore, has no obvious incentive to get prices higher.

The capacity market has been designed by regulators to provide enough income to cover the capital cost of building a new natural gas turbine (a “peaking unit”). This has been accomplished by specifying a “Demand Curve” for each month, and it specifies the price that will be paid to purchase different amounts of capacity. The Demand Curve for June 2006 is shown in Figure 1, and it is calibrated to the “Minimum Requirement” needed for generating capacity (8798 MW UCAP) to meet the load in NYC reliably. Note that the Demand Curve is specified in terms of “Unforced CAPacity” (UCAP), equal to 94.58% of the “Installed CAPacity” (ICAP), and both measures are shown in Table 1 for all existing generating capacity in NYC (9483 MW UCAP).

The market price for “All Capacity Sold” from Figure 1 is $8,600/MW, but the actual market price was $12,712/MW because only 9054MW UCAP were sold. Why did this happen? The answer is simple. Generators get paid more money for selling less. The income from selling all capacity (9843 at 8600) is only $82million/month, compared to the actual outcome (9054 at 12712) of $115million/month. The difference of $33million/month is substantial and corresponds to a 40% increase of income. It is perfectly rational for generators to sell less than the total installed capacity.

The Demand Curve implies that generators could get paid even more money by selling as much capacity as possible at the highest price allowed (8006 at 22740) to provide $182million/month, well over twice as much as selling all capacity. This raises the question of why didn’t the generators withhold more capacity and get more money? The answer is not simple and involves a combination of economic theory and additional regulatory restrictions on what the generators are allowed to do in the capacity market. First, competition among the different companies reduces their ability to exploit the market, and second, there are additional restrictions on how high the market price can be for individual merchant companies. The following discussion addresses two issues: 1) if the Demand Curve is used in the capacity market, how many different companies are needed to make the market reasonably competitive, and 2) is the Demand Curve the best way to meet the regulatory goal of ensuring that the supply of electricity is reliable in NYC?

Given the shape of the Demand Curve in Figure 1, it is possible to determine the best strategy for a firm in a simplified market that assumes all firms want to maximize their profits without colluding with each other. Figure 2 shows the income paid to the “last” firm in the market assuming that the other firms have already submitted their optimum levels of capacity. Looking at the case with 5 firms, the optimum amount of capacity to sell is 1730 MW UCAP for each firm. Hence, the first four firms sell 6920 MW UCAP and the economic problem for the fifth firm is to decide how much additional capacity to sell. The amount of income received by the fifth firm increases to a maximum and then decreases to zero as this firm submits more capacity into the market. The amount of capacity that maximizes profits for the fifth firm is 1730 MW UCAP, corresponding to a total of 8650 MW UCAP from all five firms. The implication is that with five firms, the optimum capacity is less than the Minimum Requirement in Figure 1 and the “optimum” market price, $16,570/MW, is much higher than the actual market price, $12,712/MW. Customers would have to pay more in this market than they currently do in the existing market, and in addition, the amount of capacity sold would not meet the standard required for reliability.

Table 2 summarizes the market outcomes that correspond to the maximum levels of income for the different numbers of firms shown in Figure 2. For a pure monopoly with only one firm, the maximum income occurs at the kink in the Demand Curve, and the dotted line (No Price Cap) shows what would happen without a limit on how high the price can go. As the number of firms increases, the total amount of capacity sold also increases. Consequently, the price paid decreases, and more importantly, the total cost for customers (the Total Income for generators) also decreases. As expected, economic theory shows that more firms make the market more competitive with lower prices.

The total capacity sold in a market with 10 firms corresponds quite closely to the amount of existing capacity in NYC. The implication is that redistributing existing capacity among 10 firms, each one owning less than 950MW UCAP, would eliminate the economic incentives to withhold capacity from the market. This in turn demonstrates why the current proposal to sell all of KeySpan’s 2252MW UCAP to one company is not a good idea. Selling individual power plants to different companies would make the capacity market, and the regular spot market for electricity, more competitive and more difficult to exploit. This is an issue that should be considered seriously by regulators if they plan to continue the use of a Demand Curve in the capacity market.

Real markets are much more complicated than the idealized conditions used to derive the results in Table 2. To really understand what happened in the capacity market in NYC, one would need private information about existing contracts and the bids and offers submitted into the different capacity auctions (six month, one month and spot auctions). Only regulators have access to this type of information. Hence, any analysis based solely on observed market outcomes and the structure of ownership of capacity (Table 1) must be largely speculative, and any results depend on making assumptions about how individual companies behave in the capacity market. In spite of these limitations, it is still possible to come up with a plausible explanation of what happened.

The basic assumption about behavior in the capacity market follows from the earlier discussion below Table 1. Only the three largest merchant generators, NRG Power, KeySpan and Astoria Generation, are likely to have the incentive to exploit their market power and increase the market price. As a result, it is assumed that all 3868MW UCAP from the other four companies will be sold in the capacity market. Since the other companies know this quantity, their economic problem is how to maximize their income assuming that 3868MW UCAP has already been contracted. The red line in Figure 3 (3 Companies) shows the income for the “last” company in the market. The basic characteristics of the results are the same as they were in Figure 2. As the company submits more capacity into the market, the company’s income increases at first, reaches a maximum and then decreases to zero. The optimum amount of capacity to sell is 1629MW UCAP at a price of $15590/MW to given an income of $25million/month. These results are similar to the results for five competing firms with no existing contracts in Table 2, and the prices in both cases are substantially higher than the actual market price ($12712/MW).

One of the three merchant generators, NRG Power, is relatively small and owns only 1356MW UCAP. Since this quantity is less than the optimum, the best strategy for NRG Power is to sell all of its capacity. The economic problem for the remaining two companies is how to maximize their income assuming that 5224MW UCAP has been contracted. The green line (2 Companies) in Figure 3 shows the resulting levels of income per company. The two companies make more money than they did with a third competitor because the market price ($16460/MW) is higher and they sell more (1720MW UCAP). The blue line (1 Company) assumes that only KeySpan is trying to manipulate prices and the resulting price ($15080/MW) and the amount sold (1576MW UCAP) are lower than they were for the other two cases.

Overall, the market results for the three cases considered in Figure 3 are very similar to each other, and the market prices are always higher than the observed market price ($12712/MW). Does this mean that there was no manipulation of the price, and if there was manipulation, why is the observed price so low? The answers to both questions can be found by considering the implications of having additional regulatory restrictions on the prices that the three largest merchant generators can be paid in the capacity market. The price observed in the capacity market is equal to the price cap imposed by regulators on KeySpan. Given this price cap, the best that KeySpan can do to maximize profits is to sell as much capacity as possible at the price cap of $12712/MW, and at this price, 429MW UCAP of existing capacity remains unsold. Hence, one or more of the merchant generators did manipulate the capacity market by raising the price, but not as much as they could have done if they had maximized their income without a price cap.

By setting a price cap on the incumbent merchant generators, the regulators have set an arbitrary limit on how much manipulation is allowed in the capacity market. In other words, the regulators consider that a payment of almost $700million over the summer (assuming the price cap is paid to all capacity sold) is an acceptable amount to pay to generators for being available to generate electricity. The answers to the two questions posed earlier are 1) Yes, the capacity market is being manipulated by generators to get the price and their total income higher than the competitive levels, and 2) the level of manipulation is restricted to the level allowed by regulators. If regulators had wanted to provide better incentives for generators to offer more capacity into the auction and lower the price of capacity, they could have implemented different rules on behavior. For example, if the rules were that all generators owning more than 1000MW UCAP had to submit this capacity at a minimal price and be price-takers in the market, Calpine or SCS would be the merchant generators that could set the market price. However, both of these companies are too small to benefit from withholding capacity from the market. As a result, the market would be competitive, all existing capacity would be sold, the market price would be $8600/MW and the total payment to generators would be less than $500million, corresponding to savings of $200million for the public.

An important implication for public policy is that the amount of money changing hands in the capacity market is very sensitive to how regulators specify the rules. The current rules allow the generators to manipulate the market price to a limited extent, but there is no obvious economic justification for allowing this level of manipulation over another level. It is clear, however, that the current market is not truly competitive and that generators are getting paid more for capacity than they would be in a competitive market. The additional income paid to generators in the capacity market increases the market value of existing generating units. Consequently, the current market value of generating assets for a merchant generator is inflated by the expected value of the extra income, above competitive levels, that can be obtained from the capacity auction.

For an old, inefficient generating unit with a high heat rate, this “excess” income in the capacity market can be the major component of its market value. For a company like KeySpan that is considering selling its existing assets to another company, the extra income from the capacity market, up to the limit allowed by the regulators, is an important component of the total market value of the company’s generating assets. Once a sale has occurred, this extra income will be capitalized into the purchase price of the generating assets and it will be much harder for the public to recover the excess income though more enlightened regulation. This is a crucial issue for public policy that should be resolved through open debate rather than behind closed doors. There is no economic reason why the capacity market could not be made substantially more competitive.

A second, and more fundamental, issue for public policy is whether the capacity market is accomplishing its regulatory goals. The original purpose for implementing the Demand Curve was to encourage the construction of new generating capacity when it is needed to maintain reliability. Unfortunately, this goal has not been met, and the current projection made by the New York Independent System Operator (NYISO) shows a shortfall in 2008. Something will be done to deal with this imminent problem, and the most likely solution is that the retirement of one or more existing power plants will be delayed. However, this solution does not change the conclusion about the capacity market. The public is paying over $1billion/year in the capacity market in NYC. This is enough money to finance a substantial amount of new generating capacity.

Since the Demand Curve in Figure 1 is calibrated to give a price at the Minimum Requirement that is equal to the monthly cost of financing a new peaking unit, the annual cost of capital is $18,1944/MW (12x15612). For an expenditure of $1billion/year, it would be possible to finance 5,500MW UCAP of new generating capacity, corresponding to over 50% of the existing capacity. There is more than enough money being spent in the capacity market to maintain the reliability of supply and have a substantial amount left over for upgrading the distribution system. Instead of spending the money on these important items, the money is being given to generators to ensure that existing capacity is available to generate electricity. This seems an excessive amount to pay for relatively little in return. The main effect of the current expenditure in the capacity market is to inflate the market value of existing capacity. The whole rationale for implementing the Demand Curve should be reevaluated at this time before the inflated values of existing generating units are locked in by the sale of these units to new owners.

To summarize the conclusions, first, it is clear that the capacity market in NYC can be and is being manipulated. This market is not competitive. Second, if regulators continue to use the Demand Curve, it would be possible to make the capacity market more competitive by ensuring that individual power plants are sold to different, new owners. A basic economic principle is that having more merchant generators will make the market more competitive. Finally, the cost of current capacity payments to the public is substantial, and it is almost certain that there are better ways to maintain the reliability of supply than using the Demand Curve. A serious evaluation of the effectiveness of the current capacity market is needed at this time to ensure that there are tangible benefits for the public from future expenditures.

It interesting to note that the new capacity market proposed for New England has a substantially different structure than the capacity market in NYC, and has features that make it much more likely that investment in new generating capacity will occur. For example, the New England market is a forward market that sets the price of capacity three years ahead, and it also allows investors in new capacity to lock in this price for up to five years. A major design flaw of the market in NYC is that the Demand Curve only sets the price of capacity for one month ahead. Even though the specifications of the Demand Curve are published three years ahead, this does not provide potential investors with a guaranteed forward price, and therefore, it does little to reduce the financial risk of building new generating capacity.

-----------------------

[pic]

[pic]

[pic]

[pic]

[pic]

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download