Center for the Study of Energy Markets

Underlying the current debates over the appropriate organization of the electricity industry and its wholesale markets is a need for metrics that allow for comparisons of the markets that are already operating. One such metric is provided by competitive benchmark analysis. The basic idea behind a competitive benchmark is to estimate the price that would result if no firm attempted to exercise market power and to compare it to observed market prices. In this paper we estimate competitive benchmark prices for the electricity market overseen by the Independent System Operator of New England (ISO-NE).

We study the period from May 1999 to September 2001. Using the Energy Clearing Price (ECP) of the ISO-NE as a measure of market price, we find the demand-weighted markup between the ECP and the competitive benchmark to be 12%. However, the ECP reflects adjustments for factors, including transmission congestion and other unit operating constraints, that are not explicitly considered in forming our competitive benchmark. Alternative measures of market prices that more closely match our methods for estimating a competitive price can be derived by intersecting market demand with the aggregate supply curve taken from the offer prices of generation units. One such measure intersects the aggregate offer curve of all generation units with overall market demand. Using this measure of price, we find a demand-weighted markup of 4% over our estimate of competitive price.

Another measure of price intersects the offer prices only of large fossil-fuel generation units with the actual demand served by that set of units. This price is usually higher than the previous offer-based measure because transmission and other operating constraints cause many fossil-fuel units to be called upon to operate even though their offer prices are higher than other units that are not operating. The margin produced from this latter offer-based measure is 11%. However this latter margin does not reflect ISO-NE's ex-post reduction of offer prices from certain transmission constrained units. Because of these reductions, the revenues earned by these units will be lower than indicated by this measure. It is extremely difficult to say how large the impact of ex-post market-power mitigation on this index is without more detailed data on specific mitigation events.

The differences among these three measures indicate that factors other than energy bids play an important role in market outcomes in New England. Unit operating constraints are raising the ECP above the levels produced by a simple aggregation of energy bids. In addition, the set of units actually called upon to operate appears to be working larger margins into their energy offer prices than are reflected in the bids of other units. It is likely that the strategic bidding of other unit operating parameters, such Low Operating Limits (LOL) also plays a role in these outcomes.

Other studies using similar methods have examined the competitive performance of the California and PJM electricity markets. The only time period over which all 3 studies overlap is May to December 1999. Over this period the performance of the New England market compares favorably to that of the California and PJM electricity markets. Further comparison of the New England and California markets during 2000 reinforces this impression.

From the perspective of market efficiency the results to date are encouraging, particularly when compared to California, but need to be considered in context. The continued vertical integration of some suppliers and the transition contracts imposed on others provide a powerful mitigating influence on the incentives of these firms to exercise market power. The pending expiration of transition periods and potential consolidation of supply portfolios will reverse this effect. It is difficult to predict the extent to which new entry will offset these trends. The market power created by transmission congestion and other unit operating constraints presents an ongoing challenge to the efficient operation of electricity markets. Last, the operations, monitoring, and market power mitigation functions of the ISO no doubt has contributed to the outcomes detailed in this report. It is important that we continue to examine both the short and long term impacts of these activities, particularly as widespread changes to the ISO-NE's operations and pricing protocols are adopted.