Public Utilities Reports

PUR Guide 2012 Fully Updated Version

Available NOW!
PUR Guide

This comprehensive self-study certification course is designed to teach the novice or pro everything they need to understand and succeed in every phase of the public utilities business.

Order Now

Fingerprinting the Invisible Hands

Opaque markets inflate power prices.

Fortnightly Magazine - July 2009

delay in disclosure will not cause a loss of public confidence because much of the time prices in the ERCOT-administered markets are not subject to the type of price spikes that could create an impression of market power abuses or other market failures. In some cases, however, prices may spike to higher than usual levels and cause public concern and the need for more public information. To address such events, the proposed amendment includes an event trigger that would require the public release of entity-specific information on a much quicker timeframe. The proposed amendment requires that, when the trigger is exceeded, the portion of every market participant’s offer curve that is equal to or exceeds the trigger level will be disclosed seven days after the day for which the information is submitted. The commission finds that the disclosure of this limited type of entity-specific information is sufficient to retain public confidence in the ERCOT markets while minimizing early disclosure of entity-specific information. 9

Implementation of the order took place with market data for September 22, 2007. 10 For the first time, a situation existed in which there was a statistically testable hypothesis. Using available data concerning bidding behavior on an hourly basis both before and after the change in disclosure delay, regression analysis can test whether additional transparency does reduce bids, and indirectly, prices.

Shame Caps

This analysis uses a set of ERCOT bidding rules and market conditions to determine bidding behavior. It doesn’t attempt to model ERCOT’s pricing algorithm since the algorithm is considerably more complex, and less transparent, than the bidding data it uses as part of its calculations (see Figure 2).

First, the analysis calculates two measures of bidding behavior in the ERCOT balancing energy services market. “Maximum bid” represents the highest bid during the hour and “average bid” represents the average bid during the hour. The simplicity of these two measures constitutes their primary value. There are an infinite number of possible measures that could be designed to characterize the bid curves. Opening the analysis to each one of these would eliminate the significance of the statistical results, since each alternative potentially would have a high t statistic. The best course is to choose and test the simplest hypothesis to avoid biasing the statistical estimates.

Following the same argument, the independent variables also are simple. The first two independent variables are natural gas prices and ERCOT load. These two variables are standard choices for independent variables in wholesale electricity markets and have been used in many studies. The analysis adds three other independent variables:

Shame Cap : For years, ERCOT published bids over a specified price. The price level has changed over time to its current level of 100 times natural gas prices.

Reporting Delay : The number of days until bid data is revealed.

Price Cap : The maximum bid accepted by ERCOT’s computer algorithm.

The models tested are:

• Max Bid = A + B x Gas Price + C x Load + D x Shame Cap + E x Reporting Delay +