Michael Giberson
Among the presenters at the Harvard Electricity Policy Group session on market power monitoring, Speaker 4 was the one who had it all together. Note that the session summaries characterize the discussions but do not name names, so theoretically I can’t link the discussion to a person and I suppose doing so isn’t in the spirit of the thing. But Speaker 4, whoever he or she is, hit the nail on the head:
A major role for the market design process is to refine the local market power mitigation mechanism. It is the kluge that we’ll need.
Local market power mitigation mechanisms require a kluge because the theory and practice of market power mitigation in networks remains underdeveloped. That said, it is fair to say that both theory and practice related to monitoring for and mitigating local market power has improved substantially over the last decade or so.
I’ve found two solid, relatively brief surveys of the theory, such as it is, and practice of electric power market monitoring and market power mitigation. I’ll only comment about the first of the two here, a paper by Frank Wolak, and I’ll post about the other one – by Paul Twomey, Richard Green, Karsten Neuhoff, and David Newbery – at a later date.
Like Speaker 4, Wolak in “Lessons from International Experience with Electricity Market Monitoring” demonstrates a pretty good grasp on the practice of market monitoring. It’s a good thing too, since Wolak chairs the California ISO’s Market Surveillance Committee.
He sets the stage by describing what makes a market monitoring process essential to a well-functioning electric market, and also explains why typical antitrust or competition law alone is inadequate as a basis for electricity market monitoring. He then draws five key lessons from international experiences with electric power markets integrated into transmission system operations. Finally, he illustrates his lessons with examples from various market regimes.
Somewhat condensed, his five lessons are:
- The market monitoring process should be forward looking, seeking to root out market design flaws before they occur.
- The market monitoring process must by fully supported by the government regulatory process.
- The market monitoring unit should prepare consistent measures of market and system performance that are comparable over time and across markets.
- The market/system operator should release all data submitted to or produced by it, excepting only certain limited information.
- The market monitoring process should be indepedent of the market operator, the system operator, and the political process.
Wolak comes close to getting it right when he’s talking about “why electricity is different.” He says because of the way almost all electricity is produced and delivered, market participants share a common interest in reliable operation of the network, even though market participants may often find they can profit from behavior that degrades overall grid reliability. A little later in the paper, as he warms up to his topic, his concern expands to encompass threats to both system reliability and market efficiency.
Personally, I don’t think market monitors have much to do with system reliability, and everything to do with promoting market efficiency. In fact, while the five lessons provide generally useful practical advice, they do not clearly reveal the core function of the market monitor in these integrated wholesale power markets/transmission system operations. That core function is “the kluge that we need”: local market power mitigation. A market monitor that does its core function well will naturally be well positioned to offer useful market design advice, publish market statistics, release information, and so on, but most of these tasks could be done by others as well.
As I explained last week (see also here and here), it isn’t something about electricity per se, but the nature of the transactions and the especially the nature of the network that accounts for the rise of the market monitoring institution with market power mitigation responsibilities. In the particular case of electric power networks, elements like the continuous flow of transactions, the central scheduling service provided by the system operator with the associated information flows, the clearinghouse functions provided by the market operator, and the extensive measuring of system conditions, all contribute to the ability of the network operator to serve as an effective defense against opportunistic behavior.
(Some reader will no doubt object to my dismissing of a central concern for reliability, saying, “if a system isn’t reliable then the market won’t be efficient.” Obviously, however, it is possible for a system to be too reliable, given that reliability doesn’t come free. In any case, reliability concerns should rest primarily with the transmission system operator and their customers, and not the market monitor.)