Report on POWER Conference

I attended the U.C. Berkeley POWER (Program on Workable Energy Regulation)
Conference on March 16. Here is a brief report on the relevant parts.
Notably, the Lieff, Cabreser duo of Bill Bernstein and Barry Himmelstein
were there, listening carefully.

The program consisted of the presentation of 8 economic studies, followed by
comments from two discussants, followed by audience questions/comments. The
following is a list of the papers and the discussants who presented with
slides, taken from the website of the U.C. Energy Institute,
http://www.ucei.berkeley.edu/ucei/conf2001/order_papers2001.html: You can
click on this to see the slides, in PDF format. The papers in green are not
particularly relevant for our purposes, and I won't say more about them
beyond the note in brackets. The presenter's name is in bold.

"Bidding Asymmetries in Multi-Unit Auctions: Implications of
Bid Function Equilibria in the British Spot Market for Electricity" by Greg
Crawford, Duke University, Joe Crespo, NERA, and Helen Tauchen, University
of North Carolina
"Pricing and Firm Conduct in California's Deregulated
Electricity Market" by Steve Puller, UC Berkeley
Discussion by Anjali Sheffrin, California
Independent System Operator
"Identification and Estimation of Cost Functions Using
Observed Bid Data: An Application to Electricity Markets" by Frank Wolak,
Stanford University [A paper about economic "tools" to study power markets.]
"Forward Contracts and the Curse of Market Power" by Jeffrey
Lien , University of Maryland
"The Impact of Retail Rate Deregulation on Electricity
Consumption in San Diego" by James Bushnell and Erin Mansur, UC Berkeley
[This looks at whether consumption declined with price increases. Not
much.]
"Consumption and Home Energy Costs: How Prevalent is the
'Heat or Eat' Decision?" by Julie Berry Cullen, University of Michigan,
Leora Friedberg, University of Virginia, and Catherin Wolfram, UC Berkeley
[A macro study on how consumers change overall spending patterns when they
have to shell out more $$ for power.]
"A Quantitative Analysis of Pricing Behavior in California's
Wholesale Electricity Market During Summer 2000" by Paul Joskow, MIT, and
Edward Kahn, Analysis Group
"Electricity Restructuring and the Cost of Pollution
Reduction" by Dallas Burtraw, Karen Palmer, Ranjit Bharvirkar, and Anthony
Paul, Resources for the Future

Before I discuss the specific papers and presenters, one big picture point
needs to be made. There was an unchallenged consensus at this conference
that the generators have exercised market power to the tune of billions of
dollars. The focus was on how and how much, not whether. The good news is
that I heard no evidence supporting any collusion theory; the thought was
that generators are making independent output and pricing decisions knowing
they could influence the market price given the auction rules and the
completely inelastic demand. On the other hand, I would have to say that
the economic work on the tacit collusion hypothesis in incomplete at best.
The bad news is that the scale of potential overcharges is pretty staggering
-- <$5 billion. The plaintiffs will be able to put together quite a damage
study.

Crespo, Bidding Asymmetries in UK: This is a marginally relevant paper
examining whether bidders in the UK electricity auction markets behaved in a
leader-follower mode, i.e., asymmetrically. Crespo's model shows that in a
uniform price-setting auction with clearing price rules, a price setter will
emerge and take all prices above marginal cost. Thus, above marginal cost
pricing does not require coordination. Crespo (from NERA) appeared
knowledgeable, but is not an inspiring speaker. His paper also got roughed
up a bit in the audience questioning segment.

Puller, Pricing and Firm Conduct in California's Deregulated Electricity
Market: This is a very relevant paper, as it tries to determine whether
market power (presumably) exercised in California was static or dynamic,
meaning the product of individual firm decisionmaking (static) or tacit
collusion (dynamic). The period studied was 4/98 to 12/99. Puller found
evidence of static market power consistent with so-called "Cournot pricing."
This theory posits that in an oligopoly firms will take their rivals'
observed price/output decisions as a given and decide how to maximize
profitability given that behavior. He then tried to determine whether any
dynamic games were occurring, meaning a game where firms recognize their
interdependence and try to follow a "supergame trigger strategy" in which
firms try to induce favorable responses from rivals. It's quite complicated
how he goes about this, but fundamentally he tries to correlate observed
output decisions with how a firm at that time might have expected a change
in its behavior to affect its future share of the market. With this
methodology, Puller finds what he called "weak evidence" of forward-looking
dynamic pricing for a brief time in 1998, but not otherwise. My impression
was that the evidence for 1998 was very weak, and the logic used to arrive
at this conclusion was also weak.

Puller then went at this a second time with a theory that attempts to
determine what a firm's "supply function" would look like if it was
exercising static and dynamic market power, and then comparing this to an
estimate of that firm's actual supply function. I found this even more
speculative than the first theory. Nonetheless, the results are basically
the same: evidence of static, but not dynamic market power.

Puller took a lot of heat for the methodology of this paper during the
questioning segment. He's not a dynamic speaker and does not appear to me
to be a strong expert candidate.

Anjali Sheffrin's commentary was very important. She is the Director of
Market Analysis for CAISO. After general comments on Crespo and Puller, she
launched in to a discussion of whether their models explained the California
experience. This turned out to be a preview of the CAISO FERC filing of
last week, in which they allege $5.5 billion in market power-related
overcharges from May 200 to Feb. 2001. That filing and Sheffrin's report
follow. They are essential reading.

<<2001032214541122276.pdf.pdf<< <<2001032214585222924.pdf.pdf<<

Sheffrin maintains that it was the absence of imports during this period
that left the market power of the California generators unchecked. This
permitted the in-state generators to engage in either economic or physical
withholding of power. (Economic withholding is bidding a higher-than-needed
supply curve; physical withholding is cutting output at the plant.) Her
study was intended to (1) Identify individual firms engaging in market power
activity, and (2) Analyze how each firms' actions set market clearing
prices. She utilized full bidding data in CA ISO real time market for
each hour between May and Nov 2000, defined and categorized bidding patterns
and identified economic or physical withholding, and then calculated
bid-cost mark-up and a monopoly rent. She found what she claimed was strong
evidence of both types of withholding, but that economic withholding is the
dominant bidding pattern used by the five large California generators.
While Sheffrin's study does not name names, it claims that "most of the five
in-state suppliers and many of the [16] large importers displayed bidding
patterns which were consistent with the exercise of market power."

Bernstein and Himmelstein were positively gleeful during this presentation.

On the question of collusion, Sheffrin's study is not terribly illuminating.
She maintains that "the dominant bidding pattern is consistent with two
characteristics of a supply function equilibrium model of oligopolist
pricing." I take that to mean Cournot, which is a static, non-collusive
model. However, when you read her study, you'll see the picture is not
entirely clear.

Lien, The Curse of Market Power: This is only marginally relevant for us,
as its thesis is that forward-looking supply contracts are better for both
producers and society. Everyone seemed to agree -- and were confused why we
needed a paper to prove it. Lien (U. Md.) is young and not expert material.

Joskow and Kahn: http://www.mit.edu/people/pjoskow/JK_PaperREVISED.pdf
(paper); http://www.ucei.berkeley.edu/ucei/conf2001/Slides/Kahn_Slides.pdf
(Kahn's slides). I'm sure many of you have already read this; everyone
should. Prepared for SoCal Edison, it contends that 4 in-state generators
(Reliant, Dynegy, AES/Williams, and Southern/Mirant) exercised market power
by withholding capacity during the summer or 2000.

J&K use publicly available data on loads, market prices and generation to
(a) quantify combined effects of "market fundamentals" on market prices, (b)
calculate "price gap" (difference between actual prices and "competitive"
market benchmarks), © account for quantify effects of ISO's ancillary
services requirements and forced outages, and then (d) calculate the "output
gap" for high priced hours, meaning the difference between observed and
maximum profitable levels of generation. They conclude that prices were far
in excess of the competitive benchmark and that the 4 in-state generators
could have produced more power at competitive prices but chose not to.
Hence, market power was exercised.

Kahn, the Analysis Group economist who presented, is a very colorful and
rather undisciplined advocate of his position. He threw around allegations
of conspiracy rather casually, but mostly to be funny. I couldn't tell
whether he believed his study proved that; I don't think it even speaks to
it. But Bernstein and Himmelstein applauded him when he finished, grinning
ear to ear.

Severin Borenstein, UC Berkeley and the conference director, was supposed to
comment on J&K, but hardly did. He said that the generators would have been
stupid not to exercise market power given the supply/demand conditions and
the market rules, and argued that permitting long term contracts and
requiring real-time residential pricing were the solutions to all of this.
Borenstein was the best expert material I saw at this conference, and the
fact he took a pass on the details of the J&K paper is puzzling to me. It
made me wonder whether he already has a horse in this race, or perhaps is
trying to "stay above it all."

Hope this is useful.

Regards,
Dan

Daniel M. Wall
Latham & Watkins
505 Montgomery Street, Suite 1900
San Francisco, CA 94111-2562
Direct: (415) 395-8240
Main: (415) 391-0600
Fax: (415) 395-8095
dan.wall@lw.com

This email may contain material that is confidential, privileged and/or
attorney work product for the sole use of the intended recipient. Any
review, reliance or distribution by others or forwarding without express
permission is strictly prohibited. If you are not the intended recipient,
please contact the sender and delete all copies.

- 2001032214541122276.pdf.pdf
- 2001032214585222924.pdf.pdf