This sounds like a good idea.



	"Daniel Douglass" <douglass@ArterHadden.com>
	03/12/2001 06:43 PM
		 
		 To: <Barbara_Klemstine@apsc.com>, <Bob_Anderson@apses.com>, 
<Vicki_Sandler@apses.com>, <berry@apx.com>, <dcazalet@apx.com>, 
<billr@calpine.com>, <jackp@calpine.com>, <Ken_Czarnecki@calpx.com>, 
<gavaughn@duke-energy.com>, <rjhickok@duke-energy.com>, <gtbl@dynegy.com>, 
<jmpa@dynegy.com>, <jdasovic@enron.com>, <susan_j_mara@enron.com>, 
<Tamara_Johnson@enron.com>, <curt.Hatton@gen.pge.com>, <foothill@lmi.net>, 
<camiessn@newwestenergy.com>, <jcgardin@newwestenergy.com>, 
<jsmollon@newwestenergy.com>, <rsnichol@newwestenergy.com>, 
<nam.nguyen@powersrc.com>, <Curtis_L_Kebler@reliantenergy.com>, 
<rllamkin@seiworldwide.com>
		 cc: 
		 Subject: Fwd: Review of FERC Analysis

Gary Ackerman asked that the attached proposal from Ben Zycher be sent to the 
board.
 
Dan
----- Message from "Benjamin Zycher" <bennyz@pacbell.net> on Sun, 11 Mar 2001 
17:45:55 -0500 -----
To:	"Daniel Douglass" <Douglass@ArterHadden.com>, "Gary Ackerman" 
<foothill@lmi.net>
cc:	"Benjamin Zycher" <bzycher@rand.org>
Subject:	Review of FERC Analysis
Benjamin Zycher                                                  (310)
393-0411 ext. 6436
March 11, 2001                                                  (818)
889-8163 home

bzycher@rand.org
bennyz@pacbell.net



Dear Dan and Gary,

     Given the FERC order for refunds from the power producers, I think
that the time has come for a serious paper critiquing the underlying
analyses and explaining market fundamentals.  Analytic errors are a
prominent feature of the Cal ISO "Report" on supply costs (February 28,
2001), and I would be surprised if the same were not true of the FERC
order and any supporting analyses that may have been prepared.  (Dan is
sending me a couple of relevant FERC documents now, but I will not be
able to review them in detail until Tuesday or Wednesday.)  In general,
however, it is fair to say that a common set of analytic errors
underlies such bureaucratic thinking, and the dangers inherent in the
ISO and FERC orders suggest that a detailed refutation is something that
ought to be done.  The central points that I believe I would make
include the following, among others.  (I'm using a small amount of
economic jargon here that I would not use in a report, but I really am
quite experienced at making such arguments in plain English.)

    1. Such analyses typically are incomplete, in that they ignore (a) a
risk premium for non- or under- or late payment.  They ignore also (b)
the fact that part of the marginal cost of selling today is the forgone
opportunity to sell tomorrow, given that units have to be taken down for
maintenance and the like.  In this sense, electricity can be "stored" in
the form of dispatch availability, and it is in the interest of
consumers for suppliers to produce when supplies are needed the most,
i.e., when (expected) prices are highest.  This is true even during a
stage 3 alert: Suppose, for example, that even greater problems are
foreseen for the ensuing summer.  Accordingly, this is not
"manipulation."  In addition, as explained in my paper in the current
issue of Regulation, (c) market prices under competition include a price
premium for the provision of optimal reliability.  These three
components of cost are important, with measurement requiring econometric
and other tools; and they perhaps are impossible to "verify" in
regulatory methodology.

    2. The ISO and FERC analyses confuse marginal and (average) variable
cost, average and marginal producers, average and marginal units of
production, etc.  Other errors are likely as well: The ISO, and, I
suspect, FERC implicitly assume away rising input prices as demand
conditions increase.  This would yield in perfectly competitive markets
a supply curve steeper and a market price higher than measured "marginal
cost."  Moreover, the ISO, and, again, FERC in all probability,
implicitly assume perfectly elastic supply at whatever input "costs"
happen to prevail.  That is simply incorrect: In the short run,
production capacity is limited, so that even if input prices do not rise
with demand, there is a "kink" in the competitive supply curve.  If
demand intersects supply above the kinkthat is, if the amount demanded
exceeds production capacity at the "marginal cost" that the regulators
measureprice will exceed that "marginal cost."  Think about the
wholesale market for wheat, for example, which is close to a perfectly
competitive market.  In the face of bad weather in some agricultural
regions, the market price of wheat will rise, even though "costs" for
the remaining wheat farmerslabor, insecticide, land rents, fuel, water,
etc.will not have changed, and indeed might actually fall because of
reduced demand from farmers affected by the adverse weather.

     3. The ISO report admits explicitly that it does not know market
marginal cost, because it complains that some suppliers bidding above
the "soft cap" have not submitted cost data.  Moreover, marginal cost
properly defined is for the market as a whole, rather than just for the
producers who bid in California.  This means, quite literally, even
apart from all of the other problems, that the conclusion of
"uncompetitiveness" or "unreasonableness" cannot follow from the ISO
analysis or from what is likely to be the underlying FERC analysis,
because price will equal the marginal cost (defined correctly) of the
last unit produced in a competitive market.  In short: The sort of
analysis offered by the ISO and FERC cannot even in principle determine
competitive prices because the regulators lack data for marginal cost
for the marginal producer.

     4. The implicit definition of consumer welfare ("wellbeing") used
by most regulators is incorrect.  Consumers are interested not in low
electricity prices alone, but in a maximization of the value of the
entire consumption basket.  Average electricity prices lower than
marginal cost reduce consumer welfare by engendering resource use in the
electricity sector that is too big, and thus prices in other sectors
that are too high.  Moreover, price controls of whatever form must
create shortages even in markets with inelastic demand, and the true
price with a shortage is unambiguously higher than the price that would
clear the market in the absence of the controls.  (Think about the
"cost" of electricity during a blackout.)  This is easy to demonstrate
analytically.
     The ISO and FERC clearly are shunting aside the effects of price
controls masquerading as "refunds" upon long run expectations and other
parameters, which easily can be shown to make consumers worse off over
time.  Part of this problem is the creation of an asymmetric
distribution of expected returns, since suppliers clearly will not be
subsidized when prices are low.

     5. The ISO errs in its assumption that bids above the soft cap
reflect higher capital costs.  I would not be surprised if FERC is
making a similar error.  Capital costs, having been sunk, cannot affect
short run pricing; the differing engineering efficiencies of generating
units will be reflected in their market values, and thus the opportunity
cost of production.  This means, in a strict textbook sense, that
different generating units cannot have costs that differ, as
counterintuitive as that sounds.  More efficient units will have lower
fuel costs (for example), but higher opportunity costs for employment of
the capital assets.  I would make this point in a report, simply to
erode the credibility of the regulators, but I would not push it too
hard, since some people cannot understand it.

     6. The Report ignores the market power of the respective regulatory
bodies as monopsonists under the implicit system of price controls and
refunds.

     7. The definition of "market power" is entirely obscure, except
implicitly as bids higher than marginal cost as measured by the
regulators.  This is incorrect, and I think that this needs to be
explained clearly.



     I am sure that other errors and omissions will emerge as the
relevant documents and arguments are reviewed.  Anyway, I'd like to
prepare such a report.  I would have to see precisely what it is that
must be refuted, but my guess is that this is about a two-week project
in terms of required analytic and writing time.  An effort to measure
marginal cost correctly would be a good deal more involved and
time-consuming, and my sense is that at this stage a refutation is more
important.  If it is a two-week project, then I think that $15,000 would
be reasonable.

    Looking forward to hearing from you.  Ben