California
Dreaming: Will FERC's Plan Work?
February 15, 2001
By Howard Spinner
And even if it does, it
probably won't.
Deregulation
has derailed in California, while back in Washington, the feds
have formed a plan to put it back on the tracks. Yet success remains far
from assured. After all, it can take a long time to pay off an $11 billion
mistake, even if competition brings prices down.
Meanwhile, according to the
Federal Energy Regulatory Commission and its staff, the train wreck that
has snagged the state's investor-owned utilities stems in large part from
three factors: (1) a lack of investment in new generation, (2) too much
reliance on the spot market for power supplies, and (3) no adequate way
for consumers to tailor their demand to respond to wholesale prices. The
wholesale market was found to be flawed. To remedy the situation, the
FERC staff proposed certain policy options1 and the commissioners
voted to adopt them.2 One of these policies—to "remedy
the over-reliance on spot market purchases"-seems a case of the pot calling
the kettle black.
But I digress. The task here
is not to speculate, but to test the sufficiency of the proposed reforms,
employing the basic tools of economic and financial analysis. At a minimum,
let's inquire first whether California's experiment with deregulation
has indeed blown up the laboratory. Next, let's ask whether the alleged
deficiencies identified by the FERC are ones that can be rectified by
the policy options proposed by the staff. Is it possible that the alleged
deficiencies are not deficiencies at all, but instead represent real economic
costs-costs so inherently insurmountable that they led the industry for
many years to embrace the very structure that California abandoned with
the passage of Assembly Bill 1890 and its mandate for deregulation back
in 1996?
Paying Down
an $11 Billion Mortgage
As last year drew to a close,
it was widely reported that California's two largest investor-owned electric
utilities, Pacific Gas & Electric (PG&E) and Southern California Edison
(SCE), claimed to be on the brink of bankruptcy, having spent approximately
$11 billion more to procure power for their customers on the deregulated
wholesale market than they were allowed to collect from those customers.
And as this paper went to press, the deficit was expected to expand. Ironically,
a large portion of the $11 billion shortfall stemmed from payments made
to the new owners and operators of electric generation plants that the
IOUs once owned—before they sold off many power plants that ran on
fossil fuels. The new owners of these plants came up as the big winners
in California, capturing huge cash flows produced by wholesale prices
that the FERC eventually termed "unjust and unreasonable."3
If we view that $11 billion
shortfall as a leakage from what formerly had been a relatively closed
system consisting of California IOUs and their retail ratepayers, then
all it takes is a look at some basic numbers to relate a sense of how
serious the problem truly is.
In 1995, before the enactment
of California's landmark restructuring law, the retail electricity rates
for PG&E, SCE, and San Diego Gas & Electric Co. (SDG&E) each averaged
around 10 cents per kilowatt-hour ($100/MWh). Assume that for each 10-cent
rate, about 70 percent,4 or 7 cents per kilowatt-hour ($70/MWh),
represented the cost of generation, while the remaining 3 cents per kilowatt-hour
covered the cost of delivering that energy. Further, assume that the market
fixes identified by FERC are implemented immediately and in fact prove
successful, so that as a result, generation costs fall 10 percent below
what they would have been without AB 1890, and begin to produce savings
of $7 per megawatt-hour.5 The question then arises, how long
will it take customers to amortize $11 billion with savings of $7 per
megawatt-hour?
California's relevant retail
market buys approximately 180 million MWh annually and appears to have
grown at about 3 percent per year since 1995. Using these numbers and
assumptions—and a reader-supplied discount rate—one can see
that the $11 billion leakage represents a loss that will be difficult
to recover even if the problem is fixed right now. Given any reasonable
discount rate, if California consumers are asked to pay the $11 billion
IOU shortfall, it is difficult to see how (former) ratepayers will benefit
from California's electric industry restructuring.
The above analysis, though
admittedly simple, does point to the risk-return profile of moving away
from cost-of-service regulation for electric utilities, given the magnitude
of the possible increase in price in an open market. (Editor's Note:
At a savings of $7 per megawatt-hour, and at a discount rate of 6 percent,
customers buying a constant 180 million MWh each year-no annual growth
in demand-would take 12 to 13 years to amortize $11 billion.)
Now that we can better appreciate
the seriousness of the problem, even if we assume that the FERC hits a
home run, let's look at the proposed remedies in greater detail to see
if they are indeed likely to help, and help soon.
According to the FERC and its
staff, the California market is flawed due to a lack of new generation,
a lack of demand response during periods of high prices, and an over-reliance
on spot market purchases in wholesale markets. In its report of Nov. 1,
the FERC staff proposes certain policy options to remedy these flaws.
In general, two of the remedies fall within state jurisdiction. They are
designed to mitigate the demand/supply imbalance that has led to very
high market clearing prices. These two proposals would have California
adopt policies that "encourage and facilitate the investment in new generation"6
and "increase retail demand responsiveness to price."
The third proposal seeks to
remedy market defects allegedly caused by the over-reliance on spot market
purchases. In its final order of Dec. 15, the FERC terms the requirement
that California IOUs sell all of their generation into and buy all of
their energy needs from the California Power Exchange (PX) as "the most
serious flaw in the market design created by AB 1890 and the California
Commission's implementing orders."7 Below, we examine each
alleged flaw and associated proposed remedy in more detail.
Better Demand
Response?
The purpose of encouraging
the California Public Utilities Commission to promote policies that allow
consumers of electricity to respond to high prices is to lower California's
electricity bill by better balancing demand and supply during periods
that otherwise would exhibit high wholesale prices. Residential and small
commercial customers have never been subjected to high levels of prices
or price volatility on an hourly basis as have recently been observed
in the PX, but if they were, they would probably reduce demand. And since
retail customers of PG&E and SCE do not currently see the hourly transmission
and distribution costs that they impose on the IOU, it is logical that
retail price signals that pass along wholesale generation costs would
lead to significant demand reductions during high-cost hours. (And perhaps
result in a reduction in essential electric service provided to some of
the most vulnerable of California's citizens.)
Another result of sending more
accurate hourly price signals to consumers might be load shifting. Load
may be shifted from high- to low-cost hours within a particular day or
to an entirely new day, week, or season. Aggregate, system-wide electricity
consumption usually exhibits familiar patterns of seasonal and diurnal
demand. The likely success of load shifting as a cost mitigation tool
depends on many factors, primarily the amount of load that would have
to be shifted, the cost to customers of shifting that load and the impact
on prices that will now prevail during the periods into which the load
was shifted.
Consider Figure 1. Here we
have average hourly zonal energy prices for ZP26 during the summer of
2000.8 The good news is that if load shifting were economically
feasible and physically possible, it appears that consumption could have
been shifted to the wee hours of the morning to avoid high afternoon prices.
There are lower-cost hours available during the daily cycle. The bad news
is that even during the low-cost wee hours, average hourly prices approach
the average $70 per megawatt-hour price for generation that Californians
paid prior to industry restructuring.
This example points out a key
issue: How low are the load levels during the low-cost hours targeted
for shifting? Is there enough space on the load curve to accommodate more
service at the same low rate, without "chasing the peak around the clock?"
In order to gain insights here we must look at the loads. Consider three
related pieces of load data and, for some perspective, compare California's
results to like measures derived for PJM.
Figures 2 and 3 depict summer9
load duration curves for the California ISO and for PJM. These plots show
that California already exhibits a lessened tendency for "needle-peakness"
than does PJM. In fact, California's summer of 2000 load factor of 68.5
percent was 8 percent higher than PJM's summer of 1999 load factor of
63.5 percent. Stated differently, California had 92 hours last summer
in which load fell within 95 percent of system peak last summer. By contrast,
during the summer of 1999, PJM had just 21 such hours. Also, as shown
below in Figure 4, California's load patterns already have flattened considerably.
From 1999 to 2000, California's summer load factor improved by 11 percent-from
61.7 percent to 68.5 percent.
Look also at the average peak
day load factor during the five days containing the highest loads for
each ISO during the summer being analyzed. By this measure, PJM appears
to have a "flatter" load profile than California, with an average peak-day
load factor of 82.22 percent (for the summer of 1999) vs. California's
value of 79.21 percent for last summer. (From my own experience at my
former employer, I know that after almost 20 years of extensive deployment
of seasonal, time-of-day, and interruptible rates, by the early 1990s
Central Vermont Public Service Co. exhibited peak-day load factors in
the range of 90-93 percent.)
None of this is meant to throw
cold water on the idea of improving price signals. Rational prices always
make sense. No, the point here is that because prices are high around
the clock and, by some measures, California already is less "peaky" than
might be expected, transmitting wholesale price signals down to the retail
market might not solve the problem of high wholesale prices. Further,
while customers certainly would pare back on demand and thus mitigate
wholesale costs if they faced retail prices approaching $1, $2, or $5
per kilowatt-hour, the results might not prove helpful to society.
Less Reliance
on the Spot Market?
Recall that the FERC also cited
the mandate (imposed on the IOUs) to buy only from the PX as the "the
most serious flaw in the market design created by AB 1890 and the California
Commission's implementing orders." This is a statement that reaches out
for analysis.
At first blush, it seems far
more troublesome that, for a host of reasons, little has been added to
California's generation fleet in recent years even as load has grown.
According to recently published U.S. Census data, California's population
grew by about 4.1 million people, or 14 percent, during the 1990s. Generation
markets apparently are very tight. The lack of meaningful retail price
signals also sounds like a more serious flaw. Further, it remains to be
seen what kind of relief will be provided by allowing California IOUs
to contract forward for a greater, rather than a smaller, portion of their
load requirements. Potential relief depends on what those contracts will
look like as well as the analyst's frame of reference. As of this writing,
FERC is trying to bring generators and IOUs together to work out the details
and pricing of longer term, forward contracts.
Modern portfolio theory suggests
a relationship between forward prices and expected spot prices as set
forth below.10The theory suggests that if holding the commodity
in question poses positive systematic risk, futures (forward) prices must
be lower than expected spot prices. The posited relationship follows:
Fwd = E(P)t * [(1 + Rf) / (1+k)]t
where
Fwd = forward price
E(P)t = tomorrow's expected
spot price
Rf = risk free rate of return
k = return required to hold
the asset
t = holding period
Thus, the amount by which forward
prices will be reduced relative to today's expectation of tomorrow's spot
price [E(P)t] turns on the value of the bracketed term [(1 + Rf)/(1+k)],
or whether or not k is greater than Rf. Also note that, since k = Rf +
beta * (Rm - Rf), where Rm equals the market rate of return, Fwd
will be less than E(P)t for any positive beta asset. Thus, if California
electricity is indeed a positive beta asset, the forward price
contracts that IOUs shortly may enter will be at prices lower than today's
expectation of spot prices out in the future. Of course, how today's expectation
of future spot prices compares with historical spot prices might be impacted
greatly by FERC's proposed changes to the Cal PX market structure and
operations.
Rather than debating the potential
values of "Rf" and "k," what will probably turn out to be more important
to overall prices paid by Californians under any long-term contracts signed
by their IOUs is today's expectation of tomorrow's spot price. Policies
that lower the parties' expectations of future spot prices will lower
the cost assumed by Californians in accepting the risk of having the IOU
enter long-term contracts. Thus, if the FERC's proposed market reforms
succeed in cutting the expectations of spot prices, Californians will
benefit—not really because forward contracting is to be allowed,
but rather because market reforms reduce spot prices that are expected
to prevail in the future. This observation suggests that it is not the
over-reliance on spot market purchases that is causing the problem. Rather,
it is a poorly functioning spot market itself that is allowing persistently
high "market clearing" prices to prevail at levels that are far above
marginal production costs.
Encouraging
New Generation?
Depending on the ownership
pattern, more generation supply surely could help remedy California's
current and expected future ill-functioning spot market. The current California
spot market is marked by relatively few sellers and even fewer buyers.
That the buyers are not the ultimate consumers and do not have to pay
the cost of their purchases provides the opportunity for poor results.
California's wholesale power
prices climbed very high last summer—far above the prices paid the
year before. And any number of stakeholders will offer any number of reasons
for that. Higher natural gas prices, higher environmental costs, higher
loads, lower hydroelectric water flows, and a lack of generation—they
all are mentioned. Others have estimated the impact of these factors on
price and have concluded that generators have exhibited substantial market
power in raising prices far above competitive levels.
Figure 4 depicts California
ISO load duration curves for the past two summers. It shows that while
the ISO served about 6.0 percent more load last summer than the year before,
the actual peak loads were higher during the summer of 1999. That suggests
that the much higher prices observed this past summer stemmed in large
part from generators learning how to sell into the market in a way that
increases profits for one and all.
More generation certainly would
help dilute market shares. Yet it may prove difficult to impose such a
solution on the Californians themselves. Generation has been extremely
difficult to site in California and that problem likely will continue.
Given the state's past enthusiasm for demand-side management (DSM) and
its insistence on environmental protection, no one should be surprised.
According to data available
on the Web site of the U.S. Energy Information Administration (EIA),11
total electric utility DSM program costs in the United States hit $2.4
billion in 1995. PG&E, SCE, SDG&E and the Sacramento Municipal Utility
District accounted for $273 million, or 11.3 percent of that national
total. Yet the 1999 retail sales for these four utilities represent only
about 5 percent of total U.S. retail sales. Few would argue that, whatever
its merits, California in the years immediately prior to electric restructuring
seriously had embraced DSM. In fact, one might reasonably claim that for
several years prior to restructuring, it was public policy in California
that electricity should be provided not at minimum cost, but rather with
minimum plant. And if that indeed should describe California's
historic policy—some might say its current policy too—then the
current shortage of generation, if real, should not surprise anyone.
Just because Californians have
been reluctant to add generation plants in their communities does not
make them bad people. These facilities impose both costs and benefits
on the community, state, and region that host them. Not having these facilities
forgoes those costs and benefits. To the extent that the desire to minimize
the deployment of utility plants represents the collective will of people
who have reasonably evaluated these costs and benefits, Californians'
desire to minimize plant deployment should be accommodated rather than
condemned. The question becomes, how can resources best be allocated subject
to this politically imposed constraint? Ideally, this question should
be answered based on analysis from the consumers' perspective including
all costs and benefits, appropriately quantified.
Certainly this minimalist sentiment
places plant owners in an envious position. With generation priced at
market, plant owners have a good chance to capture economic rents. More
(market) power to them—literally. Success in California should encourage
power producers in other regions to learn how to bid in a competitive
wholesale power auction. Unfortunately, however, the experience gained
from California to date suggests some of the classic symptoms of a less
than perfectly competitive market.
Howard Spinner works as
a senior utilities specialist with the Division of Energy Regulation of
the Virginia State Corporation Commission in Richmond. He worked previously
for Central Vermont Public Service Corp., focusing on rates, load research,
and special contracts, and before that for AT&T Communications, in forecasting
and rate case support for intrastate interLATA services. Spinner's views
expressed here are his own. They do not necessarily reflect the views
of the Virginia State Corporation Commission or the Commission staff.
1 Staff Report to
the Federal Energy Commission on Western Markets and the Causes of the
Summer 2000 Price Abnormalities—Part 1, Nov. 1, 2000 (Staff Report),
p. 6-1.
2 Order Directing
Remedies For California Wholesale Electric Markets, Docket No EL00-95-000,
et al, 93 FERC ¶61,294. Dec. 15, 2000. Note that while FERC
recognizes that siting matters and retail demand response programs reside
within exclusive state jurisdiction, FERC urges state officials to take
certain actions in these areas.
3 Order Proposing
Remedies For California Wholesale Electric Markets, Docket No EL00-95-000,
et al, 93 FERC ¶61,121, Nov. 1, 2000, p. 3.
4 The pre-restructuring
average embedded generation cost component of California IOU rates were
about $74 per megawatt-hour. See FERC Order Dec. 15, 2000 at
p. 27.
5 For simplicity,
let's assume that the pre-restructuring estimated generation cost of
around 7 cents per kilowatt-hour would have prevailed; a 10 percent
savings represents about $7 per megawatt-hour.
6 FERC Staff Report,
p. 6-1.
7 FERC Order, Dec.
15, 2000, at p. 25.
8 Throughout this
paper, I define summer as calendar months June, July, August, and September.
These data can be found at the California ISO website.
9 For California,
I show the months of June-September, 2000. For PJM, I use June-September,
1999 in an effort to include hot weather impacts that simply did not
occur in the Northeast during summer 2000.
10 This discussion
is taken from "Investments," by Bodie, Kane and Marcus, Richard D. Irwin,
Inc., 1989, pp. 648 - 651.
11 See www.eia.doe.gov/cneaf/electricity/dsm/t22.txt.
Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.
Public
Utilities Fortnightly & Spark
