Transcript Wolak Presentation

Managing Market Power
Frank A. Wolak
Department of Economics
Stanford University
Stanford, CA 94305-6072
[email protected]
http://www.stanford.edu/~wolak
Chairman, Market Surveillance Committee
California ISO
Outline of Talk
• Market power mitigation is part of general market design
problem
• What is market design problem?
• Major dimensions of market design problem
–
–
–
–
Regulation versus Competition (Focus on this issue)
Private versus Public Ownership
Market rules
Market structure
• When does market mechanism versus administrative procedure
set price that come closest to achieving market designer’s goal?
– Mitigation applied when administrative procedure is superior
• Choice between two imperfect worlds
– Imperfect market and imperfect regulatory process
– If perfect regulation existed there would be no need for markets
• Apply above logic to assessing ERCOT market power mitigation
mechanism
Conceptual Framework
• Technology of electricity production and delivery given
– How many units of input energy needed to produce one MWh
– How many MWh can be delivered across a transmission or distribution
line
• Market design challenge—How to cause profit-maximizing
producers to supply electricity to achieve
– Technically efficiency = produce the maximum amount of output for a
given quantity of inputs—capital, labor, input energy, and materials
– Allocative efficiency = produce fixed amount of output at least cost
given input prices
– Output prices that only recover minimum costs of production yet
maintains long-term financial viability of industry
Conceptual Framework
• Suggested market design goal of a industry (or
market) design process is to achieve
– Lowest possible retail prices consistent with longterm financial viability of industry
• Fundamental market design challenge
– It is not in the unilateral financial interest of firms
to achieve this goal
– Does not serve the financial interests of their
shareholders
• Does not maximize shareholder value
Market Design Problem
• Maximize market designer’s payoff function (which
depends on market outcomes) by setting
– Number and size of market participants
– Rules for determining revenues each firm receives
• Subject to constraints that all market participants will
choose their strategies to maximize payoffs given
rules set by market designer (Incentive constraints)
• Market mechanism may not always maximize market
designer’s payoff function
– Market power is major reason
– For extreme cases of market power regulation is superior
What is Market Power?
• Ability of a firm to increase the market price and profit from
this price increase
• In all environments (market or regulated), privately-owned
firms continually exercise market power
– Management has fiduciary responsibility to their shareholders to
exploit all profitable legal business opportunities (exercise all
available unilateral market power)
• Publicly-owned firms have an incentive to exercise market
power in a wholesale market
– Serve interests of captive customers
• Question is not whether or not firms exercise market power
– Question is when does the exercise of market power cause significant
harm to consumers
• Market design problem is concerned with setting up
incentives for market participants to produce and price in an
economically efficient manner
Adam Smith on Market Design
• “It is not from the benevolence of the
butcher, the brewer, or the baker, that we
expect our dinner, but from their regard to
their own interest. We address ourselves, not
to their humanity but to their self-love, and
never talk to them of our necessities but of
their advantages.”
The Wealth of Nations, Book I Chapter II
Regulation is Imperfect
• Major problem with regulation
– Firm usually knows its technological capabilities and the
demand that it faces better than the regulator
– This leads to dispute between firm and regulator over
minimum cost mode to serve demand firm faces
– Regulator can never know minimum cost of providing service
• Regulator can know incurred or actual cost of providing service
– There are laws against confiscating regulated firm’s assets
• Impossible to tell difference between regulator setting
– Output prices that confiscate firm’s assets
– Output prices that provide strong incentives for least-cost operation
– Long history of legal disputes in US that attempt to define
process for setting prices that do not confiscate firm’s assets
– Firm understands value of superior information about its
demand and technology in regulatory price-setting process
• Regulated firm exercises market power by exploiting informational
asymmetry between firm and regulator
Regulation versus Competition
• Benefits of competition
– There are no laws against a firm’s competitors confiscating the
firm’s assets through their output and pricing decisions
• Any firm that is unable to cover its costs at the price set by market will
exit industry
• Nature of competition among firms leads high cost firms to exit the
industry and be replaced by lower cost firms
– Contrary to regulated regime, no need to determine if a firm’s
incurred production costs are the least-cost mode of
production
• If market is competitive, then any firm that is able to remain in
business must be producing at or close to minimum cost
– Possibility of exit from industry provides strong incentives for
minimum cost production under competition
Markets are Imperfect
(Particularly for electricity)
• Costs
– Firms in a competitive market have little incentive to pass on
cost reductions to consumers in the form of lower prices
– Firms exercise all available unilateral market power
• Same as serving fiduciary responsibility to shareholders
• Actions can set prices far in excess of competitive levels
• Existing firms may takes actions to prevent entry by new firms
– Competitive markets make it virtually impossible to
maintain cross-subsidies in prices across services and/or
consumers.
Regulation versus Competition
(Concluding Comments)
• When minimum cost of providing service is known, little reason
to run a market for service
– Cost-of-service regulation can be used to set price
– It is cheaper for consumers than running a market for the service
• When minimum cost of providing service is unknown, run a
market to learn this cost
– Competitive markets provide strong incentives for minimum cost
production
• Strong incentives for firms to find this minimum cost
• Explains David Patton’s observation from yesterday that suppliers bid below
regulated marginal costs
– Not necessarily strong incentives to pass-on lower costs in lower prices-market power problems
• Unless potential for significant cost reductions exist, introducing
competition makes little sense
– Fortunately in virtually all industries this is the case
Regulation versus Competition
• If market participants don’t change their actions
under competitive market versus regulated
regime, then a market is more expensive for
consumers
• Market design is about “getting incentives
right”
– For competition to be superior to regulation, some
market participants must change their behavior
– Regulation pays costs, but markets pay marketclearing prices
Lessons for Designing Mitigation
Mechanisms
• If mitigate too much can destroy incentives for least-cost
production by suppliers
– Suppliers take actions to maximize profits by exploiting mitigation
mechanism
– When market can be relied upon to set prices this will lead to higher
prices
• If mitigate too much in spot market can destroy incentives for
load-serving entities to hedge risk of high spot prices
– Stronger argument against mitigation of system-wide market power
versus local market power
– Signing a forward contract with supplier that knows it will be needed to
provide local energy will not result in a lower price unless this supplier
faces competition to supply this energy at some time horizon up to
delivery
• Local market power in San Francisco
• Prospective mitigation mechanism provides transparency in opportunity cost
of forward contract for local supplier
Lessons for Designing Mitigation
Mechanisms
• Mitigation procedure should be easy for market
participants to verify for themselves
– Enables them to take actions to avoid being
mitigated
– Argues in favor in simple process used to determine
when a supplier is mitigated
• If mitigate too little can allow significant harm
to occur
– Important to understand cost of failing to mitigate
– Compare this to net benefit of mitigating
Market Mitigation Questions
• What are system conditions when suppliers possess significant enough local
market power that market mechanisms cannot be relied upon to set prices?
– Element Competitiveness Index (ECI) attempts to determine these conditions
– Advantage is that it is prospective in the sense that sets a standard that depends
on system conditions
– Disadvantage is that is very difficult to know what the right value of ECI should
be for making this determination
• Is ECI monotone in degree of competitiveness?
• Does ECI = X, mean the same thing in terms of competitiveness on all interfaces?
• Is critical value of ECI invariant to forward contracting levels and market rules?
• Proposed solution—Use existing zonal boundaries as competitive constraints
during first year of TNT market
– Compute ECI over first year of market operation and use this information to set
critical levels of ECI for determining whether an interface in competitive or not
Market Mitigation Questions
• Mitigation is a constraint on market participant
behavior that is relevant in real-time market
– Will create incentives for one side of market to avoid dayahead market if don’t impose mitigation in day-ahead market
• Create divergence between day-ahead and real-time market prices
– LMPM constraints are not different mathematically from
other operating constraints
• Certain amount of energy or ancillary services capacity required
from collection of units
• With virtual bidding at all nodes in network don’t need
to impose mitigation in day-ahead if it is real-time
market
• Not best solution during initial stages of market
• Relies on high level of sophistication of market participants
• Can have severe reliability concerns—See California
Market Mitigation Questions
• What is appropriate bid price for mitigated units?
– Minimum cost marginal cost, but this is unknown, for reasons discussed
earlier
– Using generic technology based bid prices introduces market
inefficiencies
• If minimum cost at location is $30/MWh and generic technology bid is
$50/MWh this can distort all LMPs in system and certainly LMP at node
– Different from zonal world, with congestion bids at many locations can
influence LMPs
• Reference bids also introduce inefficiencies for same reasons
– They do not equal minimum marginal cost of mitigated supplier
• These solutions can also create leveraging opportunities for supplier with
mitigated units
– Increase price earned by non-mitigated units by bidding to let mitigated unit set
price
• Preferred solution--verified marginal costs
• Spot market input fuel price (Henry Hub gas + transportation) and heat rate
plus O&M costs
• This regulated solution, so should regulate as well as you can
Market Mitigation Questions
• When and where should mitigation take place?
– Local market power mitigation in day-ahead and real-time market for energy and
ancillary services
• System-wide mitigation
– If load-serving entities are adequately hedged, this is likely to be unnecessary
• Cap contracts as resource adequacy and spot price hedging mechanism
• Cap contract pays load-serving entity Max(0,P(spot)-P(contract))*Q(contract) in
exchange for up-front payment to generation unit owner
– Loads more likely to be hedged if system-wide does not exist
• Dulls price signals for demand-side involvement
• Price $1/MWh too high for 400 hours per year, versus a price that is $400/MWh too
high for 1 hour per year
• Former is more difficult to detect, but just as harmful to consumers
• Former provide little incentive for investment in demand responsive technologies
• Suggested belts and suspenders solution
– Phase out system-wide market power mitigation after first year of market
operation
– Replace with “guardrails for competition” approach described below
Market Mitigation Questions
• Preference for very stringent local market power mitigation with limited
system-wide mitigation
– Little hope for competitive entry or demand response even in longer term to
solve many local market power problems—Power plants in San Francisco
• Use guardrails for competition approach for system-wide protection
• Compare 12-month rolling average actual price to 12-month rolling average
of competitive benchmark price
– Take rolling average of hourly market prices over entire 12-month period and
compare this to average hourly competitive benchmark price over same 12month period
– If difference in P(actual) and P(benchmark) exceeds some critical value then
automatic regulatory intervention occurs to protect consumers
• Requires no hour-to-hour regulatory intervention by ERCOT
– Can set high bid cap or price cap and therefore allow hourly price signals
• Consumers protected from excessive market power
• Recommended level--$5/MWh difference between 12-month average
P(actual) - P(benchmark)
• This would have not triggered regulatory intervention until June of 2000 in
California
Guardrails for Competitive Market
• Recommended intervention if index is exceeded
– All market participants must submit cost-based bids and be paid the
resulting market-clearing price
– Any unit earning insufficient revenues to cover total costs under this
scheme must cost-justify its annual cost shortfall to regulator
– Payment scheme must be sufficiently unattractive to generation unit
owners so that they do all they can to avoid triggering its imposition
• This scheme creates a self-regulating market
– Generators want to work to fix market rather continue to exercise
unilateral market power
– Prevents a California market meltdown yet still provides hourly price
signals needed to
• Simulate development of price-responsive demand
• Provide incentives for load-serving entities to hedge spot price risk
– Goal of setting this compensation scheme is to provide strong incentives
for generators to avoid implementing it
Market Mitigation Questions
• Bid mitigation in RUC?
– Integrate RUC constraints into day-ahead market
• Including demand forecast constraints
– Mitigation takes place within context of day-ahead market
– With sequential RUC mitigation in day-ahead market should occur with
respect to ERCOT forecast of load, not bid-in load
• If run sequential RUC can either make start-up and minimum
load bids
– Cost-based on verifiable engineering criteria and spot market prices
– Freeze bids for month, six month or year
• CRR mitigation is largely unnecessary if
– Allocate CRRs to load-serving entities (LSEs) accounting for embedded
generation owned by LSE
– Allow loads to request CRR configuration they desire
– Duration of initial CRR release relatively short to preserve need for
ERCOT and PUCT to revisit allocation process given market outcomes
Market Mitigation Questions
• To the greatest extent possible all mitigation should be
prospective and automatic
– Market monitor should have very little discretion to mitigate bids ex post
– Much easier not to give money to a market participant, than it is to take
money back
• Ask California
– Write into protocols prospective procedure for market monitor to
intervene in market
• Increases transparency of market
– Intervention dulls incentives for investment in technology necessary to
solve problems
• Forward contracting
• Demand-side involvement
• Transmission upgrades
Market Mitigation Questions
• Procedure should have listed set of actions
market monitor must first take before ex post
intervention can occur
– Overarching Criterion should be that significant
harm to system reliability and market efficiency that
cannot be solved by any other means is occurring
• See “Managing Market Power in Electricity”
available from web-site for more on this issue
– http://www.stanford.edu/~wolak
Questions, or Comments?
Cost-Based Dispatch Market
• Costs
–
–
–
–
Increase incentive to put fixed costs into variable costs
Limited incentive to reduce input fuel purchase costs
Limited incentive for suppliers to invest to reduce production costs
Greater incentives for inefficient dispatch
• Withhold lower cost units to allow high cost units to set market price
• Benefits
– Limit number of ways suppliers can exercise market power
• Can only bid higher if can “cost justify” higher bid
– Reduce volatility of wholesale prices and congestion prices
• Can limit incentives for demand-side involvement in market
• Can reduce economic justification for transmission upgrades
• Both of which limit ability of suppliers to exercise market power in bid-based market
• Summary
– Potential for increased average cost of real-time system operation to limit risk of
enormous market power in spot market
Pricing Under Competition Versus Regulation
Price
TRCompetition = A + B
MC
TRRegulation = A
PCompetition
B
A
Qd
Quantity
Pricing Under Competition versus Regulation
Price
TRCompetition = A + C
MCRegulated
TRRegulation = B + C
B
PCompetition
MCCompetition
A
C
QCompetition
Quantity
Real-Time Pricing to Exercise Buyer Power
P(RTP,t) = real-tme price announced to consumers in hour t
P(W,t) = wholesale purchase price in hour t
Q(t) = quantity of energy RTP consumer purchases in hour t
Retailer makes commitment to earn zero profit on RTP
customers
P(RTP,t)Q(t) = P(W,t)Q(t)
During high load hours P(RTP,t) > P(W,t)
During low load hours P(RTP,t) < P(W,t)
Peak Period Buyer Market Power
Supply Bids
P0
Savings From Exercise of Market Power
by Buyer Made Possible by Real-Time
Pricing Demand Reduction
PM
QM Q0
Off-Peak Period Buyer Market Power
Supply Bids
Cost Increase from
Exercise of Market Power by
Buyer Made Possible by
Real-Time Pricing
PM
P0
Q0 QM
Real-Time Pricing Allows Retailers to Obtain Lower
Forward Contract Prices
Generators will recognize that effects shown on previous slides
will operate to reduce spot prices and demand
Spot market prices will be lower in future than they would be in the
absence of significant real-time pricing
Lower future spot prices creates lower opportunity cost to a generator
signing a forward contract
Generators more likely to sign forward contracts at lower prices than they
would in the absence of a large commitment to real-time pricing
The Role of Price Volatility
• Wholesale price volatility makes potential benefits of realtime pricing greater
• Real-time pricing encourages demand flexibility across
hours in the day
• In California, total annual energy demand in 2000 divided
by number of hours in year is ~27,000 MW
• Total in-state capacity is 44,000 MW and 12,000 MW
import capacity
– Price-responsive demand makes market power problem goes away
– Real-time pricing accomplishes this
• Encourages development of renewable and distributed
generation technologies
Real-time pricing contracts
• All England and Wales retail customers have option to
purchase hourly consumption according to hourly pool
price plus transmission charge
• Many large industrial customers purchase according to this
pool price contract
• “Estimating the Customer-Level Demand for Electricity
Under Real-Time Market Prices” Patrick and Wolak
• Estimate half-hourly price responsiveness of a sample of
large industrial and commercial customers in England and
Wales
– Significant price response from all classes of industrial customers-water suppliers, industrial process plants, retail stores
– Even with a small fraction of these customers bidding into demand
side of pool, market power can be mitigated.
Figure 2 (a)
0.30
0.25
0.20
P
r
o
p
o
r
t
i
o
n
0.15
0.10
0.05
0.00
15
20
25
30
Quantity-Weighted Average Price
35
40
Figure 2 (d)
0.175
0.150
0.125
P
r
o
p
o
r
t
i
o
n
0.100
0.075
0.050
0.025
0.000
15
20
25
30
Quantity-Weighted Average Price
35
40
Implications for Re-structuring
• For consumers to benefit from a
competitive market they must face to realtime hourly price signals
• In competitive market a firm must make
profits on each customer
• Regulated firm only needs to make profits
across all customers
• All customers with same cost to serve will
face same price in competitive market
– Need not be true in regulated regime
Implications for Re-structuring
• Inability to cross-subsidize under competition
– Requires greater attention to protecting lowincome consumers from high prices that may
impoverish them
– Regulator may need to require explicit subsidies
• Regulator must provide significantly more
information to consumers to help them
protect themselves
– Emphasize importance of hedging spot price risk
– Information on load-shifting technologies