Transcript Marginal Cost i

Testing Strategic Models of Firm
Behavior in Restructured
Electricity Markets:
A Case Study of ERCOT
Ali Hortacsu, University of Chicago
Steve Puller, Texas A&M
Motivation
• Empirical auction literature
– Bid data + equilibrium model  valuation
• “New Empirical IO”
– Eqbm (p,q) data + demand elasticity + behavioral
assumption  MC
• Can equilibrium models be tested?
– Laboratory experiments
• Electricity markets are a great place to study firm
pricing behavior
• This paper measures deviations from theoretical
benchmark & explores reasons
Texas Electricity Market
• Largest electric grid control area in U.S.
(ERCOT)
• Market opened August 2001
• Incumbents
– Implicit contracts to serve non-switching customers at
regulated price
• Various merchant generators
Electricity Market Mechanics
• Forward contracting
– Generators contract w/ buyers beforehand for a delivery
quantity and price
– Day before production: fixed quantities of supply and
demand are scheduled w/ grid operator
– (Generators may be net short or long on their contract
quantity)
• Spot (balancing) market
– Centralized market to balance realized demand with
scheduled supply
– Generators submit “supply functions” to increase or
decrease production from day-ahead schedule
Balancing Energy Market
• Spot market run in “real-time” to balance supply
(generation) and demand (load)
– Adjusts for demand and cost shocks (e.g. weather, plant
outage)
• Approx 2-5% of energy traded (“up” and “down”)
– “up”  bidding price to receive to produce more
– “down”  bidding price to pay to produce less
• Uniform-price auction using hourly portfolio bids that
clear every 15-minute interval
• Bids: monotonic step functions with up to 40 “elbow
points” (20 up and 20 down)
• Market separated into zones if transmission lines
congested – we focus on uncongested hours
Who are the Players?
Generator
% of Installed Capacity
TXU Electric
24
Reliant Energy
18
City of San Antonio Public Service
8
Central Power & Light
7
City of Austin
6
Calpine
5
Lower Colorado River Authority
4
Lamar Power Partners
4
Guadalupe Power Partners
2
West Texas Utilities
2
Midlothian Energy
2
Dow Chemical
1
Brazos Electric Power Coop
1
Others
16
Incentives to Exercise Market Power
• Suppose no further contract obligations
upon entering balancing market
• INCremental demand periods
– Bid above MC to raise revenue on
inframarginal sales
– Just “monopolist on residual demand”
• DECremental demand periods
– Bid below MC to reduce output
– Make yourself “short” but drive down the
price of buying your short position
(monopsony)
Price
Si (p)
MRi(p)
E
B
C
D
MCi(q)
A
RDi(p)
QCi
Quantity
Methods to Test Expected Profit
Maximizing Behavior
Difficult to compare actual to ex-ante optimal bids
–
Wolak (2000,2001)  solving ex-ante optimal bid strategy
(under equilibrium beliefs about uncertainty) is computationally
difficult
Options
1) Restrict economic environment so ex-post optimal =
ex-ante optimal
•
Intuitively, uncertainty and private information shift
RD in parallel fashion
2) Check (local) optimality of observed bids (Wolak,
2001)
• Do bids violate F.O.C. of Eε[π(p,ε)]?
3) Can simple trading rules improve upon realized
profits?
Overview of Model
• Setup
– Static game, N firms
– Marginal Costi is public information
– Contract quantity (QCi) and price (PCi) are
private information
– Generators bid supply functions Si(p,QCi)
• Sources of uncertainty
– Total demand D(p) stochastic
– Rivals’ bids S-i(QC-i)
Market clearing price (pc) is uncertain
(application of Wilson 1979 share auction)
Sample Genscape Interface
Overview of Model (contd)
H ( p, Si ( p))  Pr{ p c  p Si ( p)}
 Pr{ S j ( p, QC j )  Si ( p)  D( p) Si ( p); FQC j (.), FD (.)}
ji
Firms choose supply function to maximize expected profits:
p
max  { pSi ( p)  Ci ( Si ( p))  ( p  PCi )QCi }dH ( p, Si ( p))
Si ( p )
p
If H(.) is differentiable, necessary condition for optimality:
*
H
(
p
,
S
S
i ( p))
*
*
p  Ci( Si ( p))  ( Si ( p)  QCi )
H p ( p, Si* ( p))
Overview of Model (contd)
CLAIM: If we restrict the class of supply functions:
Si ( p)   i ( p)  i QCi
then (ex ante) equilibrium bids are ex post best responses:
RDi ( p)  QCi
*
p  Ci( Si ( p)) 
RDi( p)
where
RDi ( p)  D( p) 
 S ( p)
ji
j
Computing Ex Post Optimal Bids
Ex post best response is Bayesian Nash Eqbm
 Uncertainty shifts residual demand parallel
in & out
 Can trace out ex post optimal / equilibrium
bids
Unknown


*
S
i ( p)  QCi
*
p  MCi ( Si ( p)) 

RDi( p)
Unknown
("inverse elasticity rule")
Data (Sept 2001 thru July 2002)
• Bids
– Hourly firm-level bids
• Demand in balancing market – assumed perfectly
inelastic
• Marginal Costs for each operating fossil fuel unit
• Fuel efficiency – average “heat rates”
• Fuel costs – daily natural gas spot prices & monthly
average coal spot prices
• Variable O&M
• SO2 permit costs
– Each unit’s daily capacity & day-ahead schedule
Measuring Marginal Cost in Balancing Market
• Use coal and gas-fired generating units that are “on” and
the daily capacity declaration
• Calculate how much generation from those units is already
scheduled == Day-Ahead Schedule
Price Residual MC Total MC
Day-Ahead
Schedule
MW
Reliant (biggest seller) Example
TXU (2nd biggest seller) Example
Guadalupe (small seller) Example
Calculating Deviation from
Optimal Producer Surplus
$
Optimal
Profit  P EPO  qiEPO  TC(qiEPO )  ( P EPO  PC)QCi
Actual
Profit  P BAL  qiBAL  TC(qiBAL )  ( P BAL  PC)QCi
Avoid
BAL
BAL
Profit  PAvoid
 0  TC(0)  ( PAvoid
 PC)QCi
(1) Foregone Profits   Optimal   Actual
 Actual   Avoid
(2) Percent Achieved  Optimal

  Avoid
Measures of Foregone Profits
.9
1
Percent of Potential Gains from Not Bidding
Bryan
TXU
BP
Mirant
.3
.4
.5
.6
.7
.8
Reliant
.2
Austin
0
.1
WTU
Lamar
Calpine
CPL
0
100
200
300
400
500
Absolute Value of Optimal Output
600
700
1
Learning by Larger Players?
Reliant
TXU
TXU
Reliant
Reliant
Reliant
TXU
Reliant
CPL
.5
Austin
Reliant
WTU
TXU
Austin
TXU
Austin
Calpine
Calpine
TXU
Calpine
Austin
TXU
Lamar
WTU
Calpine
0
WTU
Lamar
Calpine
Lamar
WTU
Calpine
Lamar
Calpine
Austin
WTU
Lamar
Lamar
CPL
WTU
Austin
Calpine
Lamar
TXU
Austin
WTU
CPL
WTU
CPL
Austin
Calpine
TXU
Calpine
WTU
WTU
CPL
Lamar
CPL
Austin
Austin
Lamar
TXU
-.5
Lamar
2
4
6
8
10
Months Between Aug 2001 and July 2002
12
Testing Expected Profit Maximizing
Behavior
1) Restrict economic environment so ex-post
optimal = ex-ante optimal
2) No restrictions -- Check (local) optimality of
observed bids (Wolak, 2001)
3) Can simple trading rules improve upon
realized profits?
Generator’s Ex-Ante Problem
• Max Eε[π(p,ε)]
s.t. (1) monotonicity of bids
(2) transmission congestion
(3) physical operating contraints
• Restrict our sample  ignore constraints
• Wolak test for (local) optimality:
– Ho: Each bidpoint chosen optimally
– Changing the price of each (pk,qk) will not
incrementally increase profits
Reliant (biggest seller) Example
Guadalupe (small seller) Example
Test for (Local) Optimality of Bids
Choose bid vector   ( p1... pK )
 (,  )  RD ( p( , ),  ) p( , )  C ( S ( p( , )))
 ( p( , )  PC )QC
 (,  ) 
  0
E 
 pk 
Moments for GMM:
 (,  ) 
S  p
  RD() p  ( RD ( p )  QC )  C ( S ( p )) 
pk
p  pk

p
" MarginalRe venue ( p )  MarginalCo st ( p )
"
pk
Test for (Local) Optimality of Bids
Firm
Reliant
TXU
Guadalupe
J-stat
0.131
0.302
0.005
d.o.f.
9
5
2
• Fail to reject (even for Guadalupe!)
• Test is lower power in our setting
– Future work: use *quantity* moments
p-value
0.99
0.99
0.99
“Naïve Best Reply Test” of
Optimality
• Bidders can see aggregate bids with a few
day lag
• Simple trading rule: use bid data from t-3,
assume rivals don’t change bids, and find ex
post optimal bids (under parallel shift
assumption)
• Does this outperform actual bidding?
How Much Does Trading Rule
Increase Profits?
Bryan
$200/hr18
Calpine
$1,325/hr18
City of Austin
$1,129/hr18
Reliant
TXU
$957/hr18
$1,770/hr18
Preliminary
Learning in Second Year?
Sep02Sep01-Jul02 May03
Reliant/Texas Genco
82%
27%
TXU
53%
77%
Bryan
44%
56%
Calpine
34%
44%
City of Austin
27%
28%
Fair bit of month to month variability by firm.
Note: Second year excludes Aug’02, Dec’02 (data not clean yet) and Feb ’03 (“weather incident”)
What the Traders Say about Suboptimal Bidding
1. Lack of sophistication at beginning of market
•
Some firms’ bidders have no trading experience; are
employees brought over from generation &
distribution
2. Heuristics
•
•
Most don’t think in terms of “residual demand”
Rival supply not entirely transparent b/c
•
•
•
Eqbm mapping of rival costs to bids too sophisticated
Some firms do not use lagged aggregate bid data
Bid in a markup & have guess where price will be
3. Newer generators
•
If a unit has debt to pay off, bidders follow a formula
of % markup to add
What the Traders Say (contd)
4. TXU
•
•
“old school” – would prefer to serve it’s customers
with own expensive generation rather than buy
cheaper power from market
Anecdotal evidence that relying more on market in 2nd
year of market
5. Small players (e.g. munis)
•
•
“scared of market” – afraid of being short w/ high
prices
Don’t want to bid extra capacity into market because
they want extra capacity available in case a unit goes
down
Counterfactual Welfare Calculations
(Not Yet Completed)
• Productive inefficiencies under alternative bidding
(1) Actual vs. Competitive (Vickrey multiunit)
(2) Actual vs. Unilateral Best-Reply (Uniform-Price)
(3) Actual vs. "Large Unilateral" and "Small
Competitive"
Conclusion
• Electricity markets are a great “field”
setting to understand firm behavior under
uncertainty and private information
• Stakes appear to matter in strategic
sophistication
• Both sophistication (“market power”) and
lack of sophistication (“avoid the market”)
contribute to inefficiency in this market
The End
"Bid-Ask" Spread For Largest Sellers
50
Reliant
$/MWh
40
Calpine
TXU
30
20
10
0
Jul-01
Sep-01
Nov-01
Dec-01
Feb-02
Apr-02
May-02
Difference betw een INC and DEC bid prices at q=0, excluding hockey stick hours.
Jul-02
Sep-02
Avg. Number of Bid Points
Average Bid Points Used Per Period
For Three Largest Sellers
Reliant
25
Calpine
20
TXU
15
10
5
0
Sep01
Oct01
Nov01
Dec01
Jan02
Feb02
Mar02
Apr02
May02
Jun- Jul-02
02
Dispersion of “Money on the Table”
Reliant
Density
.000146
0
-3000
0
5000
10000
15000
xpo_minus_ignore
Kernel Density Estimate
20000
25000
Dispersion of “Money on the Table”
Reliant
TXU
Density
.000246
Density
.000146
0
0
-3000
0
5000
10000
15000
xpo_minus_ignore
20000
25000
-3000
0
Kernel Density Estimate
5000
10000
15000
xpo_minus_ignore
20000
25000
Kernel Density Estimate
Calpine
Bryan
Density
.001846
Density
.000353
0
0
-3000
0
5000
10000
15000
xpo_minus_ignore
Kernel Density Estimate
20000
25000
-3000
0
5000
10000
15000
xpo_minus_ignore
Kernel Density Estimate
20000
25000
.0004
Quantity Traded in Balancing Market
.0002
0
.0001
Density
.0003
Mean = -257
Stdev = 1035
Min = -3700
25th Pctile = -964
75th Pctile = 390
Max = 2713
-4000
-2000
0
Net MW in BES Market
2000
4000
Sample: Sept 2001-July 2002, 6:00-6:15pm, weekdays, no transmission congestion
Zones in ERCOT 2002
Source: Public Utility Commission of Texas, MOD Annual Report (2003)
Sample Bidding Interface
Do We Expect to See Optimal Bidding?
• First year of market
– Some traders experienced while others brought over
from generation and transmission sectors
• Many bidding & optimization decisions being
made
• Real-time information?
– Frequency charts & Genscape sensor data  rival costs
– Aggregate bid stacks with 2-3 day lag  “adaptive
best-response” bidding?
• Is there enough $$ at stake in balancing market?
– Several hundred to several thousand per hour
• “Bounded rationality”
Smaller Players
• Appear to bid to “withhold capacity” to avoid the
balancing market
 productive inefficiencies
• Not unilateral market power because
markups/markdowns are too large given their
small inframarginal sales
• Policy implications:
– Fixed costs to participation?
– But some small players are closer to optimal
• Bidders lacking trading experience?
• Sticky market for managerial efficiency?
“Testing” Explanations for Suboptimal Bidding
1. Not enough $$ at stake  avoid the balancing
market
– Potential profits for each 6-7pm
•
•
•
Reliant = $6,165
Lamar Power Partners = $1,391
But Bryan = $315!!
2. Learning
– Exercising market power on DEC side (“monopsony”)
may not be obvious
•
Bid to DEC low so you’re short but at a low price
– Decrease in bid-ask spread
– Profitability over time
– Use more bid points over time
“Testing” Explanations (contd)
3. Adjustment costs?
•
Marginal generating unit most often is gas (very
flexible)
4. Transmission congestion is important
–
–
–
We analyze only periods with no interzonal
transmission congestion
Congestion changes residual demand
If cannot perfectly forecast congestion, the bidding
strategy under congestion may “spillover” to
uncongested times
5. Collusion?
•
Would be small(!) players - unlikely
Sample Bidder’s Operations Interface
Residual Contract Positions
Summary Statistics on QC (residual contract position)
Firm
Reliant
TXU
Calpine
West Texas Utilities
Guadalupe
Mirant
Central Power and Light
BP
Bryan Texas Utilities
Lamar Power Partners
Mean
Std Dev
448
413
147
264
38
106
11
51
-1
10
-6
5
-16
130
-26
37
-28
26
-54
68
Min
-500
-400
-300
-268
-50
-20
-661
-100
-100
-250
Max
1801
850
380
293
25
-1
286
0
0
150
Daily Weighted Average Price
Uncongested Hours September 2001 - July2002
Average Price ($/MWh)
50
30
10
-10 32
-30
82
132
182
232
282
332
Intervals INCing in all zones
Intervals DECing in all zones
-50
Days Since Opening of Market on 7/31/01
Difference in average system loads: INC = 33GW DEC=29GW
Can marginal costs differ by that much?
Average BES MW
Uncongested Weekdays September 2001 - July 2002
1400
1000
800
600
400
200
Interval of Day
96
91
86
81
76
71
66
61
56
51
46
41
36
31
26
21
16
11
6
0
1
MW Traded
1200
30000
40000
50000
Avg Load Over Day
0
5
10
15
Hour Interval
20
25
30000
40000
50000
Avg System Load Over Day During July-Sept
0
5
10
15
Hour Interval
20
25
Medians of Reduced-form ‘conduct’ measures
p  MCit ( S
obs
it
RDit ( p)  QCit
( p))  it
RDit ( p)
it  1  Optimal
Medians of Theta
Actual Realizations of Residual Demand
Interval
INC
DEC
Reliant
1.78
1.23
TXU
0.79
2.85
Calpine
15.90
9.64
20 Simulated Residual Demands Each Period
Interval
INC
DEC
Reliant
1.40
1.37
TXU
0.77
2.99
Calpine
12.54
7.68
Example of Data We See
Sept 14, 2001 6:00-6:15pm
Total Balancing Demand = -996 MW
Aggregate Bids and MC
One Firm’s Bids and MC
Reliant on February 8, 2002 6:00-6:15pm
40
35
Residual demand
Bid curve
Ex-post optimal bid
MC curve
30
Price ($/MWh)
25
Optimal (p,q)
20
15
Actual (p,q)
10
5
0
-2500
-2000
-1500
-1000
-500
0
Balancing Market MW
500
1000
1500
2000
Calpine (3rd biggest seller) Example
Test for (Local) Optimality of Bids
Who are the Players?
Generator
Average Balancing
Sales** (MWh)
% of Installed
Capacity
TXU Electric
156
24
Reliant Energy
473
18
City of San Antonio Public Service
*
8
Central Power & Light
28
7
City of Austin
40
6
Calpine
78
5
Lower Colorado River Authority
*
4
Lamar Power Partners
23
4
Guadalupe Power Partners
8
2
West Texas Utilities
10
2
Midlothian Energy
*
2
Dow Chemical
*
1
Brazos Electric Power Coop
5
1
Others
*
16
* Cannot uniquely identify the bids
** Sales in zones where bids can be uniquely identified