Auctions and Trading in Energy Markets: an Economic Analysis

1
Auctions and Trading in Energy Markets an
Economic Analysis

• David Newbery and Tanga McDaniel, University of
  Cambridge
• 8th POWER Conference Electricity industry
  restructuring - Berkeley, March 14 2003
• http//www.econ.cam.ac.uk/dae/electricity

2
Auctions in energy markets

• Electricity
• NETA BM replaces Pool -changes auction
• Gas
• auctions of entry capacity since Sept 1999
• long term auctions since Jan. 2003 (2004-2016)
• Electricity Interconnection
• Germany/Netherlands
• England/France

3
Criteria for successful liberalisation

• deliver the lowest possible sustainable prices
  to all customers, for a supply that is reliable
  in the short and long run
• minimise entry and exit barriers
• minimise transaction costs and avoidable risks
• charges should be cost-reflective and paid by
  those causing cost

4
The case for NETA

• The Pool is too transparent and discourages
  bilateral bargaining
• Making balancing market a poor guide to SMP will
  encourage contracting
• If there is no market of last resort then
  must-run stations have to accept lower bids

5
1998 critique

• The root problem is lack of competition
• If this is resolved the Pool may work better
• Pool replacement may then be unnecessary, costly
  and counterproductive. It will
• accelerate vertical integration
• deter entry so equilibrium prices will rise
• raise transaction costs and hence prices

6
Events from RETA to NETA

• Competition intensified
• Jul 99 Edison buys 4GW 472/kW
• raises load factor from 25 to 40
• AES buys Drax, then offers for sale
• SMP falls 20-30 year-on-year
• Oct 01 Edison Mission sells at 190/kW
• Interconnector raises UK gas prices
• CCGT at margin
• more dispersed ownership ? more competition

7
Source John Bower (Oxford Institute for Energy
Studies)
8
(No Transcript)
9
Did NETA causes prices to fall?

• Fall before NETA, after increased competition
• endgame or too many to collude?
• Was NETA needed to encourage plant sales?
• Quid pro quo for attractive vertical integration?
• Did ending Pool cause (modest) overcontracting
  rather than (modest) undercontracting?
• If so, is this a Good Thing?

10
(No Transcript)
11
(No Transcript)
12
Critique of Balancing Mechanism

• volatile and risky
• SSP low, moderately predictable
• SBP unpredictable, can be very high
• each agent penalised for imbalance
• incentive to over-contract, spill at SSP
• excessive self-balancing, reserves

13
(No Transcript)
14
(No Transcript)
15
Rationalised defence of NETA

• dual cash-out prices ? asym risk
• ? over-contracting ? spot price?
• over-contracting discourages market power
• spot market sets contract price then prices?
• inefficiencies small price for more competition?

16
(No Transcript)
17
UKPX reflects contract prices
18
Costs of NETA

• costly to implement 1 billion and rising
• small genco costs up 16
• wind power can be charged for selling
• BM imbalance exceeds energy value
• self-insure with own spinning reserve
• loss of system multiplexing
• Demand forecasting decentralised
• system accuracy 5, individual 15

19
Incentives to over-contract and part-load

• Supplier forecast error ? 5- 15
• Optimally over-contract x?
• F(x?) (b-p)/(b-s) x 0.1 to 1.6
• - Cost approx 1-3 of purchase costs
• Part-loading a 500MW coal fired plant
• thermal efficiency falls from 35 to 32 (net)
• fuel cost increase 10 ( 1/MWh)
• 0.5-1 million tonnes extra C, 40 million/year

20
Improving NETA

• Reform Balancing Mechanism
• only charge those who add to imbalance
• re-instates statistical averaging
• targets charges on those causing imbalance
• single cash-out price SSP or SBP
• Ofgem may discourage contracting
• But competitive risky markets normally encourage
  contracts
• changes due March 8 2003

21
NETA assessment

• NETA benefits large vertically-integrated (GS)
  companies with smart traders
• excess reserves raise costs
• damages renewables, entry, small companies
• competition, not NETA, lowered prices
• NETA a costly way to buy competition?

22
Auctions for network trading

• For gas entry to replace LRMC charge
• increase transparency, reduce constraint costs
• efficient allocation and rent capture
• For electricity interconnection
• can reduce market power
• single price better than pay-as-bid
• but market integration potentially better still
• Unreliable for financing network investment

23
Monthly vs daily access prices St. Fergus
24
Comparison of costs and returns to buying
interconnection
25
Cost profit of trading from France to England
26
Conclusions

• NETA- costly iteration towards a better system?
• Auctions assign property rights for access
• Auctions work best if
• capacity fixed and well-defined
• enough competitors, liquid spot markets
• constraints otherwise costly to manage
• Good for gas entry, electricity over links
• Less good for meshed networks, investment

27
Auctions and Trading in Energy Markets an
Economic Analysis

• David Newbery and Tanga McDaniel, University of
  Cambridge
• 8th POWER Conference Electricity industry
  restructuring - Berkeley, March 14 2003
• http//www.econ.cam.ac.uk/dae/electricity