1. E QUILIBRIUM MODELLING OF BENEFICIARY - PAYS TRANSMISSION CHARGES P ROF. A NDY P HILPOTT, D R. A NTHONY D OWNWARD EPOC W INTER W ORKSHOP 2013

2. 2 of 28 B ACKGROUND / M OTIVATION The EA has proposed that, in order to fund current/future investments in the transmission grid, a beneficiary-pays scheme ought to be introduced. The fundamental aim of this scheme is that those who benefit from the investment will be required to pay for the investment (in proportion to their benefit). While the aim of this pricing mechanism may be fair, problems arise in actually being able to compute the benefits created by adding an asset to the grid. The key proposal was that SPD would be run, with and without an asset (e.g. a transmission line), and the difference in an offers’ infra- marginal rent would be treated as the benefit.

3. 3 of 28 O VERVIEW Beneficiary-pays Pricing Computation of Rentals / Benefits Incentives to Reduce Rentals / Benefits Supply Function Modelling Uniform Pricing Pay-as-bid Pricing Supply Function Equilibrium Examples Single-node – tax on rentals Two-node symmetric quadropoly – tax on benefits Conclusions

4. 4 of 28 B ENEFICIARY - PAYS P RICING Transmission Pricing Methodology Consultation Paper (2012)

5. 5 of 28 B ENEFICIARY - PAYS P RICING Solve 1Solve 2Change DemandA + B + C +DAB + C + D SupplyE + F + GB + EF + G – B Transmission Pricing Methodology Consultation Paper (2012)

6. 6 of 28 I NCENTIVES TO R EDUCE C HARGES After the transmission pricing proposal was announced, there were concerns over the way the benefits would be computed. Particularly, the ‘profit’ of a firm would be assumed to infra- marginal rental. Quantity (MW) Price P D

7. 7 of 28 I NCENTIVES TO R EDUCE C HARGES Thus, in a context where firms are charged based on infra-marginal rentals, there may be incentive to mark-up infra-marginal offers so as to reduce these rents. We will explore these incentives through a supply function equilibrium duopoly. Quantity (MW) Price P DQuantity (MW) Price P D

8. 8 of 28 I NCENTIVES TO R EDUCE C HARGES EA TPM Consultation Presentation November 2012: “Parties may be able to alter their offers to avoid the charge e.g. South Island generators could reduce their beneficiaries-pay charge for Pole 3 by offering as if only Pole 2 was available. “To the extent parties can do this it would reveal the asset is not economically justified unless the SPD charge recovered costs from other beneficiaries e.g. costs of Pole 3 may be able to be recovered through the SPD charge from consumers.”

9. 9 of 28 S UPPLY F UNCTION A UCTION

10. 10 of 28 S UPPLY F UNCTION A UCTION Quantity ( MW ) Price ∑S(p)D(p) p D(p)+h

11. 11 of 28 M ARKET D ISTRIBUTION F UNCTION Anderson and Philpott (2002), Wilson (1979)

12. 12 of 28 U NIFORM P RICE D ISPATCH Anderson and Philpott (2002)

13. 13 of 28 P AY - AS -B ID D ISPATCH Anderson, Holmberg and Philpott (2013)

14. 14 of 28 I NFRAMARGINAL R ENTALS

15. 15 of 28 M ODELLING A T AX ON R ENTALS

16. 16 of 28 S YMMETRIC D UOPOLY

17. 17 of 28 S YMMETRIC D UOPOLY Symmetric equilibrium with no tax (black) and with 25% tax on observed profit (blue). Generators mark-up low priced offers.

18. 18 of 28 S YMMETRIC D UOPOLY Welfare with 25% tax on observed rentals (times 4860). Taking into account the tax on rentals, suppliers offer to improve their actual after-tax payoffs. A side-effect of this is a transfer of some wealth to consumers. No TaxWith TaxIncrease Suppliers’ rentals3,2402,560–680 Suppliers’ payoffs6,4806,344–136 Tax810640–170 Suppliers’ payoffs after tax5,6705,70434 Consumer welfare3,2403,376136

19. 19 of 28 T AX ON B ENEFITS However, in the proposed transmission pricing methodology, the benefits would be computed based on the difference between the rentals in the current market, and those computed for a counterfactual without a transmission asset. In this context, the incentive to increase one’s offer curve is reduced. Quantity (MW) Price P DD’ P’

20. 20 of 28 I NCENTIVES TO R EDUCE C HARGES However, in the proposed transmission pricing methodology, the benefits would be computed based on the difference between the rentals in the current market, and those computed for a counterfactual without a transmission asset. In this context, the incentive to increase one’s offer curve is reduced. Quantity (MW) Price P DD’ P’ Quantity (MW) Price P DD’ P’

21. 21 of 28 I NCENTIVES TO R EDUCE C HARGES However, in the proposed transmission pricing methodology, the benefits would be computed based on the difference between the rentals in the current market, and those computed for a counterfactual without a transmission asset. In this context, the incentive to increase one’s offer curve is reduced. Quantity (MW) Price P DD’ P’ Quantity (MW) Price P DD’ P’

22. 22 of 28 I NCENTIVES TO R EDUCE C HARGES However, firms may be incentivised to mark-up supra-marginal offers. Consider the situation where a transmission investment has allowed additional supply into a node. In the current market, a generator at that node is unlikely to be dispatched for their final tranche; whereas in the counter-factual they could be. Quantity ( MW ) Price D(p) p pcpc pcpc counterfactual

23. 23 of 28 T RANSMISSION E XAMPLE

24. 24 of 28 P REDICTED B EHAVIOUR ?

25. 25 of 28 T RANSMISSION E XAMPLE

26. 26 of 28 T RANSMISSION E QUILIBRIUM

27. 27 of 28 T RANSMISSION E QUILIBRIUM

28. 28 of 28 C ONCLUSIONS On the surface, if you consider a single dispatch point, clear incentives to avoid transmission charges exist. However, since the charge is based on benefits, the incentive to increase infra-marginal offers is limited. Ability of firms to mark-up in a supply function, with many demand realisations is restricted. Furthermore, competition restricts ability of firms to mark-up to reduce transmission charges. Asymmetric players Demand response F UTURE E XTENSIONS

29. 29 of 28 I NCENTIVES TO R EDUCE C HARGES In this work, we wish to investigate how these incentives to change one’s bid are affected by uncertainty, as well as competition.