Design of Auction Safety Valve and Revenue Recycling

Design of Auction, Safety Valve and Revenue Recycling within the Energy Performance Commitment (EPC) Environmental Resources Management Ltd & Market Design Inc 26 January 2007

Presentation Overview • Project objectives • Interim findings for each EPC design aspect: • revenue recycling options • auction design • integrated analysis (including safety valve) • Next steps

Project Key Objectives • Many key EPC policy and design decisions have been taken − Fixed price introductory phase (sales), cap-and-trade fixed volume phases (auctions) − Use of “safety-valve”; preference for “buy-only” link to ETS − Revenue recycling to participants on basis of emissions and/or energy management performance • Design details have not yet been determined • how does EPC incentivise ‘good practice’ amongst participants? • Overall objective is therefore: to provide guidance on the auction, ‘safety-valve’ and revenue recycling aspects of the proposed Energy Performance Commitment (EPC).

3 Aspects Linked – Impact Effectiveness Task 1: Task 2: Task 3: Design of auction method Design of revenue recycling Design of “safetyvalve” Integrated analysis Task 4: Final Scheme Proposal

Revenue recycling

Revenue recycling Assessment of alternative options and proposed formulae • Option 1 (RR 1): Auction revenues returned to participants in proportion to average annual emissions • Option 2 (RR 2): Performance ‘banding’ based on emissions reduction performance (allows % reward/penalty) • Option 3 (RR 3): Performance ‘banding’ based on emissions reduction performance and other criteria e. g. use of HH metering Key criteria include: effectiveness, fairness and simplicity, minimising cost/admin burden to Gov and participants, rewarding early action/incentivising the installation of AMR

Revenue recycling We model revenue recycling for x participants over 5 years Key variables: • Basic recycling option design details - no. of bands, band allocation, rate of reward/penalty • • Number and size distribution of participants Changes in emissions/behaviour across period Key outputs: • • • Determine net winners/losers and extent of wealth transfer Identify perverse incentives/’unfair’ outcomes Identify optimal option according to key policy criteria

Revenue recycling • Different options result in different winners/losers • Simple payment proportional to emissions - insufficient incentive to reduce emissions (mixed signals) • Performance banding option - “league table” − creates incentive to reduce emissions − i. e. reinforces the signals of cap and trade − equal number of participants in each band creates incentive discontinuity (i. e. 1 st place rewarded same as 500 th place) − equal number of participants in each band also requires corrective factor to % reward/penalty − equal number of emissions in each band may be an alternative option, although outcomes are then highly dependent on participant size distribution

Revenue recycling • Use of continuous performance banding − alternative is to have continual linear incentive, i. e. x bands for x participants; simple, fair and no discontinuities − Participants could still be reported in ‘performance bands’ • Increasing the % performance factor does the following: − reduces auction price due to additional incentive to abate along MACC − shifts money from those with higher abatement costs to those with lower abatement costs. − shifts money from growing businesses to declining businesses

Revenue recycling • Multi-criteria recycling option − Participant administrative simplicity maintained (e. g. small increase in reporting requirement only, payments still simple) − encourages uptake of auto metering and may encourage early action − however, non-transparent in outcomes and adds complexity to optimal bidding and abatement strategy − winners and losers will depend on MAC of HH metering and existing availability of the option for each participant

RR 1: proportion to average annual emissions Revenue recycling RR 2: emissions reduction performance

RR 1: proportion to average annual emissions Revenue recycling RR 2: emissions reduction performance RR 3: RR 2 + other criteria

Auction design

Auction design Key attributes of emissions allowances: • • Product is a homogenous, divisible good Product is actively traded in secondary market Market is un-concentrated Downward sloping demand curves

Impact of revenue recycling (1) • Reduces the burden of the scheme on participants but complicates bidding incentives • Bidder’s quantity choice impacts its revenue share in complex way • Performance factor: % bonus or penalty based on relative performance • Increasing performance factor • Reduces auction price • Shifts money to those with lower abatement costs • Causes some shift of money from growing businesses to declining businesses

Impact of revenue recycling (2) • Scheduling of payments is important consideration − Auction at start of year 1 + 6 month reconciliation with payments in mid year 2 = 18 months without cash − However, it is desirable that some time pass between auction and recycled payment − Solution 1: defer participant payment till start of year 2 − Solution 2: participants make net payments only: “settle-up” − Solution 3: auction in mid-year or multiple auctions • Solution 1 recommended: achieves simplicity whilst preserving the strong signal/financial decision associated with payment

Auction design Leading auction candidates: • • One static (one round of bidding) One dynamic (multiple rounds of bidding)

Uniform price auction

Ascending clock with proxy bids

Auction design Proxy bids: • Bidders can participate as in a static auction • or can bid dynamically Other features: • • Internet auction Demand schedules (step function) Proportionate rationing Report excess demand distribution of % reductions − can estimate revenue recycling impact • Recommendation: clock auction with proxy bids, so bidders can better estimate marginal incentive to abate

Integrated analysis

Integrated Analysis • • • Auction Design, Revenue Recycling and the Safety Valve all impact on each other Tasks 1 -3 therefore being conducted as a whole using Case Study modelling Case Studies aim to predict potential outcomes of design options 1. identification of the incentives to change behaviour - investing in abatement, participating in sales/auctions must be understandable to participants › i. e. participants must be able to form an initial strategy and optimise their position within scheme 2. sense checking - are there perverse incentives or unintended consequences? - e. g. too much/too little abatement capacity - how do abatement costs and auction prices interact?

Case Study Fundamentals • • Uses marginal incentives • marginal costs modified by recycling payments Informed by real data (NERA/Enviros etc) Uses scenarios (recycle design, participants, ceilings, etc) Outcomes include auction prices, wealth transfer, etc. Case Studies developed by ERM in Microsoft Excel, with VB coding Models hypothetical cases rather than the system as a whole • Aim is to analyse the effects of design details Very flexible: multi-year, assumptions re: MAC curves, behavioural aspects, etc. Model largely operational but needs testing

Key Outstanding Questions (1) • • • Recycling Formulae Multi-Criteria League Tables Distribution of Sizes of Participants Distribution of Emissions Projections of Participants Baseline Alternatives • • Average emissions since start of scheme, Phase 1 period only, rolling x-year average, etc. Payment Scheduling Safety Valve Buying-in (of alternative allowances)

Key Outstanding Questions (2) • Public/Private Sector split • Ensures that public sector is not a net loser • Technically possible - 2 revenue pots Both revenue neutral • • • Likely to reduce market efficiency Ensures that there will definitely be public sector losers Level of participation in abatement/auctions Achievability of caps Participant behaviour

Participant Behaviour • Perfect Solution • • • Companies act independently of recycling rules Risk aversion • • optimal system solution, perfect rationality and information Over-abate Others?

Next Steps • Draft Final Report due week commencing Feb 5 • • Major addition will be Case Study results/conclusions Will include final design recommendations Final Report due week commencing Feb 19 Meetings scheduled after each Report delivered