Electricity Policy Options - Evaluation of Regulated Market, Single Buyer and Amalgamated SOE models

Transcription

1 Electricity Policy Options - Evaluation of Regulated Market, Single Buyer and Amalgamated SOE models 17 July 2006 Preface (added December 2006) This document was prepared by MED and does not represent government policy Some of the information contained in these slides might now be dated Disclaimer The Ministry does not accept any responsibility or liability whatsoever whether in contract, tort (including negligence), equity or otherwise for any action taken as a result of reading, or reliance placed on the Ministry because of having read any part, or all, of the information in this document or for any error, inadequacy, deficiency, flaw in or omission from the document.

2 Contents Background Context Evaluation criteria Models to be evaluated Regulated Market Single Buyer Amalgamated SOE Assessment Security of supply Competitive and efficient pricing Minimising transition cost and risks Summary and key judgments Next steps Appendices NZ experience of significant supply issues Contracting in Single Buyer Model examples of key issues Accountability choices in Single Buyer model Amalgamated SOE possible amendments 2

3 Background

4 Background Context A review of past experience indicates that the current Regulated Market model appears to have delivered worthwhile efficiency gains, significant levels of new generation investment, and generally efficient pricing However, there are question marks over: the degree of certainty over further new investment a seeming vulnerability to crises and whether this arises because no single entity is in charge whether hydro/thermal optimisation could be improved whether generation and transmission investments need closer coordination This presentation compares the Regulated Market to two possible alternatives, (Single Buyer and Amalgamated SOE) using a common set of criteria The assessment recognises that decision-makers in all three models will face challenges arising from underlying uncertainties over issues such as: the rate of growth and location of electricity demand the extent and timing of further economic gas finds in NZ the direction of national and international climate change policy the effect of technology change on electricity demand and supply 4

5 Background Underlying characteristics - costs and prices Electricity cannot be stored easily and must be produced at the instant it is required - necessitating a centralised system for coordinating production Electricity is a key enabler for society its value is very high relative to its cost Even though electricity is homogenous, different generation technologies have markedly different cost drivers this creates: potential for big profit swings, e.g. change in thermal fuel prices can cause big gains/losses for renewable generators Forward price formation for wholesale electricity is similar to other homogenous commodities: Buyers seek low prices, sellers seek high prices Arbitrage ensures price convergence for a given place and time On average, wholesale prices are expected to approximate the cost of adding new generation capacity (which can go up or down through time) The presence of sustained market power could enable prices above cost-of-entry Network prices (transmission, distribution) are regulated Regulation aims to allow risk adjusted return on efficient asset base (and, hopefully, incentives for secure and efficient operation) Retail prices: built up of wholesale electricity cost, lines cost, plus retail costs including a profit margin 5

6 Background Policy goals and evaluation criteria The Government s policy objectives for the electricity sector span economic, social and environmental matters This presentation focuses on effect of different industry models on the attainment of economic objectives: Security of supply (reliability) Competitive and efficient prices over time (costs are as low as possible on a sustainable basis and prices reflect costs) Minimising undue transition risks and costs (e.g. any delay to key generation investment) Social ( fairness ) and environmental objectives can be met under any model (though the choice of policy instrument may differ). These objectives may include: Mitigating climate change Assisting lower income groups Pricing water used for hydro generation 6

7 Background Assessment criteria All three models are assessed against a uniform set of criteria: 1. Security of supply Certainty of timely investment in generation* Certainty of timely investment in transmission* Efficient management of dry years* Effective demand side response to tight supply 2. Ability to achieve competitive and efficient pricing Efficient capital investment in generation* Coordination of generation and transmission investment Sustained pressure on operational costs* Least-cost dispatch Pressure on retail costs and innovation Pressure for energy efficiency and demand side management Efficient wholesale prices* Availability of contracts for buyers 3. Minimising transitional issues Minimise transition costs and risks* Minimise transition time * Denotes key criteria for evaluation of relative performance 7

8 Models to be evaluated

9 Models to be evaluated Regulated Market Model Competitive markets for generation and retail, subject to regulatory oversight by Electricity Commission (established 2004) Transmission is managed by Transpower, with new investment proposals subject to approval by the Electricity Commission Refinements to the model are possible (e.g. streamlined transmission approval process, strengthened reserve energy policy), but are likely to deliver incremental improvements only 9

10 Models to be evaluated Single Buyer Model Government agency replaces (or substantially reconfigures) the Electricity Commission Agency undertakes following functions: sole buyer of wholesale electricity (from generators) sole seller of wholesale electricity (to retailers and major users) system operator (schedules and dispatches all plant on grid) planner and procurer of all investment in new generation and transmission procurer of demand side response Agency is sole decision-maker on dispatch of power plants based on its own water optimisation models and contracts with station owners latter will not include any variable charge for use of stored water Agency is not an asset owner, but procures supply under long term contracts which cover cost of thermal fuel, other variable cash operating costs and recovery of capital cost for assets 10

11 Models to be evaluated Amalgamated SOE Combine all existing generator/retailer SOEs into a new entity (around 70% market share) Obligation placed on new entity to offer electricity sale/purchase contracts out to ~10 years, with limits on the maximum spread between buy and sell prices 1 New SOE could be required to divest existing retail customer bases, but allowed to compete for new customers New SOE could become grid owner, with system operator separated into stand-alone independent entity New entity structured as an SOE commercial objective is primary goal (no explicit security of supply obligation) 1. Note : no precedents for this mechanism have been identified in other markets and there is a question mark over its feasibility 11

12 Models to be evaluated Indicative implementation timeframes Timeframe Policy design Legislation Implementation Amalgamated SOE ~1 months Can draw on previous experience (except contract mechanism) Expected to be required (especially for contract mechanism) Structural element relatively straightforward - contract mechanism more difficult Single Buyer ~24 months Significant issues involved no international precedent identified Will be required Major work effort due to complex contracting issues (see appendix), large number of parties and structural change elements Amalgamated SOE Merged SOE Single Buyer Single Buyer EC formation EC formation ECNZ 3 way split ECNZ 3 w ay split ECNZ/Contact ECNZ 2 w ay split split Wholesale Wholesale mkt formation market Expected Implementation timeframes compared (months) to previous experience Key observations from past Policy design Legislation Implementation months 30 Requirement for legislation adds significant time element Timeframe tends to expand as number of significantly affected parties increases Initiatives involving new issues tend to require more time Notes : 1. Policy design includes conceptual design work, consultation, and drafting of legislation. Legislation phase covers period from introduction to passage of Bill. Implementation covers detailed design, resolution of legal, commercial, staffing, information technology and other operational issues. 2. EC formation occurred in phases because initial policy decisions contemplated EC as a backstop policy option. Graph shows time spent in active phases, rather than total elapsed time (46 months) between Caygill Inquiry report (June 2000) and initial EC operation (March 2004). 3. Time spent in policy design phase for previous initiatives based on judgment of when active work began in many cases work began earlier (e.g. WEMDG work for establishment of wholesale market) and policy design built on that foundation. 12

13 Assessment - Security of supply

14 Security of supply Certainty of timely generation investment Policy objective Consumers have a high degree of confidence that sufficient investment in generation is occurring Observations RM has delivered significant new investment, and is backed by the reserve energy scheme. Nonetheless, there have been perceptions that security is at risk and concern at absence of a plan SB expected to provide highest perceived security, due to agency s objectives and ability to present a plan SB is also likely to deliver highest level of actual security, but margin will be less clear-cut and even SB model is not able to guarantee absolute security due to real world uncertainties such as gas availability, RMA consents etc (see appendix for NZ history of supply disruptions) RM and AS expected to deliver similar outcomes, though AS might provide slightly higher perception of security due to its size and implicit security obligation Certainty over timely plant investment Regulated Market 9 Perceived.5 Actual.5 Single Buyer Amalgamated SOE 14

15 Security of supply Certainty of timely transmission investment Policy objective Consumers have a high degree of confidence that sufficient grid investment is occurring (coordination of transmission and generation covered in later slide) Observations NZ has seen relatively little grid investment in last 15 years under both RM and previous arrangements More recently, grid investment processes have been changed and are now a regulated/planned approach Arrangements under SB and AS would be similar to current RM, although SB would combine regulator and grid ownership functions (these are separate in RM and AS) Grid investment outcomes are expected to be similar under all three, although single decision maker in SB should enable faster decisions (leaving aside any delay caused by transition - which is addressed under Transition Issues ) Certainty over timely grid investment.5 Regulated Market 9 Single Buyer.5 Amalgamated SOE 15

16 Security of supply Efficient management of dry years Policy objective Hydro and thermal generation optimised to minimise risk of energy shortage Observations SB model allows all information to be processed by a single decision-maker facilitating consistency of approach But, placing control of all supply-side decisions under single party magnifies the consequences of any mis-judgement or error Concentration of decision making also makes it harder to detect any error beforehand (less contestability and visibility) As a result SB and RM rate similarly, but for different reasons AS may perform slightly better than alternatives because it puts coordination of most (but not all) hydro storage and swing-thermal into one company, but retains some diversity of views Efficient management of dry years Regulated Market Single Buyer.5 Amalgamated SOE 16

17 Security of supply Effective demand-side response to tight supply Policy objective Consumers encouraged to respond effectively to changing conditions through mechanisms such as: provision of short term price signal to encourage conservation reward for providing interruptible load sell-back of discretionary energy Observations 1 RM model has facilitated demand side response by large users, although some parties claim that arrangements are inefficient in periods of high spot prices While SB model allows explicit emphasis to be placed on demand-side, a supply-side focus is likely to predominate monopoly position and administered price structure expected to make it more difficult to provide clear price signals during tight periods Accordingly, despite potential to give SB explicit demandside objectives, performance is likely to be weaker than RM in practice AS is expected to favour supply-side responses even more than RM and SB, because of weaker competitor pressures and absence of explicit demand-side mandate respectively Effective demand side response to tight supply Regulated Market 7 Single Buyer 6.5 Amalgamated SOE 1. It is expected that conservation campaigns could be managed in much the same way in any of the models 17

18 Assessment Competitive and efficient pricing

19 Competitive and efficient prices Efficient capital investment in generation Policy objective Decisions about timing of new investment, choice of technology, and station size deliver desired security at lowest possible national economic cost over the long term. Observations SB model likely to result in some over-investment in supply, due to asymmetry of risk/reward on decision maker (i.e. security failure very visible, but unnecessary cost is virtually impossible to detect) Stronger incentives under RM to invest well and innovate: Investors bear risk if they make poor decisions under RM, but SB is able to pass on costs of stranded assets generators likely to be more innovative to get competitive advantage under RM SB likely to be less innovative and more conservative Other cost factors: investors in new plant likely to get lower cost of capital under SB due to lower revenue risk but expect this to be offset by SB s ability to pass on costs of poor investments to consumers AS expected to be intermediate between RM and SB as it allows for some competitive dynamics, but in a more limited form than RM due to smaller number of competitors Pressure to innovate/invest well 9 Regulated Market 7 Single Buyer Amalgamated SOE 19

20 Competitive and efficient prices Generation and transmission coordination Policy objective Decisions about investment in generation and grid are coordinated to ensure lowest total cost is achieved Observations SB model allows direct coordination, though it will still face a wide range of inherent uncertainties such as demand growth, future climate change policy, RMA decisions etc As a result, while SB is expected to perform better than RM over time, margin is unlikely to be substantial AS is an intermediate option, in that some generation is coordinated directly with transmission decisions, and some (that undertaken by other parties) is not this will also create some competitive conflicts Generation/grid investment coordination Regulated Market Single Buyer Amalgamated SOE 20

21 Competitive and efficient prices Sustained pressure on operational costs Policy objective Cost of running stations and grid is minimised by efficient operational, fuel procurement, maintenance and staffing decisions Observations In RM station owners have strong incentive to manage costs due to direct financial exposure SB is cost-plus. Also, contracting creates two inefficiencies: it will be extremely difficult to write a contract at the outset that properly captures all of the interactions between operating decision (controlled by plant owner) and production decisions (controlled by SB) Even if the initial contract is a good approximation, it will be impossible to foresee all future issues and so will create increasing inefficiencies over time In other words, station owner will respond to contract terms rather than actual circumstances For this reason, SB will produce some operating inefficiencies relative to RM, resulting in higher costs (and potentially prices) than would otherwise apply (see Appendix for example) AS is expected to be intermediate avoids efficiency loss of SB contracts, but competitive pressures will be somewhat weaker due to firms dominant position, and reduction in number of competitors Pressure to minimise costs 9 Regulated Market 6.5 Single Buyer 7.5 Amalgamated SOE 21

22 Competitive and efficient prices Least cost dispatch Policy objective Plant operation and fuel choice decisions (especially hydro vs. thermal) are made to minimise total cost Observations SB model puts decisions on station running order under single party (+ve), but that party is likely to have a bias to conservative water management to meet security objective this will increase hydro spill and thermal fuel burn (-ve) RM model does not give rise to conservative water management bias, but reliance on contracting between swing thermal producers and large hydro generators for hydro-firming service means that some efficiencies are unlikely to be captured AS is expected to reduce some of the contracting costs (especially between Waitaki system and Huntly), and absence of explicit security objective may reduce tendency to over-use thermal plant relative to SB Least cost dispatch.5 0 Regulated Market Single Buyer Amalgamated SOE 22

23 Competitive and efficient prices Pressure on retail costs and innovation Policy objective Retailing has low barriers to entry and expansion, to facilitate cost minimisation and innovation Observations RM has produced strong pressure on retail costs and some innovation in service operation but retailing has been limited to vertically integrated players SB model is likely to facilitate retail entry by new players and allow existing companies to expand coverage (because contracts will be readily available from SB) However, in SB model, retailers will compete on narrower element of value chain, since they all buy from common seller AS contracting mechanism should facilitate retail entry as compared to RM, but likely to be less than SB Pressure on retail costs and innovation Regulated Market Single Buyer Amalgamated SOE 23

24 Competitive and efficient prices Pressure for energy efficiency & DSM Policy objective Facilitate the uptake of energy efficiency and demand-side management Observations SB model allows explicit emphasis to be placed on demand-side, but supply-side focus is likely to predominate in practice due to organisational preference for bricks and mortar to satisfy security objective especially the need for tangible evidence to address perception issues Accordingly, despite ability to give SB explicit demandside objectives, performance likely to be similar to RM in practice because of less dynamic pricing environment and supply-side preference AS expected to perform similar to SB and RM Pressure for energy efficiency and demand side management Regulated Market Single Buyer Amalgamated SOE 24

25 Competitive and efficient prices Efficient wholesale prices Policy objective Wholesale prices for consumers/producers reflect the true costs of using/producing more electricity. Observations Under RM, prices are generally efficient in the long run but market power is sometimes present and exercised Able to price to LRMC, signaling need for new capacity (but not sustainably above LRMC) Not artificially low, which would induce over-use Less competitive pressure during tight periods Contract prices are substantially above spot during wet years a distinct margin is appropriate SB has the ability to replicate this, but less incentive to do so As a monopolist of wholesale electricity, SB has ability to design highly efficient pricing schemes But likely to find it difficult to promulgate these due to its governmental status, especially during periods of tight supply or rising fuel costs Cost plus pricing to recover contract costs (even where LRMC declining) AS performance expected to be similar overall Likely to suffer similar, but weaker, social pressures as SB regarding the use of efficient pricing structures Cost reflective prices Regulated Market Single Buyer Amalgamated SOE 25

26 Competitive and efficient prices Availability of contracts for buyers Policy objective Wholesale electricity buyers can obtain contracts on reasonable prices and terms Observations RM has provided high level of contract cover for retail customers, but larger users argue that generators priority for their retail bases means that prices/terms for residual volume are less attractive SB expected to be indifferent between contracting with major users and retailers SB model will also allow higher level of contract availability due to larger scale of generation portfolio within SB s total control AS contracting mechanism expected to provide sufficient contract volume, but not clear how customers will perceive contract terms Contract availability Regulated Market Single Buyer Amalgamated SOE 26

27 Assessment Minimising transition issues

28 Transition issues Minimise transition costs and risks Policy objective Minimise transition costs and risks Observations SB implementation would be complex and challenging in part because NZ not able to draw on experience from any jurisdiction that has moved to SB from RM Initial contracting between SB and counterparties is also likely to be challenging due to complex nature of issues (see appendix) Transition would require the coordination of extensive change across a large number of parties, many of whom will have significant value at stake creating high potential for contractual disputes If compulsory price setting powers are conferred on SB, there is significant risk of triggering allegation of expropriation by existing generators, with adverse impact on NZ reputation SB direct implementation costs are expected to be significant due to extensive change required, large number of contracts and participants and associated litigation risks AS would also entail significant transition issues, but expected to be less complex and costly than SB Minimise implementation costs & risks 9 Regulated Market 2 Single Buyer 5 Amalgamated SOE 2

29 Transition issues Minimise transition time Policy objective Minimise transition time to reduce scope for any investment pause in sector Observations SB likely to require at least 2 years, and probably longer to fully implement SB will create uncertainty, require high level of engagement from key decision-makers and therefore reduce focus on current investment initiatives AS expected to have lower risk of investment pause because it is shorter to implement, entails less fundamental change, and is more of a known quantity reducing risk for those considering investments Implementation time required Regulated Market Single Buyer Amalgamated SOE 29

30 Summary of evaluation

31 Summary of evaluation Summarising overall picture Key criteria In red No option offers clearly outstanding performance on all criteria Choice depends on relative weightings of criteria Amalgamated SOE tends to be intermediate between other models Security Cost/efficiency Transition Regulated Market Single Buyer Amalgamated SOE Timely plant invt (perceived) Timely plant invt (actual) Timely grid investment Management of dry years Eff dem response to shortage Efficient generation investment Coord of grid and gen invs't Operating cost minimisation Least cost dispatch Retail innovation Energy eff & DSM Efficient wholesale prices Contract availability Min Implem'n cost & risk Min Implem'n time Chart summarises relative performance of models based on assessments in previous sections. Bars depict Regulated Market line above bar indicates superior performance from that model relative to RM line below bar indicates inferior performance from that model relative to RM 31

32 Summary of evaluation Key judgments required Favoured if priority given to: Requires comfort with: Regulated market Cost minimisation and encouragement of innovation Inability to point to definitive plan due to dispersed decision making Noise factor arising from multiple views Single Buyer Amalgamated SOE Addressing security concerns especially perceptions about need for an explicit plan Addressing security concerns (especially perception issues), but concerned about transition risk and implementation cost of Single Buyer model Costs and risks associated with complex implementation process Risk of pause in sector investment momentum during transition Risks arising from placing control with single decision-maker Degree of cost inefficiency Implementation costs and risks (though less than Single Buyer model) Weaker competitive pressures Possibility that it will need to evolve in future either back to Regulated Market or to Single Buyer 32

33 Summary of evaluation Commentary No model offers stand out performance on all criteria SB does best on perceived security, but there is less clear-cut margin on actual security AS likely to allow modest improvement in optimisation of hydro/thermal usage, but expected to perform in intermediate fashion between RM and SB on other criteria RM expected to be best option at minimising costs and promoting innovation over time Implementation costs and risks are significant issues for SB and AS especially an investment pause during the transition phase. Costs and risks are materially higher for SB compared to AS To choose either SB or AS, should be satisfied that expected benefits exceed costs, and that key drivers for change cannot be satisfied by other means 33

34 Summary of evaluation Next Steps Depend on Preferred Option Regulated Market Single Buyer Amalgamated SOE Strengthen Electricity Commission function in relation to medium term security, e.g. provide definitive projection of supply margin 5-10 years out (to provide early warning of any investment problems) Identify ways to speed up decision-making processes for transmission investment (e.g. unify investment and costrecovery approvals for new grid investment with EC) Streamline EC processes by allowing more flexibility and discretion in consultation requirements under certain circumstances Review reserve energy regime Determine accountability arrangement between Minister and Single Buyer (see appendix for more detail) Determine scope of compulsory sale obligation on generators (i.e. mechanism to determine price of power from existing plant) Determine role of Single Buyer regarding retail functions Determine how to address risk of investment pause (e.g. put in place additional contracts to secure investment coming to decision point in transition phase) Prepare detailed implementation plan and schedule Determine scope of generation merger (i.e. all SOEs, or subset) Determine whether grid assets are in or out of new SOE Review practicality of marketmaker contract mechanism as a restraint on new SOE s potential market power Identify other options for addressing market power issues if market maker not viable Determine how to handle retail functions i.e. is SOE required to divest retail or not Prepare detailed implementation plan and schedule and consider other generic issues such as whether to relax generation/retailing restrictions on lines companies 34

35 Appendices

36 Appendices NZ experience of significant supply issues Date , 1975, 1976, Issue A 15% cut was imposed in the North Island Supply authorities were requested to save up to 6%. Broadcasting hours were reduced, ripple control was increased and in some areas there were daily blackouts Government requested voluntary savings Large publicity campaign used to reduce demand. Water heating was generally cut for 1 hours/day, and Comalco closed one of its three potlines (=5% of national demand) Power supply to Auckland CBD was reduced to minimal level for three weeks with restrictions for another month due to multiple distribution cable failures Voluntary demand reductions and buybacks programmes were undertaken during a dry winter (Jan-Aug inflows were lowest on record) Voluntary demand reductions and buyback programme were undertaken to address a period of relatively low inflows and concerns about thermal fuel availability Power cut to half of Auckland for up to ten hours caused by transmission earth wire failure Single decision maker or market-based process Single decision maker (NZ Electricity Dept) Single decision maker (NZ Electricity Dept) Single decision maker (NZ Electricity Dept) Single decision maker (ECNZ) Single decision maker (Mercury Energy) Market based (Multiple parties) Market based (Multiple parties) Single decision maker (Transpower) 36

37 Appendices Contracting examples of key issues Station owner wants Single Buyer wants A station owner has planned for a major thermal plant outage. However, it is very dry in the leadup to the scheduled outage A vibration alarm is triggered on a large gas turbine there have been recent false alarms taking the plant down will cause higher cost oil-fired plant to run Steam production from existing geothermal wells is declining (natural phenomenon). New wells costing ~$500k-$1m each will be needed to restore production levels Retain original timing to avoid additional costs and/or increase risk of plant failure Check plant immediately vibration could indicate a serious problem A contract with SB that prespecifies when to drill new wells, to enable forward planning for drilling rig hire etc. Defer the outage to allow increased thermal production Defer gas turbine outage to a lower demand period (e.g. weekend) to avoid fuel cost (>$1m/day) Enough steam winning to sustain generation levels but impossible for SB to contract right level of drilling ahead of time Long term contracts are not well-suited to addressing the conflicts that arise between Single Buyer s desire to control station running patterns, and owner s desire to look after its plant investment 37

38 Appendices Accountability choices in Single Buyer model 1 Strategic level Operational level Hierarchy of Control Parliament Minister Single Buyer board Single Buyer staff Contracted parties Need to decide who controls which levers Security standards longer term (e.g. 7% reserve margin) within year (e.g. 1:60) real time (e.g. n-1) Pricing decisions level of prices allocation among customer classes allocation by region Investments generation transmission DSM/energy efficiency Station merit order Maintenance practices Broad Alternatives Ministerial power to direct Ministerial power to fire board Comment Crown Agent (e.g. Electricity Commission) Independent Crown Entity (e.g. Commerce Commission Yes No (though entity must have regard to government policy) Yes Very limited ( bad or mad ) Allows high degree of control by Minister. But creates high degree of political accountability Considered best practice. Government sets up institutional arrangements and objectives, but not involved in operational issues. Minimises risks of politicisation of operational decisions (e.g. on pricing issues) 3 1. The same accountability issues arise with the other models because there are choices to be made about the accountability framework for the Electricity Commission. However, the issues are most apparent in the Single Buyer model because of its wide ranging powers and responsibilities.

39 Appendices Amalgamated SOE Possible amendments Original proposal Combine all generatorretailer SOEs to form new entity Obligation on new entity to offer sell/buy electricity contracts at fixed maximum price spread New entity to divest existing retail customer bases but permitted to compete for new customers New entity to own grid Possible revision Place fewer assets into new SOE to reduce market power (e.g. Meridian + Huntly) allocate balance to Mighty River Power Obligation not imposed due to concern over workability (no international precedents) reduce market power by avoiding undue size No divestment requirement but reallocate customer bases among SOEs resulting from smaller scale merger noted above Grid ownership to remain separate entity increase planning information provision to facilitate grid/generation investment coordination Impact on assessment of AS Probably beneficial - narrowing merger scope should allow most benefits to be retained, while reducing detriment associated with large market power No major impact original mechanism may be difficult to implement (though if workability issues could be overcome, mechanism could be applied to all generators above a minimum size) No major impact in substance original approach may create more perceived competition for a period, but does not alter underlying fundamentals Probably beneficial grid ownership by new SOE would create potential competitive barriers for other generators these are likely to outweigh any coordination benefits from original proposal Refinements to the original Amalgamated SOE model are possible though they are unlikely to fundamentally alter the model s relative performance 39