The Impact of Capacity Market Auctions on Wholesale Electricity Prices.

• Author: Moraiz, Francisco

1. INTRODUCTION AND BACKGROUND

   In 2014, government runs the first of a number of Capacity Market auctions for delivery from 2018/19 onwards

   Prior to the introduction of the Capacity Market, secure supplies were entrusted to wholesale markets (providing adequate capacity) in conjunction with National Grid's deployment of balancing tools (managing the challenge of continuously balancing supply and demand). (1)

   In 2013, government identified a risk to future supply adequacy. (2) In particular, government was concerned that at times of scarcity, wholesale prices may be too low to sufficiently reward generators that could provide power, and this 'missing money' (3) or even the perception of it may reduce planned investment in the capacity required to cover peak demand.

   To address this risk, government announced in 2013 its commitment to introduce Capacity Market auctions delivering from 2018/19, with auctions held four years (4) in advance, to pay generators that make capacity available during these years. These payments provide an additional revenue stream for generators who continue to be able to sell power in the wholesale market. Three major four-year-ahead auctions have been held so far for delivery years in 2018-19 to 2020-21 costing around [pounds sterling]1 billion each (see Table 1).

   The first indication from government of the possibility of a Capacity Market for 2017/18 was on 1 March 2016, when government consulted on its proposal. (5) The government confirmed its intention in the summer. The auction was held in February 2017, around 9 months ahead of delivery. 54.43GW cleared at a price of [pounds sterling]6.95/KW, determining payments of around [pounds sterling]380 million.

   Importance of efficient interaction between Capacity Market and wholesale market

   The rationale presented in 2016 for introduction of the Early Auction in 2017/18 was to address market failures and other drivers (6) in causing a lack of investment in traditional generation facilities. An Early Auction was identified as necessary to ensure sufficient existing capacity and provide incentives for new-build capacity and thereby enhance security of supply for winter 17/18. The proposal was consulted on against a back-drop (peaking in intensity in February 2016) of rumours and announcements of impending plant closures, which were reflected in scenarios (notably significantly enhanced closures) in DECC's analysis. (7)

   A crucial component of ensuring a Capacity Market that functions in the interests of the consumer is its efficient interaction with the wholesale market. In practice, Capacity Market payments may be greater than the amount of 'missing money' in the wholesale market. (8) But the additional capacity procured in the Capacity Market should also enhance supply and thereby reduce wholesale prices (expectations of the supply shift should similarly lower forward (9) wholesale prices). In theory, the net injection of revenues should equal the missing money.

   DECC's 2016 impact assessment for the Early Auction modelled an expected reduction in wholesale costs ([pounds sterling]1.5 bn) equal to roughly half the expected cost of the capacity market payments ([pounds sterling]3.2 bn). (10) However, there is uncertainty on the extent to which the wholesale market will respond to Capacity Market introduction. Academics and other stakeholders (11) have identified this as an important area of research--Newbery (2016) for instance argues that wholesale costs may not fully compensate for these payments.

   The size of the sums involved combined with the uncertainty over how and whether practice will play out according to theory emphasise the importance of research such as ours in shedding light in this area.

   The academic literature contains many papers focusing on the rationale and design of capacity mechanisms. Cramton and Stoft (2006), Joskow (2007), and Bushnell et al. (2017) analyse the economic properties. Fabra (2018) provides a formal model, and Newbery (2016) analyses some key policy aspects and market distortions that could be exarcebated by capacity auctions.

   We are aware of one attempt to evaluate empirically these interventions (Scouflaire, 2018). Our paper assesses the impact of the announcement proposing introduction of the Early Auction on wholesale prices, and estimates the accompanying wholesale revenue impact. It takes advantage of the unique opportunity to study 'forward prices' for winter 2017/18 both before and after the announcement of the introduction of a Capacity Market for that winter, an opportunity not available for other auctions which were announced long in advance of formation of forward prices. It draws on theory that suggests that introducing a Capacity Market should reduce wholesale prices by enhancing expectations of supply and thereby lower expectations of scarcity and market power (controlling for other factors), and that this price effect should be more pronounced in peak prices than in base prices (explained later).

   Our analysis therefore examines how the spread between peak and base prices changes between the pre and post announcement periods, using base prices as a 'control' group against which to test the effect of the announcement on peak prices (the 'treatment' group). The possibility of a price effect on 1 (st) March builds on an assumption that the Early Auction was unexpected and therefore not already 'priced in'. This assumption, which draws on our discussions with colleagues in DECC and is consistent with trade press reports (12), we test later. Our study differs from that of Scouflaire (2018), in that it focuses on one rather than many interventions, and that the short lead-in time for the early auction suggests certain potential impacts such as mix changes and increase in baseload capacities may be relatively constrained compared to auctions with greater lead-in times.

   The rest of the paper is structured as follows

   * Section 2 outlines theory

   * Section 3 outlines empirical method, data and final model

   * Section 4 presents results and tests

   * Section 5 discusses possible interpretations, and outlines assumptions and caveats

   * Section 6 presents conclusions and outlines opportunities for further research

2. THEORY AND METHOD

   Introduction

   Theory suggests that expectations of scarcity prompt higher prices, that this increase is mostly reflected in peak (rather than off-peak, or baseload) prices, and that capacity markets should dampen expectations of scarcity. Since this is based on future expectations, we mainly focus on the impact on forward rather than day ahead prices. (13)

   A key paper that offers a promising method for identifying policy impacts such as the announced introduction of a Capacity Market is "Economic impact of enforcement of competition policies on the functioning of EU energy markets" published by the European Commission (2016). This paper seeks to identify the impact of EU competition policy enforcement in driving stronger competition in European gas and electricity markets and therefore contributing to lower prices, higher investment and improved productivity. It evaluates empirically the price effects of two individual competition policy enforcement cases using the did approach.

   Of particular note is the case study on the Commission's case against E.ON (2008) for its alleged abuse of dominant position in the German wholesale electricity market. This study examines the impact of the Commission's decision on wholesale electricity prices, using daily data of peak and off-peak prices from the European Energy Exchange (EEX). The results show that the Commission's decision, by affecting supply and competition in the EEX, led to a reduction in wholesale electricity prices in Germany. This case study is particularly pertinent because it applies a method to identify the effect of withdrawing capacity from the market on wholesale prices, a method which can similarly be applied to identify the impact of adding capacity, (14) which we spell out later.

   Scarcity and expected scarcity mean higher prices

   Among other important variables such as gas and coal prices, scarcity--and expectations of scarcity--should also influence prices (particularly in the face of constraints to storing energy). This is because greater scarcity, which reflects a reduction in the margin between capacity and demand, implies a higher risk of market power, with accompanying higher prices. This may allow prices to be very sensitive to information on scarcity and capacity margins, and for spot prices to move much higher than the long-term equilibrium price. Expectations of capacity margins can also influence more long-term ('forward') prices so, other things equal, heightened expectations of scarcity should translate into higher forward prices. (15)

   Peakprices reflect scarcity to a much greater extent than off-peak and base prices

   Challenges in the storage of electricity contribute to the wholesale price difference between demand during peak hours and off-peak and baseload hours. Products that serve peak hours (peak load) are higher than off-peak and baseload not only because they employ their inputs less efficiently, but also because they are targeted at moments when demand is higher and pushes against the limits of available capacity, and it is for this reason that scarcity value--and the ability to express market power--is more likely to kick in. (16)

   The Capacity Market announcement should dampen expectations of scarcity and lower prices

   Expectations of scarcity should be influenced by expectations of supply and demand, which in turn may be affected by regulatory interventions that drive a revision of expectations of supply or demand. In particular, the announcement of a Capacity Market, itself driven by concerns of insufficient supply, should serve to dampen expectations of scarcity by prompting traders to expect more capacity. Given the transmission mechanism outlined between scarcity and prices, it...