Economic and Environmental Consequences of Market Power in the South-East Europe Regional Electricity Market.

• Author: Viskovic, Verena

1. INTRODUCTION

   Two major historic processes of the last four decades have shaped current electricity markets worldwide. First, the deregulation of the electricity industry, which in some cases resulted in electricity markets characterised by oligopolistic ownership structures (Wilson, 2002) potentially subject to the exercise of market power. Electricity markets are particularly vulnerable to the manipulation of prices above competitive levels due to relatively inelastic short-term demand and lack of storage (Borenstein, 2000). Furthermore, market separation due to network congestion can play an important role in the extent to which market power can be exercised (Neuhoff et al., 2005).

   Second, concerns about the effects of greenhouse gases (GHGs) on climate change led to carbon pricing through transferable property rights, e.g., allowances or permits. Emission permits are commonly traded under cap-and-trade (C&T) systems, which can also be subject to market power. For example, Hahn (1984) shows that any initial allocation of permits to a participant with market power that deviates from the quantity of permits consumed in equilibrium results in market inefficiencies, i.e., a participant with market power will either raise the permit price above or push it below the perfectly competitive level.

   The exercise of market power in a single market has attracted attention in the literature; however, the interaction of product and permit markets both subject to market power has been less investigated. Kolstad and Wolak (2003) examine the circumstances in the California electricity market in 2000 and 2001 where part of the market (Los Angeles area) was subject to the NOx C&T. They find evidence that prospects for exercising market power in the electricity market might have been enhanced via the C&T. Specifically, firms that owned plants both in the area subject to the C&T and outside of it might have intentionally paid higher permit prices. Due to the higher permit price, they were able to justify higher offers into the electricity market despite not using the more-polluting plants in a way consistent with the higher marginal cost of production given by the higher permit price. In this way, they could have earned higher profits on less-polluting plants and/or uncapped plants.

   Besides price manipulation, regional electricity markets where only a subregion of the market is subject to a C&T, such as the Pennsylvania-New Jersey-Maryland (PJM) Interconnection, California, and South-East Europe, are at risk of carbon leakage (Burtraw et al., 2006; Viskovic et al., 2017), i.e., an increase in emissions in an uncapped subregion of a regional market as a result of introducing a binding cap in the subregion subject to a C&T. A binding cap results in a positive permit price, which translates into higher electricity prices in the capped subregion and entices the uncapped subregion's production leading to an increase in emissions in the uncapped region. Since exercise of market power raises prices above perfectly competitive levels, the issue is whether the exercise of market power by a firm located in the capped subregion of a regional market can exacerbate carbon leakage through its simultaneous manipulation of the permit price. In particular, given the prominent role of C&T markets in curbing C[O.sub.2] emissions, it is crucially important to examine their vulnerability to such strategic behaviour. Towards that end, we investigate the impacts of (i) environmental regulation under perfect competition, (ii) market power in electricity markets only, and (iii) market power in both electricity and permit markets.

   We address these issues via a stylised 22-node network of the South-East Europe Regional Electricity Market (SEE-REM) that spans EU and non-EU countries in which EU members are covered by the EU Emissions Trading System (ETS) and non-EU countries are exempt from it (Figure 1). We choose a single firm with fixed capacity in the ETS part of the market as the dominant firm. In order to investigate the short-term impact of market power in both electricity and permit markets, we analyse three market settings. First, all firms are price takers in both electricity and permit markets under perfect competition. Second, a leader-follower setting provides the leader with market power in the electricity market only, whereas all other firms are price takers, and the C&T is modelled through an exogenous permit price, i.e., a carbon tax. Such a setting enables us to assess the impact of market power in the electricity market only under varying levels of C[O.sub.2] emission restrictions. Finally, a leader-follower setting provides the leader with market power in both electricity and permit markets, whereas all other firms are price takers, and the C&T is modelled through an ETS emission constraint. Via this framework, we analyse the impact of market power on prices, output, consumption, flows, emissions, and social welfare.

   We find that under perfect competition, a binding cap on ETS emissions curbs ETS production. As the cap tightens, the price differential between ETS and non-ETS areas of the market increases, thereby enticing non-ETS production and leading to higher non-ETS emissions. Consequently, there is carbon leakage between 39%-11% for caps of 10%-40% reduction, respectively, compared to the baseline (Viskovic et al., 2017). If the leader can exert market power in the electricity market only, then its strategy depends on the marginal technology, which changes with the stringency of the environmental regulation. When natural gas is the marginal technology, which occurs at lower carbon-tax levels, the leader's strategy is to withhold production from its dominant technology (coal) in order to raise electricity prices and reap higher profits on its operating power plants. Higher electricity prices entice ETS natural-gas (including the leader's) and non-ETS production, which partly replaces the share vacated by the leader's coal plants. As a result, ETS (non-ETS) emissions fall below (rise above) the perfectly competitive levels. Since the reduction in ETS emissions offsets the increase in non-ETS emissions, carbon leakage is lower compared to the perfectly competitive setting. For a carbon tax such that coal reaches cost parity with natural gas, the leader adopts an opposite strategy coal-wise. In particular, it expands coal production in order to set equilibrium prices. Higher electricity prices entice ETS coal and non-ETS production resulting in ETS and non-ETS emissions as well as carbon leakage above the perfectly competitive level. With the additional possibility to influence the permit price, the leader holds back more coal compared to the carbon-tax setting. A lower abatement cost results in higher ETS natural-gas production and ETS emissions compared to the carbon-tax setting leading to higher carbon leakage. When coal reaches marginal-cost parity with natural gas, the leader expands coal to a lesser extent compared to the carbon-tax setting as it does not want to increase the permit price. Contrary to the carbon-tax setting, since the fringe firms in the ETS cannot increase coal production because of the cap, they increase natural-gas production. This leads to lower ETS emissions and carbon leakage compared to the carbon-tax setting. Generally, the leader is able to reap higher profits when it has the ability to manipulate both markets, except in the case of a tighter cap when its expansion of coal production is limited by the effect that it might have on the permit price.

   The remainder of this paper is structured as follows. In Section 2, we review the relevant literature and explain our paper's contribution. Section 3 describes the mathematical models used and explains how the data are implemented, whereas in Section 4, we analyse and discuss the results. Finally, Section 5 summarises our work and provides directions for future research. Appendices A, B, and C contain the calibration, additional results, and the mathematical reformulation, respectively. They are available online as Supplementary Materials on the Energy Journal website.

2. LITERATURE REVIEW

   Market power in electricity markets remains a concern (Tangeras and Mauritzen, 2018) and is commonly examined through market settings such as Cournot oligopoly, dominant firm-competitive fringe, and multiple dominant firms with a competitive fringe. For example, Gabriel and Leuthold (2010) use a bi-level model to study the impact of market power by a single dominant firm in an electricity market. They find that the dominant producer is able to raise electricity prices above perfectly competitive levels by holding back capacity with the extent of the withholding bounded by the cheapest plant of the follower firms. Meanwhile, Sauma and Oren (2007) devise a tri-level model with a regulated planner deciding upon transmission expansion at the uppermost level, power companies making generation-capacity investment decisions at the middle level, and market clearing at the lowest level. They implement this framework in a Chilean case study to demonstrate the complexities of attaining politically feasible transmission plans when generators exert market power.

   Market power in an electricity market can indirectly affect even a perfectly competitive permit market. Limpaitoon et al. (2011) develop an equilibrium model of an oligopolistic electricity market with a C&T where firms are price takers to study the impact of market structure on market outcomes. Although it is generally expected that more competitive markets have higher permit prices, Limpaitoon et al. (2011) find that this might not be the case when the ownership of relatively cleaner power plants is concentrated among fewer firms. In particular, if cleaner firms withhold generation in order to raise electricity prices, then dirtier firms increase production to compensate partly for the share vacated by the...