Market Power and Renewables: The Effects of Ownership Transfers.

AuthorBalm, Olivier


Over the past few years, electricity markets around the world have seen important changes in their energy portfolios as new sources have been introduced (e.g. wind and solar) and others have been retired or penalized through taxes (e.g. non-refurbished nuclear plants and coal plants). These changes continue nowadays as a number of incentives to curb the greenhouse gas emissions associated with the production of electricity have either been put in place (e.g. production subsidies such as feed-in-tariffs (FiT) and mandates such as renewable portfolio standards (RPS)) or in other cases been dismantled. (1) Although some of the consequences of these policies have been studied, little is known about the effects of these mechanisms in the presence of market power and under different market structures. (2)

In this paper we quantify the net result on wholesale electricity prices of two opposite effects: market power and the merit order effect (henceforth MoE). The latter occurs when there is an expansion of the amount of renewable energy sources (RES): in that case, the system's supply curve shifts to the right and its intersection with the demand curve occurs at a (weakly) lower price than before the expansion. However, partial ownership of RES from firms with market power may counteract the MoE. In fact, as we show theoretically, the best response for firms with market power is to reduce production from conventional sources, which has a positive effect on market prices. We find upper bounds of these effects throughout a series of simulations using an equilibrium model for the Ontario electricity market. First, we quantify the market power effect on its own by measuring the effect on market prices when holding the system's capacity constant and changing the ownership structure. Second, we compare this effect to the MoE by expanding net capacity and allowing different firms to hold the additional capacity. By simulating uniform incentives for the adoption of RES across different ownership structures, we show that allowing market participants with high market power to hold an increasing fraction of RES capacity mitigates the MoE in a way such that wholesale prices do not decrease significantly (about 7% drop in prices, relative to our baseline Cournot simulation). On the other hand, introducing RES through firms that cannot exercise market power would decrease wholesale prices by large amounts (in the order of 30% in the perfectly competitive case). These simulation results demonstrate that the wholesale price savings that final consumers would receive depend crucially on who owns this additional RES capacity, hence the importance of the ownership structure and market power in the design of uniform incentives for RES adoption.

The effect of market power on prices is one of the main takeaways in the theoretical work of Acemoglu et al. (2017) for the case of symmetric firms and symmetric portfolio compositions. In the presence of market power and a more diversified portfolio (provided an initial low share of RES), strategic firms partially internalize the shift of the supply curve caused by the MoE by decreasing output from their conventional generation sources, thus effectively shifting their individual bid curve to the left and causing an upward pressure on prices. Brown and Eckert (2018) expand on this by allowing for asymmetric amounts of RES in the firms' portfolios(3) and Gene and Reynolds (2019) allow for asymmetric strategic firms. (4)

Our paper continues this line of work by providing a simulation-based analysis of the effect of RES ownership transfers and expansions on wholesale electricity prices. RES transfers and expansions arise in current markets due to the different incentives to adopt RES such as RPS and FiTs. Moreover, the long-term existence of such incentives is threatened in several places around the world, implying a risk of even more ownership transfers. One possibility is that payments cease to exist for FiTs and the idle capacity from non-strategic players gets transferred to the strategic players. (5) The effect of these policies on wholesale prices is an open empirical question. To the best of our knowledge, our work is the first to empirically simulate the effects of portfolio changes from RES additions and transfers on the wholesale electricity market. One of the advantages of our work is that, by using actual market data, we simulate the effects of transferring or introducing RES in the presence of asymmetric firms, and at different levels of correlation between RES output and load. From an economics perspective, our paper also contributes to the literature by confirming some theoretical results on the interaction of diversification and market power in electricity markets. From a regulatory perspective, our results contribute to environmental policy analysis by quantifying hidden or ambiguous price effects from the introduction of large-scale renewable sources into the electricity production mix.

As Fowlie et al. (2016) point out, since the seminal work of Buchanan (1969) regarding the effect of taxes to correct for externalities on market outcomes in the presence of market power, only a small number of empirical studies have quantified these interactions in the context of pollution taxes. (6) We contribute to this strand of the literature by focusing on the impacts on wholesale electricity prices had the market structure changed. Specifically if the ownership distribution were different. Recent examples of changes in the market structure in the electricity sector have taken place in Western Australia (Leslie, 2018), in the U.S., with a series of restructuring cases from utilities into investor-owned companies (Knittel et al., 2015) and because of the retirement of coal-powered plants (Kim, 2019) or in Denmark, where the company 0rsted changed its portfolio composition from 17% to 80% RES in 2017. (7) Ownership transfers are also observed in electricity markets: for example, the German company RWE acquired E.ON's and innogy's RES assets in 2019 to become the third-largest firm in Europe by RES capacity. (8)

Although the exact impact of expanding RES is most likely market specific, we believe we can extract general qualitative impacts from the study of a particular market. We choose the case of Ontario and its Independent Electricity System Operator (IESO) to answer these questions. This market has a similar structure to other North American markets such as CAISO, PJM, MISO, and NYISO in that there is a system operator that constructs supply curves for each time interval by combining each of the producer's individual supply curves and taking into account transmission constraints. Those markets have been also the object of other market power studies (see for instance Joskow (2019) where it is recognized that the Alberta and Ontario electricity markets are similar to those bid-based markets in the U.S.). Some European electricity markets also share large similarities with the Ontario market (wholesale market, nuclear-dominant, with increasing amounts of renewables). Among them, Germany is the closest, Belgium, and the UK are similar as well.

We model the market following the methods developed by Borenstein and Bushnell (1999) and Bushnell et al. (2008), and more recently by Brown and Eckert (2016). In particular, the first paper introduces an industry marginal cost curve expressed as a step function that can be fully characterized using engineering parameters and fuel prices. Then, they separate producers into two categories: strategic and non-strategic (price-takers) firms. The former have the capacity to influence the hourly equilibrium price because of their importance in the market or because of their atypical cost structure (e.g. low start-up costs); the latter are firms that own very small nameplate capacities or that have long-term contracts with the system operator or with downstream firms (therefore they have negotiated prices much before the spot price is formed). This second category of firms is important to the specific case of Ontario as a large proportion of producers belong to this category. The other two papers refine those estimation techniques by exploring the consequences of forward contracts and the use of different instrumental variables.

The presence of imports and exports in the Ontario system implies that firms face a different demand than just the domestic one. Following Bushnell et al. (2008), we account for this by estimating a net exports supply function and add it to the domestic demand to get the residual demand that firms face. We estimate this model using data from multiple and publicly available sources. We use weather data from the National Oceanic and Atmospheric Administration (NOAA), hourly production, demand and capacity data from the Independent Electricity System Operator (IESO), firm-level aggregate capacities from financial statements and fuel spot prices from public databases. We use actual production data from wind and solar sources and these are the only RES sources we consider. (9)

Expanding on the stylized theoretical Cournot competition models in Acemoglu et al. (2017), Brown and Eckert (2018), and Gene and Reynolds (2019), we add a competitive fringe to the model, and show that their results also hold in this extended setting. However, we warn that this result holds under strong market assumptions such as symmetry among the firms. (10) This provides a natural motivation for our analysis as only an empirical analysis can shed light on more realistic configurations, otherwise intractable to solve analytically. In particular, we provide a framework to study RES capacity changes regardless of the degree of asymmetry among the competitors.

We estimate a detailed model of the Ontario electricity market to run simulations that consist of finding the new hourly equilibrium prices under different allocations of RES among...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT