Reciprocal Dumping under Dichotomous Regulation.

AuthorDebia, Sebastien

    An essential ingredient to net-zero-emissions policies is to regionally integrate electricity markets for allowing a better access to renewable energy sources, and hedging their intermittency (IEA, 2021). In other words, it has become essential to internationalize the trade of electricity as resources allocation becomes more important in decarbonized electricity sectors. Yet, the electricity sector's regulation is designed at an intra-jurisdictional level and cross-border transactions are generally not included (Wilson, 2002; Joskow, 2008). When they are monitored, cross-border transactions are often assessed as inefficient. While uncertainty and differences in market design are certainly a part of the explanation, bidding behavior also raises an issue.

    For example, the power scheduled between New York and New England--two markets regularly assessed to perform competitively--on their common market platform (CTS) has flown in the efficient direction 58% of the time in 2019. At a price spread of $0/MWh, the net flows of CTS transactions would have been 889 MW from New York to New England, on average. And ISO New England to conclude that "CTS transactions continue to flow in the uneconomic direction due to bidding behavior" (ISONE (2020), pp. 150-151). Empirical evidences of imperfect arbitrage over explicit auctioning of interconnectors capacity in Europe can be found in Turvey (2006), Bunn and Zachmann (2010), Gebhardt and Hoffler (2013), Gugler et al. (2018) and Gissey et al. (2019). Gugler et al. (2018), notably, documented that, when cross-border trade in Europe was organized through explicit auctions, interconnection capacity was binding in both directions at the same time for the majority of occurrences.

    More generally, the efficiency of international trade is monitored by the World Trade Organization, which main objective is to eliminate public policies that (1) restrict access to a national market by foreign firms, and (2) promote over-exports by national firms, notably through discriminatory measures. Countries' competition policies, as long as they are non-discriminatory, are not considered detrimental to trade's efficiency (Bagwell et al., 2016). When it comes to trade, national competition policies typically differentiate between local and foreign activities (Levinsohn, 1996). For example, the United States Foreign Trade Antitrust Improvements Act of 1982 provides exporting firms with a significant exemption to the Sherman Act, as long as the domestic effects are incidental and insubstantial (Becker, 2007). Likewise, Articles 81 and 82 of the European Commission Treaty focus on anti-competitive behavior only within the European Union. Hence, competition policies do not formally prohibit regulated monopolists' rent-seeking, or several producers cooperating, for their export activities.

    In this paper, we show that such dichotomous (domestic--foreign) regulations provide firms with a rationale to distort their domestic market. Firms are willing to dump their product in the foreign market as an indirect means of creating scarcity in their local market. The trade regime between markets is a function of their relative price-elasticity. The lower the relative price-elasticity of a market, the stronger is the incentive to dump for the local producer, and the stronger is the incentive to withhold for the foreign producer. This effect can be substantial as a two-player symmetric equilibrium is Pareto-dominated by the competitive benchmark. Further, we characterize reciprocal dumping as the equilibrium pattern of trade for weakly asymmetric players, implying that, by continuity of the analyzed functions, there exist a set of asymmetric equilibrium Pareto-dominated by the competitive benchmark. Very asymmetric situations may also be inefficient. We show that the endogenous allocation of transmission rights in the trading game between Quebec and New York developed in Debia et al. (2018), a companion paper, leads to the worst-case scenario, where investing in a new interconnection decreases welfare.

    Competition policy and regulation are treated as abstractions: they are modeled by their end and not as a tool. In the presence of market power, the goal of a regulator is to limit the potential for mark-up pricing. In this respect, we use the terms "competition policy" and "regulation" interchangeably: despite being different means, they aim to reach the same end. (1) In our case, each regulator replicates the competitive outcome in its local market by enforcing marginal-cost pricing: for a given level of exports, the first-best outcome is enforced at the local level. However, we find that the resulting equilibrium is inefficient. In that sense, our paper is theoretical--it outlines an economic problem--and leaves to further research the normative question of which particular policy to implement.

    Our work contributes to the literature on reciprocal dumping in showing that reciprocal dumping occurs when local prices are regulated at producers' marginal cost, which stands in sharp contrast to the traditional view that, for dumping to occur in a deterministic framework, it is necessary that producers be price-makers (Krugman and Obstfeld, 2008). (2) Our framework of strong-local versus weak-global regulations is fully general, as the field of application of competition policies is only limited by the extent of economic activities. The impact of competition policies as a motive for dumping should however matter the most in sectors where the local market is of relative importance. e.g., the electricity sector.

    Reciprocal dumping-defined as intra-industry trade (3) where both producers are willing to sell at a free-on-board (FOB) foreign price lower than their home price--was first demonstrated in Brander (1981) in a two-Cournot-player framework, with constant marginal costs and linear inverse demands. For the producers to equalize their perceived marginal revenue in the two markets, they sell in both of them. Since, by construction, prices are equal in the two markets, the FOB foreign price for each player is necessarily lower than the home price in presence of transport cost, and intra-industry trade is thus equivalent to reciprocal dumping. However, reciprocal dumping is considered beneficial if the competitive effect of trade is large enough. (4)

    Two assumptions are critical to obtain our results. First, the firms are able to anticipate the impact of the regulation on their strategy. To account for this anticipation, we model this situation as a two-stage game. In the first stage, each producer sets its export level a la Cournot, while paying an iceberg transport cost, that is, a part of the value is lost during the transportation process (Samuelson, 1954). In the second stage, producers decide on the quantity to sell in their domestic market, subject to regulation. In our framework, the players can thus be interpreted as (locally) regulated monopolists, up to atomistic producers allowed to cooperate for exports.

    The direct interpretation of our sequentiality is that exports are contracted forward of local sales. Forward contracts are commonly used to hedge risks. For Selten (1998), the sequentiality in a multistage game "focuses on the delay between the time a decision is made and the time at which it becomes effective". Hence, it is only natural to consider that exports take a longer delay to be effective than local sales decision. In electricity, over-the-counter contracts associated with explicit auctioning is the preferred mode of trade by far (Gissey et al., 2019). More importantly, even if implicit auctioning is used to arbitrage the price differential in spot markets, this arbitrage may be made along price signals biased by the physical position implied by forward contracts. (5)

    But sequentiality needs not to have a direct temporal interpretation (Fudenberg and Tirole (1991), p.70), and should first be interpreted in terms of "anticipations" and "information sets", two recurring quotes in Kreps and Wilson (1982). (6) Sequentially can also be motivated by hierarchy. For example, Ehrenmann and Neuhoff (2009) model strategic power producers to anticipate the reaction of the competitive fringe of producers when they set their Cournot strategy. Hortacsu et al. (2019) show that cognitive hierarchy among firms--a multistage model solved recursively from the least to the most strategic firm--explains well the bidding behavior in the Texas electricity market. (7) In our model, a hierarchy naturally appears between local production--a non-strategic variable because of an efficient local regulation--and exports, on which firms can act more freely because of the lack of international regulation.

    Second, increasing marginal cost is necessary to provide the players with an incentive to dump. It is not unusual to employ increasing marginal cost in economics. (8) It requires only that the supply-side of local market consists of various technologies with different productivity levels. Such a technological heterogeneity is even a requirement to efficiency when the product is hardly storable and the demand is volatile, e.g., electricity. (9) These differences in aggregate productivity are at the center of the new trade theory initiated in Melitz (2003). Several studies in electricity economics assume increasing marginal cost per technology (Green and Newbery, 1992; Ehrenmann and Neuhoff, 2009; Antweiler, 2016; Debia et al., 2018; Genc et al., 2020).

    The rest of the paper is organized as follows. In section 2, we state the main assumptions of the model. In section 3, we characterize the equilibrium of the game and provide our main welfare result. In Section 4, we provide sufficient conditions for existence and uniqueness of equilibrium, and show that the players' exports are strategic complements under mild assumptions. We provide a quadratic example and our numerical application in Section 5...

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