Work in progress
A Cournot Model of Complex Bidding
I study the welfare effects of a dynamic element of wholesale electricity auctions: complex bids. Complex bids allow firms to submit a minimum revenue requirement per day as well as the usual hourly quantity offers, enabling them to link valuations across time. The regulator faces a trade off when deciding whether to allow complex bids or not. Introducing complex bids allows for greater flexibility in bidding and has the potential of improving the efficiency in the market, as bidders can better express their cost complementarities. However, it also gives them another dimension to exert market power. In order to understand this trade-off, I develop a theoretical model of complex bidding in which firms compete in quantities. The theoretical model is broadly consistent with observed bidding patterns and can be computed analytically. I find that complex bids can either increase or decrease welfare in the market, depending on elements such as the tightness of the market or the relevance of cost complementarities.
Complex Bidding and Nonconvexities in Competitive Electricity Markets, with Ignacio Pérez-Arriaga
In the presence of non-convexities, marginal prices can fail to clear the market. In the context of electricity markets, this impossibility result is known as the “energy pricing problem.” Complex bids are often introduced to circumvent this issue. Our goal is to understand how different complex mechanisms affect the incentives of firms both in the short and the long run. We examine first the performance of the efficient pricing mechanism proposed by O’Neill et al. (2005), which implements efficient outcomes in the short run. We explore whether these efficient prices are compatible with the optimal technology mix in the market. We find that efficient prices do not in general support the optimal technology mix in the long run, as discriminatory or “make-whole” payments to the firms do not make cost signals visible to other technologies. We find that a mixed complex bidding design, in which the efficient outcome is achieved and the uplift payment is minimized, gives the best incentives in the long-run. Intuitively, such mechanism makes most of the cost structure visible to the firms in the market by relying substantially on the marginal price for startup costs recovery, while guaranteeing incentive compatibility by means of minimal uplift payments. This mechanism is in line with the convex-hull design suggested in Gribik et al. (2007).
Local Congestion and Market Power
In electricity markets, the presence of congestion can change the scope of the relevant market. A firm can enjoy substantial market power when located in a congested area. I study the behavior of strategic firms in the presence of congestion, based on the Spanish electricity market design. In this market, there is a day-ahead market for energy. Whenever congestion appears, an additional congestion market opens to deal with congested areas. Whereas the day-ahead market uses a uniform auction rule that pays an hourly marginal price to the firms, the congestion market is discriminatory, paying to each firm the value of its own bids. I look at a change in the auction rules in the Spanish electricity market to understand the incentives of firms in congested areas. Before July 2005, firms in the market had to choose the same bids for the day-ahead and the congestion market. After July 2005, firms could submit bids separately for each of these two markets. I test predictions of optimal bidding under joint and separate bidding rules using a regression discontinuity design. The break can also be used to structurally test for optimal bidding under the two bidding rules. I also explore the implications of these two different designs when firms face the threat of market power mitigation measures undertaken by a regulator.