Search

This paper proposes a game-theoretic model to analyze the strategic behavior of inc-dec gaming in market-based congestion management (redispatch). We extend existing models by considering incomplete information about competitors' costs and a finite set of providers. We find that these extensions do not dissolve inc-dec gaming, which already occurs in our setup of two regions. We also benchmark market-based redispatch against grid investment, cost-based redispatch, and the Vickrey-Clarke-Groves mechanism. The comparison highlights a significant inefficiency of market-based redispatch and inflated redispatch payments. Finally, we study seven variations of our basic model to assess whether different market fundamentals or market design changes mitigate inc-dec gaming. None of these variations eliminate inc-dec gaming entirely.

Auctions with endogenous rationing have been introduced to stimulate competition. Such (procurement) auctions reduce the volume put out to tender when competition is low. This paper finds a strong negative effect of endogenous rationing on participation when bid-preparation is costly, counteracting the aim to stimulate competition. For multiple auctioneer's objectives mentioned in directives, we derive optimal mechanisms, which differ due to different evaluation of the tradeoff between participation and bid-preparation costs. Thus, the auctioneer needs to decide on an objective. However, reducing bid-preparation costs improves the optimal values of multiple objective functions.

This paper derives Bayesian equilibria under general conditions for a broad family of sealed-bid combinatorial auction formats under incomplete information. Equilibria are constructed by linking an equilibrium of the combinatorial auction format to an equilibrium of its single-unit version, thereby providing a closed-form expression if the single-unit version has one. Auction formats that share the same single-unit version, like the Vickrey auction and all bidder-optimal core-selecting auctions, have common equilibria for any number of items and bidders. Furthermore, if an ex-post equilibrium exists in the single-unit version of an auction then the auction format has an ex-post equilibrium in the combinatorial auction. We characterize the ex-post equilibria for a subfamily of auctions.

We investigate equilibrium bidding behavior of bidders with reference-dependent preferences and independent private values in single-unit English and Dutch clock auctions. Bidders' reference points are endogenous and determined by their strategy and their beliefs about the other bidders. In deriving their strategy, bidders anticipate changes in their reference point due to updated information about others' values (i.e., the own winning probability) during the course of the auction, and make optimal binary decisions at each price (approve or quit in the English auction and wait or bid in the Dutch auction). First, we solve for personal equilibrium profiles, i.e., profiles of bids that contain for each bidder a bidding strategy that is optimal given the others' bidding strategies and the reference point induced by the own and others' strategies. Second, we consider locally preferred personal equilibrium (LPPE) profiles, i.e., personal equilibrium profiles where no bidder can locally find a better personal equilibrium given the others' fixed strategies by varying his reference point via considering different own strategies. The expected revenue in the Dutch auction is higher than in the English auction in the respective unique LPPE profiles. The difference is mainly driven by the aversion to losing the item in the Dutch auction.

We find new equilibria of minimum-revenue core-selecting (MRCS) auctions that, in contrast to previously identified equilibria, involve overbidding - bidding more than one's true value for some packages of goods. With full information, every MRCS auction in every possible setting has equilibria with overbidding and these equilibria have different properties than the previously known equilibria with bid shading. Namely, they can lead to strictly higher revenues for the seller and larger price differences among bidders. Previous studies of MRCS games with incomplete information assumed restricted strategy spaces that prevented overbidding. In this paper, we allow bidders access to their complete strategy sets and show that, in some settings, overbidding occurs in all Bayesian equilibria in undominated strategies. In a simple setting with independent private values, equilibrium strategies of a particular set of MRCS auctions employ a mixture of bid shading and overbidding. These new equilibria improve expected effi ciency relative to equilibria with restricted strategy spaces and lead to higher expected revenues than those from the Vickrey package auction. A second incomplete-information setting demonstrates that equilibria with overbidding can be in some sense unique. In this setting, every Bayesian equilibrium in undominated strategies of every MRCS auction has at least one bidder who overbids and there is no bid shading on winning packages. Overbidding eliminates the threshold problem, leading to an effi cient assignment and payoffs that are in the core with respect to the true values.