Resolving Failed Banks: Uncertainty, Multiple Bidding and Auction Design.

This article analyzes the Federal Deposit Insurance Corporation’s (FDIC) use of scoring auctions to resolve failed banks during the global financial crisis, focusing on how uncertainty about the FDIC’s scoring rule influences bidding behavior and resolution costs. The FDIC permits multidimensional bids combining continuous dollar amounts and discrete contract components, but bidders face uncertainty about how the FDIC weights these components, motivating multiple bids per auction. Using a structural model and data from 322 auctions between 2009 and 2013, the study estimates private valuations of failed-bank assets and decomposes bidding incentives into competition, substitution, and noise effects. Counterfactual simulations suggest that eliminating scoring-rule uncertainty—either by announcing scoring weights or restricting bids to a single package—could reduce FDIC resolution costs by 29.8% to 44.6% (approximately $8.2 to $12.3 billion), with limited impact on local market concentration. The findings highlight the trade-offs in auction design involving multidimensional bids and scoring uncertainty, offering insights relevant to other complex combinatorial auctions in financial markets.