Bank resolution is costly. In the United States, the Federal Deposit Insurance Corporation (FDIC) typically resolves failing banks by auction. If a bank is failing, healthy banks are encouraged to compete at auction to buy it. This results in a cash transfer from the FDIC to the buyer; the failing bank then continues under new ownership.

The FDIC tries to minimize these transfers by holding competitive auctions. The main feature of these auctions is that they are scoring auctions.

  • First, healthy banks place bids that can differ along multiple dimensions. These are scored based on their estimated costs.
  • Second, to foster competition, bidders are encouraged to submit multiple bids, even though only one bid can win.

This paper proposes a methodology for analyzing auction environments where bids are ranked according to multiple attributes but there is uncertainty about the scoring rule used to evaluate them. We use this framework to estimate the cost to the FDIC of having an opaque scoring rule. We find that the FDIC could reduce costs of resolution by around 17 percent by removing uncertainty.