130 S. Dasgupta and R.G. Hansen
firms: on average, creditors received 35% of their claims, with secured creditors receiv-
ing 69% and unsecured creditors only 25%. Thorburn finds that APR is maintained by
the auction procedure.
Eckbo and Thorburn (2005)construct an auction-based model to examine the incen-
tives of the main creditor bank in a bankruptcy auction. Their work addresses one fear
of bankruptcy auctions, that credit market inefficiencies will sometimes limit credit and
cause bankrupt companies to be sold at “fire-sale” prices, possibly to the benefit of the
original owner/managers. Eckbo and Thorburn show that the main creditor bank has an
incentive to provide financing to one bidder and to encourage that bidder to bid higher
than would be in their private interest. The reason for this follows from the analysis of an
optimal reserve price (see also the discussion of toeholds in Section4.4) in an auction,
for the main creditor bank is essentially a partial owner of the bankrupt company. Just
as an optimal reserve price exceeds the seller’s own valuation (see above), the optimal
bid for a main-bank financed bidder exceeds that bidder’s own valuation. The equation
specifying the optimal bid in Eckbo and Thorburn is exactly analogous to the equation
for an optimal reserve price. Eckbo and Thorburn, examining again the Swedish data,
find strong results for the over-bidding theory and no evidence that auction prices are
affected by industry-wide distress or business cycle downturns. They also demonstrate
a surprising degree of competition in the automatic bankruptcy auctions, and that auc-
tion premiums are no lower when the firm is sold back to its own owners. Overall, their
evidence—which is the first to exploit directly the cross-sectional variation in auction
prices—fails to support either fire-sale arguments or the notion that salebacks are non-
competitive transactions.
Auction theory has also been applied to study the question of optimal bankruptcy
procedures.Hart et al. (1997)propose an ingenious three-stage auction process for
bankrupt companies. The first stage solicits cash and non-cash bids for the firm, while
the second and third stages determine prices and ownership of so-called “reorganiza-
tion rights”. Reorganization rights are new securities which consolidate all the various
existing claims on the firm’s assets. This proposal differs fromAghion, Hart and Moore
(1992)in that there is a public auction (the third auction) for the reorganization rights.
The purpose here is to reduce any inefficiencies caused by liquidity constraints in deter-
mining prices and allocations of the new securities which replace the old claims.
Rhodes-Kropf and Viswanathan (2000)extend the limited work done on non-cash
bids in auctions discussed previously. While theory such asHansen (1985a, 1985b)
shows that non-cash bids such as equity can increase sales revenue, non-cash bids are
themselves subject to uncertain valuation. Building on these basic insights, Rhodes-
Kropf and Viswanathan show that in any separating equilibrium, a security auction (the
means-of-payment is a security the value of which depends on the bidder’s type) gen-
erates higher expected revenue to the seller than a cash auction. The reason for this is
that in a security auction, the low types have a greater gain from mimicking the high
types, so to separate, the high types have to bid more. However, some securities will not
separate the bidders. The authors show that there is no incentive compatible separating
equilibrium with stock alone. Debt bids, or a minimum debt requirement, can achieve