116 S. Dasgupta and R.G. Hansen
depend on the quality of the target’s assets, or if the bidders plan to bundle these assets
with other assets that they own and eventually sell these assets.^27 ,28
Other complexities also arise when applying the auction framework to the analysis of
takeover bidding. Bidders could bid for the company, or they could bid for a fraction of
the company’s shares. Different regulatory regimes permit different types of bids. Bids
could be exclusionary, discriminatory, conditional, and so on. Bids can be in cash, or in
shares of the target company. Bidders may have different toeholds, and they might have
different degrees of expertise in the target industry (a factor that could affect the degree
of information conveyed by their bids in common value environments). Finally, if the
bidders are competing with each other in the same industries, then the outcome of the
auction may impose externalities on the bidders. As we will argue below, while existing
takeover models, drawing on auction theory, have evolved to deal with these many of
these complexities, significant gaps still exist in the literature.
4.2.3. Auctions versus negotiations
Several papers use auction theory to further refine our theoretical and empirical under-
standing of the auction process in corporate takeovers. Starting at the most basic level,
Bulow and Klemperer (1996)show that in an English auction, it is always better to have
N+1 bidders in a formal auction than to haveNbidders but with a follow-on (optimal)
negotiation between the winning bidder and the seller. IfN =1, this shows that it is
better to have an auction with two bidders than to sell by posting a reserve price.^29 Ve r y
simply, the auction process is extremely efficient at extracting value from the high bid-
der, more so than even an optimally conducted negotiation. This theoretical result does
conflict with a stylized fact that companies do frequently avoid auctions and instead
negotiate with just one buyer (Boone and Mulherin, 2006a).
4.2.4. Pre-emptive bidding
Fishman (1988, 1989)considers models where one bidder has incentive to make a “pre-
emptive” bid. In the main model of Fishman, a first bidder has incentive to put in a high
bid that discourages the second bidder from bidding. The reason for this is that a high
bid can signal a high valuation on the part of the first bidder, and a second bidder will
(^27) Models of takeover bidding, when making common values assumption about the target’s “true worth”,
have often tended to assume away the free-rider problem. If a bidder obtains a large majority of the shares, she
may be able to “freeze out” the remaining minority shareholders. Also, the loss of liquidity on any remaining
shares can have the same effect as “dilution” (seeGrossman and Hart, 1980) that reduce the post-takeover
value of the minority shares.
(^28) Betton and Eckbo (2000)show that the average number of days from the initial tender offer bid to the
second bid is 15 days (counting only auctions with two or more bids). They suggest that this very short period
is evidence of correlated values. Of course, the vast majority of all cases develop a single bid only, which may
be taken as evidence of private (uncorrelated) values, or preemptive bidding (see below).
(^29) SeeKrishna (2002)for an analysis of this case.