Ch. 3: Auctions in Corporate Finance 135
size that bankers agree to provide research coverage for issuers in return for using the
bookbuilding method. What is left unstated is that issuers must be unable to buy such
research coverage on the open market at prices similar to the costs paid by investment
bankers: the authors agree that auctions would yield issuers a better price, so one must
wonder why issuers put up with an inefficient procedure simply to get a tied service.
On the empirical side of the auctions/IPO issue,Kandel, Sarig and Wohl (1999)
utilize a data set from Israel IPO auctions to document elasticity of demand and under-
pricing. The under-pricing of Israeli IPOs is intriguing, for those IPOs had their prices
set by an explicit auction mechanism. In the period 1993–1996, Israeli IPOs were
conducted much like Dutch auction share repurchases: investors submitted sealed-bids
specifying prices and quantities, a demand curve was determined, and a uniform price
was set at the highest price for which demand equaled the supply of shares available.
Kandel, Sarig and Wohl document some elasticity of demand for the reported bids: the
average elasticity at the clearing price, based on the accumulated demand curves, was 37
(relatively elastic). Interestingly, even in these IPO auctions, there was under-pricing:
the one-day return between the auction price and the market trading price was 4.5%.
Another interesting feature of the Israeli auctions is that after the auction but before the
first day of trading, the underwriters announce the market clearing price corresponding
to the offered quantity, as well as the oversubscription at the minimum price stipulated
in the auction. This essentially means that the investor can estimate the price elasticity
of demand based on two points on the demand curve. The authors find that the abnormal
return on the first day of trading is positively related to the estimate of the elasticity. The
authors argue that this reflects greater homogeneity in the estimates of value on the part
of the participants in the auction; this is “good news” either because it implies greater
accuracy of information about future cash flows and thus leads to a lower risk-premium
demanded by investors, or because it signifies greater “market depth” and hence greater
future liquidity.
Kerins, Kutsuna and Smith (2003)examine IPOs in Japan in the period 1995–1997, a
time when Japanese firms had to use a discriminatory (bidders pay the amount of their
bid) auction to sell the first tranche of newly issued shares. This first tranche of shares
would be relatively small, and the sale by auction was restricted to outside investors
only, with further limitations on the amount that could be bought by any investor. These
restrictions could be interpreted as limiting the informational advantages of any one
bidder. Under that interpretation, it is not surprising that the authors find relatively lit-
tle “underpricing” of the shares for the auction tranche: for all the issues, the auction
proceeds were only 1.6% below what proceeds would have been at the final aftermarket
price. The second stage of the Japanese process was a more traditional fixed-price offer,
and there was considerable underpricing of shares at this stage. While this might sug-
gest that the auction was a better choice of mechanism, one must recognize that costs of
a larger auction (to sell the entire issue) could well be larger than costs of just the first
tranche.