“A Study of the Effects of Competition in the Tax-Exempt Bond Market”, 1971-07-01 (; similar):
This is a study of the effects of competition upon underwriting costs, reoffering yields [the yield to maturity at which an underwriter offers to sell bonds to investors], and the financing costs to issuers of tax-exempt bonds.
It provides estimates of the marginal effects of changes in the degree of competition, as measured by independent bids submitted by underwriters syndicates, upon the terms of newly issued tax-exempt bonds holding constant default risk, issue size, level of interest rates, etc.
The paper is of theoretical interest because it applies Stigler’s theory of information to the explanation of phenomena—in particular, the behavior of reoffering yields—that cannot be explained with the neoclassical model of competition which implicitly postulates that information is a free good.
[1991 writes:
Table 2: Number of Bidders and Underwriter Spread No. of Bidders Underwriter Spread 1 $15.74 2 $12.64 3 $12.36 6 $10.71 10 $10.23 A more specific illustration of the effect the number of rivals has on price can be found in Reuben Kessel’s study of the underwriting of state and local government bonds. Syndicates of investment bankers bid for the right to sell an issue of bonds by, say, the state of California. The successful bidder might bid 98.5 (or $985 for a $1,000 bond) and, in turn, seek to sell the issue to investors at 100 ($1,000 for a $1,000 bond). In this case the underwriter “spread” would be 1.5 (or $15 per $1,000 bond).
In a study of thousands of bond issues, after correcting for size and safety and other characteristics of each issue, Kessel found the pattern of underwriter spreads to be as shown in Table 2:
For 20 or more bidders—which is, effectively, perfect competition—the spread was $10. Merely increasing the number of bidders 1 → 2 was sufficient to halve the excess spread over what it would be at the 10-dollar competitive level. Thus, even a small number of rivals may bring prices down close to the competitive level. Kessel’s results, more than any other single study, convinced me that competition is a tough weed, not a delicate flower.]