An analysis of potential Treasury auction techniques

Federal Reserve Bulletin, June, 1992 by Vincent Reinhart

Support for the second-price scheme is stronger than the balancing of these welfare triangles would suggest. Those analysts working with explicit models of bidder behavior in a Treasury-like format, rather than with reduced-form demand schedules, typically find that a second-price scheme does produce higher revenue for the seller. Further, in 1962 Milton Friedman made a persuasive argument that revenue would increase.[12] Dealers devote considerable energy to the auction only to sell those securities almost immediately to customers--and most profit from doing so. Part of the resources devoted to that distribution could be appropriated by the Treasury if it could directly deal with those customers. A second-price auction, because it is less penalizing to the aggressive or the uninformed, may be the best vehicle to attract those people.

The Consequences for Cornering

As seen previously, the current format reduces demand at auctions and makes it more sensitive to price in relation to the demand determined by the buy-and-hold ownership of the long-time investor. This reduction is the rational response to the Treasury's discriminating pricing: The investor shows less of his true consumer surplus to a seller whose stated intention is to seize it.

Moving to a common-price format permits demand at the auction to reflect the true nature of investor preference. With no friction, investors can bypass the dealer intermediaries and bid directly, sharing the resulting savings with the Treasury. Viewed in terms of the three-figured determination of Treasury prices, second-price awards would make the auction demand curve identical to the secondary market demand curve (diagram 8). Against this backdrop, the cornerer of an auction would place surprising bids that shift the demand schedule from Demand to Demand'. The horizontal distance of that shift represents the cornerer's awards, or the extent to which secondary market supply can be restricted. As seen in the tight panel of the figure, however, the investors who are unwilling to pay the auction price will be unwilling to pay the secondary market price. Now the cornerer acting as a discriminating monopolist, rather than maximizing profit, minimizes loss (the shaded triangle). Clearly, one cannot profit from cornering a market with invariant demand, because one ultimately must sell the security to those from whom it was bid away. In this simple world, cometing would be eliminated by the removal of the potential for profit.

This result, however, requires that the switch in auction technique completely unify the primary and secondary markets. Even after the adoption of common-price awards, presence at auctions may still be limited to a segment of the investor populace, perhaps to those who are more sensitive to price. Those who sold short in the when-issued market want quickly to cover their positions at the auction. Also, participants at an auction face uncertain outcomes, since they may not be awarded securities if they have not cast their bids appropriately. Those particularly averse to this quantity risk may well delay purchase to secondary trading. Most important, direct bidding requires incurnng the fixed costs of ensuring payment and arranging for the placement of bids--the prospects for which depend on the pace of automation and the nature of regulation. As a result, the infrequent purchaser may remain in the secondary market. In other words, advocates of this format assume that dealers exist solely to shade bids because of the Treasury's discriminatory pricing. If, however, dealers provide any other service in the distribution of securities, then a gap remains between the demand schedules of the auction and the secondary market. A sufficiently large gap represents an opportunity for manipulation. Indeed, second-price awards might encourage strategems should differences between primary and secondary markets remain. A would-be manipulator could place bids for a substantial fraction of an issue well above the market consensus, and thus ensure awards, but pay only that price required to allocate the remaining portion of securities to his or her unsuspecting competitors.