The next test examines whether the postauction price changes are systematically
related to the percentage of new issues won by banks and dealers. If
they routinely collude then it is likely that when they collectively bid less
aggressively, or when the percentage of winning bids tendered by this group
is low, the profits to the winning bids will be high. Interestingly, it is found
that the allocation to banks and dealers are negatively related to auction
profits. This result is consistent with the collusion hypothesis and suggests
that further analysis using data on bids by individual dealers is warranted.
Finally, the validity of claims that the secondary market prices of the
two-year Treasury notes issued in the May 1991 auction were manipulated is
investigated. Salomon Brothers admitted to having controlled 94 percent of