DISCRIMINATORY VERSUS UNIFORM TREASURY AUCTIONS - EVIDENCE FROM WHEN-ISSUED TRANSACTIONS

We use when-issued transactions data to assess the Treasury's current experiment with uniform auctions. When-issued volume is higher under u niform as compared to discriminatory auctions, suggesting a higher inf ormation release, which should;reduce pre-auction uncertainty and the winner's curse. Under uniform auctions, when-issued volatility falls a fter the auction and again after the outcome announcement. The pattern is the opposite for discriminatory auctions. This is further evidence that uniform auctions increase pre-auction information and lower the short squeeze. A direct comparison of markups in uniform and discrimin atory auctions yields mixed results.