Why the Reserve Price Should Not Be Kept Secret

This paper considers the optimality of setting a secret reserve price in ascending auctions. Contrary to intuition, an ascending auction is no longer equivalent to a second price auction when the reserve price is secret. We determine the seller's optimal reserve price policy when the bidders' values are private and independently distributed and when the bidders are risk averse. We show that an optimal secret reserve price policy can dominate an optimal public reserve price policy when the bidders' degree of constant relative risk aversion is sufficiently high and when the seller can commit to a reserve price policy before learning her type. In contrast, a secret reserve price policy can never be part of a Bayesian equilibrium when the seller is informed.