The case for multi-unit single-run descending-price auctions

We show that a multiple-unit descending-price auction in which the clock is not reset after each sale may be faster and yield more stable prices than an efficient alternative, thus providing sellers a rationale for using it in practice.

► We analyze a multiple-unit descending-price auction in which the clock is not reset after each sale. ► This is not efficient and yields smaller revenue to sellers than the standard efficient mechanism. ► We use numerical methods to compute average prices and revenue for a family of distribution functions. ► We show that the single-run auction is faster and less volatile in prices. ► This explains why impatient or risk averse sellers may prefer this method as observed in real-markets.