The Persistence of IPO Mispricing and the Predictive Power of Flipping

Abstract

Despite underwriters' efforts to balance supply and demand in the IPO price setting mechanism, we show that the market accurately predicts (in the first-day return and volume) the direction but not the full magnitude of underwriters' pricing errors. That is, first-day winners continue to be winners on average (outperforming a size-adjusted benchmark) over the first year, and first-day "dogs" continue to be relative "dogs". A trading rule of "buy first-day solid performers" beginning at the close of the IPO's second trading day outperforms the portfolio of first-day losers by about 14% in the next year during the 1988-1995 time period. An exception to this rule is the performance of extra- hot IPOs (with a first-day return greater than 50 to 60%) which are poor one-year performers on average. We also find that a measure of flipping, the dollar volume of sell-motivated block trades as a percent of total dollar volume on the first-day, has significant power to predict future returns, lending credence to investors' "flipping" of cold IPOs as a rational strategy. Furthermore, we show that first-day flipping can be predicted from ex ante factors. Thus, we conclude that underwriters' pricing errors at both extremes (initial cold and very hot IPOs) are intentional and strategic.

JEL Classification: G12, G14