The Weighting Bias

Abstract

I develop a framework to reconcile the conflicting results on long-run IPO performance based on the matching firm approach and the calendar time portfolio approach. The approaches differ in their weighting schemes: the former gives equal weight to each firm and the latter to each period. While IPOs exhibit lower buy-and-hold returns than their matched counterparts, portfolios of IPO and matched firms generate alphas of similar magnitudes. This discrepancy stems from a weighting bias inherent in the matching firm approach, driven by non-random IPO firm survival. My findings underscore the crucial role of the weighting bias when evaluating long-run performance.

Keywords: weighting bias, long-run firm performance