Extrapolation Errors in IPOs
研究发现，IPO前收入增长快的公司定价更高但后续回报更差，三年年化回报低增长公司为-1%，高增长公司为-12%或-29%，且分析师预测偏差在高增长公司中最大，支持外推误差模型。
We demonstrate that relatively high revenue growth prior to an initial public offering (IPO) is associated with high IPO prices and poor returns. In particular, low‐growth IPO returns, −1% on a three‐year annualized basis compared to −12% (equal‐weighted) or −29% (value‐weighted) for high‐growth new issues. There is no evidence that the performance differentials reflect risk premia. Rather, low‐growth firms’ returns are more stable over time. Finally, while analysts’ forecasts are upwardly biased for all firms, the magnitude of this bias is greatest for firms with rapid pre‐IPO revenue growth. Overall, these results are consistent with the extrapolation errors model suggested by Lakonishok, Shleifer, and Vishny (1994).