本文強調新上市公司股票長期報酬的衡量。Ritter(1991)及Loughran and Ritter(1995)均指出新上市公司股票長期報酬不佳,本文利用傳統的市場調整報酬(market adjusted return)、調整市場模型報酬(adjusted market model return)及多因子模式配合Jensen's alpha所衡量的報酬來探討台灣新上市公司股票的長期報酬行為。我們發現新上市公司股票之原始報酬(raw return)、市場調整報酬或Fama-French三因子模式之報酬均有一定的報酬型態,這樣結果並不符合隨機漫步假說。然而,以四因子(Fama-French三因子加上動能因子)所衡量的長期報酬並沒有一定的報酬型態。由於各種不同的衡量模式會導致不同的長期報酬結果,我們認為學者對於新上市公司股票之長期報酬行為下結論時必須審慎。
This paper focuses on the measure of the long-run performance of initial public offerings (lPOs). Ritter (1991) and Loughran and Ritter (1995) argue that the long-run performance of IPOs is worse than that of the overall market or that of the firms with the same size and age. Besides the traditional market-adjusted returns and adjusted market model returns, this paper employs multi-factor models as well as Jensen's alpha to measure the long-run performance of IPOs. Our results show that the long-run performance of IPOs measured by raw return, market-adjusted return or Fama-French three-factor model follows a specific pattern which contradicts to random walk hypothesis. However, the IPO long-run performance measured by a four-factor model (Fama-French three factors plus momentum factor) 的similar to that of a seasoned security. The empirical findings suggest that various measurement models lead to various long-run performance of IPOs. Therefore, researchers must be cautious with the empirical findings of IPO long-run performance since it is very sensitive to the measurement models.