本文使用台灣加權股價指數、重貼現率及國內生產毛額作為代表市場環境之變數,並採取非線性之平滑移轉自我迴歸模型以探討歷年來IPOs家數與市場環境之關連性。過去實證研究多以線性之模型作為研究方法,然而時間序列資料在長期可能存在結構性的轉變而不被傳統線性模型所涵蓋,故本文嘗試以非線性之平滑移轉自我迴歸方法,來探討上述幾項國內總體經濟變數對IPOs家數之影響,並了解是否存在異質平滑移轉之現象。 實證結果發現,國內IPOs家數的確存在結構性轉變之現象。而上述三項總體經濟因素中,僅國內生產毛額及利率對IPOs家數有明顯的影響。在國內生產毛額變動率作為轉換變數的情況下,國內生產毛額及利率之落後期於門檻值前後的確有產生不同方向的結構性轉變。落後1、2期之國內生產毛額變動率及落後4期的利率變動率的影響減弱,而落後3、4期之國內生產毛額之影響則是增強。 This article adopted Taiwan stock index, discount rate and GDP as representative variables of market environment. It also used nonlinear smooth autoregressive model to research connections between IPOs volumes and market environment over the years. Over the past, what has been used often was looking into linear model as the research template; it is however, that time series data in the long term could cause possible structural transformation without being covered by traditional linear model. Therefore, this article will focused on using nonlinear smooth autoregressive model to attest the effect of the above variables toward IPOs volumes and at the same time testifying whether there is the existence of heterogeneity smooth transition phenomena. The final result turned out to be that there is the existence of structural transformation within our IPOs volumes. Among those three variables mentioned above, only GDP and discount rate cause major effect toward IPOs volumes. Under the condition of using GDP as transition variables, the lag of GDP and discount rate did contribute diverse structural transformation.