本研究探討 Chen, Roll, and Ross (1986) 的 CRR 五因子模型是否適用於台灣上市櫃股票市場。該模型包含了五個總體經濟因子:工業生產成長率、未預期 通貨膨脹率、通貨膨脹率變化、利率期限結構利差、及信用利差。我們透過規模、價值、及動能等特徵來切分股票,從中建構出 50 組依照個別股票特性分類的投資組合。最後以總體經濟五因子作為自變數、和 50 組台灣上市櫃股票投資組合作為應變數,透過 Fama and MacBeth (1973) 的時間序列與橫斷面兩階段迴歸分析及 Gibbons, Ross, and Shanken (1989) 的 GRS 檢定方法來檢定五因子模型之定價能力。結果顯示總體經濟五因子模型之調整後判定係數高達 90% 以上,亦通過 GRS 檢定,整體表現優於台股組成的 Fama and French (1993) 三因子模型。從模型估算得出之風險溢酬,我們觀察到台灣股市與總體經濟數據息息相關。
This paper investigates the performance of the Chen, Roll, and Ross (1986) (CRR) five-factor model in pricing assets in the Taiwanese stock market. The model incorporates five macroeconomic factors: the growth rate of industrial production, unexpected inflation, the change in expected inflation, term spread, and default spread. After constructing 50 testing asset portfolios based on size, value, and momentum, we examine pricing ability of the CRR five-factor model through the Fama and MacBeth (1973) two-pass regression and the Gibbons, Ross, and Shanken (1989) GRS test. Our results show that the determination coefficient of the macroeconomic five-factor model exceeds 90%, and passes the GRS test. The macroeconomic risk model also has better explanatory power than the Fama and French (1993) three-factor model constructed from Taiwanese stocks. Moreover, the estimated risk premium from the macroeconomic risk factor model reflect the strong tie between the Taiwanese stock market and the macroeconomic conditions.