本文主要探討股票報酬與各項總體經濟變數之間的因果關係。本研究方法採 Chen and Lee （ 1990 ）提出的網狀因果關係檢定法（ Nested Causality Test ）， 本文亦比 較由向量自我迴歸模型（ VAR ）分析之結果。 經實證研究得到以下結論：（ 1 ） Nested 法：就短階模型來看，除了股票報酬單向影響工業生產外，股票報酬與其它各變數幾乎皆有 回饋關係。而採 Var （ 6 ）時，貨幣供給單方向影響股票報酬且二者為正相關；股票報酬 則單向影響工業生產，二者亦為正相關；利率單向影響股票報酬，而通貨膨脹則與股票報酬 獨立。（ 2 ） VAR 法：取短階模型時，股票報酬與利率呈回饋關係，以 Var （ 6 ）檢定 時顯示股票報酬領先利率， 但二者相關性不明顯； 貨幣供給、 工業生產與通貨膨脹則與 Nested 法有相似之結論。

　　　　　This study investigates the causality between Taiwan's stock return and several macroeconomic variables, including money supply, interest rate, inflation rate and industrial production by method of Nested Causality Test (NCT). The results of NCT show that, except for industrial production, all variables have feedback relationships with stock returns of 2-to-4-month lags. With 6-month lags, except inflation, all other variables show unidirectional relationships with stock returns. The results of VAR test are similar to those of NCT except that the stock returns and interest rate have feedback relationship with 2-to-4-month lags, and the stock return leads interest rate with 6-month lags.