總體數列趨勢之去除是景氣循環研究前所必須處理的重點之一，而台灣
的相關研究中大多無法棄卻總體變數具隨機趨勢，因而必須在實證分析中對變數
取差分。本文以單根檢定來檢測台灣主要總體實質變數(RealMacro-Variables)究竟
是其確定性趨勢或隨機性趨勢，實證結果支持其大致為趨勢恆定之時間數列。故
就台灣之實質景氣循環分析而言，確定性趨勢成長確有實證上的支持。

Trend-removing is important in analyzing real business cycles. Instudies
related to this field in Taiwan, most authors cannot rejectthe hypothesis that there
exists a stochastic trend among key macrovariables, which makes data differencing
necessary in their empirical studies. This paper applies several types of unit-root tests
toexamine the issue of stationarity among major real, macro variablesin Taiwan.
Empirical findings suggest that most of the variables aretrend-stationary. Hence, we
claim that it is justifiable to assumestationary trends in examining real business cycles
in Taiwan.