可預期的貨幣政策對股市的影響已不明顯,故近年來有越來越多的學者探討 非預期的貨幣政策對股票報酬在不同景氣狀態下的衝擊。國外的研究多發現緊縮 較寬鬆的政策有效,且在蕭條時期較在繁榮時期有效,但台灣的文獻則顯示寬鬆 和在蕭條時期的政策較有效。本文採用三種貨幣供給量(M1a、M1b 及 M2)、重貼 現率(rediscount rate)、金融業隔夜拆利率(overnight interbank call-loan rate)等政策變數,和固定移轉機率(fixed transition probability)與隨時間變動的移轉機率(time-varying transition probability)的馬可夫轉換模型(Markov-switching model)來 分析非預期的貨幣政策在不同景氣狀態下對台股報酬的影響,並測試三種可能的 假說,以瞭解投資人對非預期貨幣政策的解讀。另外,本文將研究時期區分為可 最早收錄不同政策變數的時間至 2012 年、及 1998 至 2012 年兩種,前者包含了台灣經濟發展的全部時期,後者為彭淮南先生擔任中央銀行總裁的期間。實證結果顯示,台灣的非預期貨幣政策變動對股票報酬的確具有不對稱的效果。在兩段 研究期間下,非預期的貨幣供給變動對股票報酬的不對稱影響大致相同,皆於景 氣蕭條時較顯著,與過去的研究發現一致。相對地,非預期的重貼現率和隔夜拆 利率變動對股票報酬的不對稱影響在兩段期間並不相同,但其效果較貨幣供給量 的效果小。
Since the effect of expected monetary policy becomes insignificant, more and more researchers have tried to explore the effect of unexpected monetary policy on stock returns under the expansion and recession states of an economy. Previous studies discover that the tight monetary policy is more effective than the easy one using data of foreign countries, and the unexpected policy employed during the recession period is more efficacious. However, the asymmetric effects found using Taiwan’s data are quite the opposite. This research uses policy variables of M1A, M1B, M2, rediscount rate and overnight interbank call-loan rate, as well as Markov-switching model with both fixed and time-varying transition probability to investigate asymmetric effects of unexpected monetary policy on Taiwan’s stock returns under different economic states. Moreover, three hypotheses are tested to help us understand how investors interpret the effects of unexpected monetary policy. This research analyzes the asymmetric effects in two sample periods. The first period basically contains the whole time span of Taiwan’s economic development, and the second period from 1998 to 2012 has Mr. Huainan Peng being the president of our Central Bank. The results show that unexpected monetary policy indeed has asymmetric effects on Taiwan’s stock returns. The effects using money supply variables are pretty similar in the two periods, and are consistent with the findings in the literature. When variables of rediscount rate and overnight interbank call-loan rate are used, the effects of unexpected monetary policy on stock returns are different in the two periods. Nevertheless, these asymmetric effects are smaller than those using money supply variables.