本研究探討總體經濟變數是否可以作為預測熊市的指標,研究對象為12工業國,包括:比利時、加拿大、丹麥、法國、德國、義大利、日本、荷蘭、挪威、瑞典、英國及美國,本文考慮的總體經濟變數有相對市場利率、相對國庫券利率、相對政府公債利率、期限利差、通貨膨脹率、工業生產成長率、貨幣供給成長率及失業率的變動作為評價指標。資料期間為1970年代早中期至1990年代晚期的月資料。 在實證上,本文採用非參數的Bry-Boschen方法,藉由虛擬變數來區分出熊市與牛市之後,接著採用樣本內及樣本外的probit預測模型檢定,最後進行預測熊市與預測股票報酬率的比較及子樣本期間的樣本內預測結果分析。 實證結果發現,樣本內及樣本外皆指出利率和通貨膨脹率在12工業國中為最具有預測能力的指標,此項結果也可為股票市場參與者提供進場時機的策略,而以總體經濟變數預測熊市與預測股票市場報酬預測能力差不多,且在子樣本的實證分析中可以發現大部分國家子樣本與全樣本預測的結果差異不大。
This study investigates which macroeconomic variables can predict bear market in the stock market across the countries. Compared to the stock market of a single country, such as Chen (2009), this study expanded to 12 industrial countries: Belgium, Canada, Denmark, France, Germany, Italy, Japan, Netherlands, Norway, Sweden, the U.K. , and the U.S. The sample consists of monthly stock market indices and series of indicator variables such as relative money market rate, relative 3-month Treasury bill rate, relative long-term government bond yield, term spread, inflation rate, industrial production growth, narrow money growth, broad money growth and change in the unemployment rate as evaluation indicators. Monthly data start in the early-to-mid 1970s and end in the late 1990s. In this paper we use nonparametric approach Bry-Boschen method to identify recession periods in the stock market and consider both in-sample and out-of-sample tests of the variables’ predictive ability. Then comparing the bear market prediction to the stock return predictability and subsample predict analysis. Empirical evidence suggests that among the macroeconomic variables this study have evaluated, term spread and interest are the most useful predictors of recessions in the 12 industrial countries, according to both in-sample and out-of-sample forecasting performance. Furthermore, the bear market prediction is about the same with the stock return predictability that all of them using the same macroeconomic variables. Then continue to analysis the evidence of subsample, the empirical results do not change over subsample periods for most of countries.