我們透過 PLS 方法建構總體經濟聯合指標，並成為一強大的熊市預測指標。藉由將馬可夫轉換模型應用於美國標普五百指數報酬率上，我們可以辨識出熊市期間與該期間對應之熊市機率。不論在樣本內或樣本外的檢定，總體經濟聯合指標在預測熊市機率都統計上顯著優於其他指標，包含投資人情緒指標以及其他二十五個總體經濟指標。此外，將 PLS 應用在不同市場指標時，此方法仍可以維持其預測能力。;We use the PLS method to construct an aligned macroeconomic index, which is powerful in forecasting the bear markets. By using the Markov-switching model to return of the S&P 500 index, we identify the bear market and obtain the probability of each period. No matter in- and out-of-sample test, this index statistically outperforms twenty-five macroeconomic variables and the investor sentiment index in prediction bear markets. What is more, this method also could be applied to different market indicators, and the results remain robust.
