使用狀態轉換模型進行特色反轉投資策略在歐洲區規模及價值風險溢酬的研究

Abstract

2008年全球金融危機後，造成資產交互去槓桿化及投資策略過度壓縮。在此不穩定的經濟環境更引發了不同特色投資組合的超常相關性和波動性，甚至特色投資的反轉現象。所以傳統的一致化特色投資策略可能無法再提供過去文獻所引述的長期獲利。本研究將關注在規模及價值投資溢酬的動態變化並建立一個能預警經理人反轉其特色投資策略的模型。我們將使用品質變數的向量自我迴歸模型去預測歐元區的股票市場規模及價值風險溢酬的方向變動，進而提供經理人買賣投資組合的訊息。我們利用品質變數的向量自我迴歸模型所建立的特色反轉投資策略之績效將和特色一致投資策略及使用傳統probit/logit 迴歸建立的特色反轉投資策略績效相比較。此外，預測期間也將使用１個月期和３個月期去測試模型樣本外的穩定度及考量交易成本的影響。

The 2008 global financial crisis induced cross-asset de-levering/sell-off, overcrowded investment strategies. The instable macroeconomic environment has resulted in abnormal style correlations and volatility, and sudden style reversals. Hence, the style consistent strategies may not provide the long-term benefits often assumed in the literature. This study aims to look at the performance of various asset classes (styles) and aims to build a model that can indicative to managers to switch styles. Qual VAR model will be constructed in order to generate the switching signal of size and value portfolios in the stock markets in the Euro area. The results of the rotation strategies are compared with the style consistent buy-and-hold strategies as well as the traditional probit/logit style rotation strategies. Furthermore, the forecast horizon, one-, three- and six-months signals is varied, which serves as an out-of-sample model-stability test, and account for the impact of a wide range of transaction costs.