本研究藉由財務研究中重要的十個因子為配置標的,透過六種投資組合建構模式,包含馬克維茲、Black-Litterman、風險平價、倒數加權、極大化分散加權和等權重方式,分析投資組合績效表現。我們觀察因子報酬週期性表現,與不同投資組合建構模式在牛熊市和景氣循環狀況下的績效,以及任一時點進場投資勝過大盤(S&P500)的機率。研究顯示,風險平價模型在牛市期間有最佳的詹森指標,而且長期而言夏普值最佳; 馬克維茲模型所建構的投組,景氣衰退期間詹森指標最好且長期累積報酬最佳,其持有期間越長,勝過大盤的機率越高。
This paper takes ten financial anomalies that are empirical to be our sample’s underlying assets. We then adopt six different portfolio constructions, including Markowitz Mean-variance, Black-Litterman, Risk Parity, Inverse weight, Maximum Diversification, and equal weight, to build up portfolio returns, and observe the factors’ cyclical performance, the conditional performances of the six portfolios in bullish, bearish, and recovery markets, and winning over the benchmark (S&P500) assessment by buying and holding portfolios for different holding period lengths regardless of which month a position is initiated. This study shows that the Risk Parity portfolio has the best Jensen’s alpha in bullish markets and the highest Sharpe ratio in the long run among the six portfolio construction methods. The Markowitz approach has the best Jensen’s alpha during contraction periods and the highest cumulative returns in the long term, and that a longer holding period has a better chance of winning over the benchmark (S&P500).