Time-Varying Factor Allocation
研究预测信息能否帮助盈利地配置跨资产因子组合，发现基于通胀和商业周期信号调整全球因子组合可提升表现，但多数预测因子无效甚至有害，建议对基于回报预测的超额收益持怀疑态度。
This study provides evidence on whether predictive information can help to profitably allocate a cross-asset factor portfolio, covering well-known factors over the asset classes of equity, commodities, fixed income, and foreign exchange. The authors investigate the performance of a meaningful set of predictors, which they broadly divide into macro and market indicators. Their analysis indicates that tilting a global factor portfolio based on inflation and business-cycle signals can enhance performance, whereas most of the analyzed predictors do not notably improve decision making or are even counterproductive. The results are validated over an extensive out-of-sample period and under practical considerations, while surviving conservative transaction cost assumptions. In sum, the authors advise exercising skepticism when assessing outperformance based on return forecasting, given the challenges of selecting and adhering to the right variables beforehand. Nevertheless, their results suggest potential benefits of conditioning an investor’s factor allocation on fundamental macroeconomic information and motivate future research to explore this link.