Conditioning Factor Exposure on Economic Regimes: Evidence from Momentum Investing

This article proposes a regime-based framework for dynamic factor allocation and illustrates its application using momentum investing in large-cap US equities. A central challenge in factor investing is that factor returns are highly state-dependent: Unconditional exposure to factors such as momentum deliver little incremental value over market benchmarks in recent decades precisely because profitability is intermittent rather than persistent. The author develops a gated factor strategy that conditions exposure to the momentum winner portfolio on market regimes identified using lagged macroeconomic, sentiment, and policy uncertainty indicators, and allocates to the market benchmark otherwise. Machine learning classifiers are employed as a flexible tool for regime identification rather than as the primary contribution. The gated strategy substantially improves risk-adjusted performance and reduces drawdowns relative to both the market benchmark and a buy-and-hold momentum strategy, achieving an annualized Sharpe ratio of 1.56 out of sample compared with 0.96 for the market benchmark.