Using Conditional Factor Performance to Analyse a Market-Beating Portfolio Strategy

Abstract

This paper revisits the literature on conditional asset pricing and portfolio management by analysing conditional returns of thirteen factors dependent on macroeconomic indicators in relevance to a market-beating equity portfolio. Fundamental-based factors – size, value, and investment, have no significant premia over the past decade, although long-term reversal, unexpected volume, and price to 52-week high factors produce return premia significant to the 5% level. On conditional asset pricing, a group of thirteen macroeconomic observables, that have appropriately been first differenced to remove unit roots, explains up to 41% of the variation on returns of significant factors. By creating a measure that ranks a market-beating portfolio’s factor exposure, the study finds this portfolio to have patterns in its ranking in key factors across U.S. equity markets. In particular, the study finds that the portfolio ranks in the 8th decile of the price to 52-week high factor. This paper finds evidence that factor investing can explain 65% of the variation in returns from a market-beating portfolio, and that interaction coefficients between macroeconomic indicators such as exchange rates, interest rates, and price growth show significant contributions to factor loadings. These results are robust to heteroscedasticity and autocorrelation controls including Prais-Winsten and Newey-West estimators.