| This thesis explores the empirical connection between macroeconomic fundamentals and asset pricing. The central question I try to answer is: what are the real, aggregate macroeconomic risks that drive expected stock returns over time and across assets? Empirical studies show that price-based predictors have limited power in forecasting future stock returns. One possible explanation is that these fixed predictors fail to sufficiently capture changes in economic conditions over time. To address this problem, I construct macro indices from a large number of economic series as quantitative descriptions of economic conditions and adaptively choose optimal indices to predict stock returns. I find that adaptive macro indices explain a substantial fraction of the short-term variation in future stock returns and beat both the historical average and commonly used predictors in real-time prediction. In addition, investment strategies exploiting the forecasts based on macro indices result in greater profits than a benchmark buy-and-hold strategy. The results remain robust under scenarios of varying transaction costs and risk tolerance. The composition of these macro indices suggests that particular sectors | interest rates, price indices, housing, and employment | contribute to the predictability of the equity premium. Furthermore, I creates new state variables from predictive macro indices to capture dynamic economic conditions. I show that the innovations in interest rate and housing are priced risk factors in explaining the cross section of stock returns. In the presence of these innovations, both Fama-French factors and innovations in dividend yields, term spread and default spread lose much of their ability to explain the cross-sectional variation. |
