UC San Diego

This dissertation seeks to better understand the underlying factors driving financial performance and economic activity in international markets. The first chapter "Predictability of Growth in Emerging Markets: Information in Financial Aggregates" tests for predictability of output growth in a panel of twenty-two emerging market economies. I use pooled panel data methods that control for endogeneity and persistence in the predictor variables to test the predictive power of a large set of financial aggregates including valuation measures, interest rates, and capital flows. I find empirical evidence that stock returns, portfolio investment flows, the term spread and default spreads help predict output growth in emerging markets. In particular, large capital inflows predict subsequent high GDP growth as do high term spreads. Conversely, higher default spreads on emerging market government debt signals lower future GDP growth. Results also suggest that the performance of global aggregates such as commodity markets, a cross-sectional firm size factor, and returns on the market portfolio contain information about the future state of the economy. I benchmark my results against the US and find that there are differences in information flows and the role of capital markets in predicting economic growth. My analysis extends previous findings in the macro-finance literature on the links between the real economy and financial market performance. Within emerging markets, a largely unexplored area of research is related to the study of mutual funds performance. In my second chapter, "Emerging Market Mutual Fund Performance and the State of the Economy" I propose a set of asset class specific predictive variables and exploit them in order to identify those funds that outperform the market in different phases of the economic cycle. I employ a comprehensive survivorship-bias free universe of global and regional emerging market funds and use a Bayesian framework that incorporates predictability in manager skills (stock selection and benchmark timing skills), fund risk loadings and benchmark returns by exploiting ex-ante business cycle related state variables. Results provide empirical evidence of return predictability and the economic value of active management in emerging markets. My final dissertation chapter studies market integration and segmentation and their effects on return predictability. In "Mutual Fund Return Predictability in Partially Segmented Markets" (co-authored with B. Gillen, A. Timmermann and R. Wermers) we generalize existing models for Bayesian asset selection by considering both integrated and partially segmented market models. We find that regional state variables can be used to identify a significant time-varying alpha component among a large sample of funds with a pan- European, European country, or European sector focus. Specifically, the default yield spread, term spread, dividend yield, short interest rate and market volatility, as well as macroeconomic variables tracking consumer price inflation and growth in industrial production prove valuable in identifying, ex-ante, funds with superior performance. Our analysis also suggests that allowing for segmentation in market risk factors enhances risk-adjusted performance