This dissertation estimates factor models with incomplete data. In this context, hidden factors map serial and cross-sectional correlations of financial and macroeconomic data. The proposed estimation methods involve two alternately applied expectation-maximization algorithms. Besides modifications of the Kalman Filter and Smoother, we utilize closed-form expressions for estimating factor moments. With factors as exogenous variables, we derive point and interval forecasts of returns, reveal their composition and develop single-market and portfolio strategies.
«
This dissertation estimates factor models with incomplete data. In this context, hidden factors map serial and cross-sectional correlations of financial and macroeconomic data. The proposed estimation methods involve two alternately applied expectation-maximization algorithms. Besides modifications of the Kalman Filter and Smoother, we utilize closed-form expressions for estimating factor moments. With factors as exogenous variables, we derive point and interval forecasts of returns, reveal thei...
»