Three essays in behavioral finance, asset pricing and macroeconomics

This dissertation proposes a novel resolution for some outstanding puzzles in the financial literature. The first puzzle is the asset allocation puzzle. Conventional portfolio choice theory typically advises investors to shift their portfolio composition towards relatively safe assets as they age. Empirical evidence shows that the fraction of financial assets which individuals hold in stocks varies in a hump-shaped pattern over the life span. The second puzzle the limited stock market participation puzzle. Conventional portfolio choice theory predicts that most or all individuals should hold stocks, given the high historical equity premium. Contrary to this, 75 to 80 percent of United States households do not hold stocks directly. The third puzzle is referred to as the international diversification puzzle. Despite the obvious merits of diversification, we observe that most investors hold most of their portfolios domestically. I argue that resolutions to the three puzzles may lie in a model of optimal asset allocation over the life cycle of an individual with a prospect theory-based utility function. Prospect theory has been found to provide one of the few acceptable solutions to the equity premium puzzle and other behavioral anomalies. I find that prospect theory offers a solution to the asset allocation puzzle and the international diversification puzzle. In particular, the optimal share of the portfolio held in domestic stock under prospect theory follows a hump shaped pattern over the life span while no international investment is observed. Interestingly, prospect theory seems to resolve the stockholding puzzle as well. Unlike the conventional consumption-based approach, it predicts the existence of non-stockholders as well as stockholders, in accordance with observed behavior.; The dissertation also explores the interaction between business cycles and economic growth, two areas which have each generated voluminous literatures, but seldom have been analyzed in tandem. The modest number of studies that have tried to connect the two have mostly considered the effect of the variance embodied in business cycles on the economic growth trend. These papers generated varying results. This work departs from previous literature by proposing theoretically and indicating empirically a positive connection between serial correlation in business cycles and economic growth. When one accounts for serial correlation in business cycles, the contradictory findings of earlier studies regarding the effect of the variance in output on the growth rate of output can be reconciled.