| My thesis focuses on two well-known anomalies in finance. The first anomaly concerns violations of the mutual fund separation theorem; the second anomaly relates to the apparent momentum in stock prices.; The first chapter provides new findings that challenge the widely accepted mutual fund separation theorem of Tobin (1958) and Markowitz (1952). Contrary to this theorem, I document that in reality the ratio of investor's high-risk to low-risk stock holdings changes with her attitude towards risk. This result is an extension of the asset allocation puzzle posed by Canner, Mankiw and Weil (CMW (1997)), in which they point out that financial advisors generally recommend a lower ratio of bonds to stocks for more aggressive investors. Drawing on the loss aversion literature, I propose a general model that can solve both the CMW puzzle and my extension. The key is that investors are loss averse and care about their expected worst-case returns.; The second chapter documents empirical evidence of within-industry momentum, i.e., the fact that a zero-cost strategy to long past winners and short past losers within an industry can generate significant profits. This finding is contrary to the widely cited conclusions in Moskowitz and Grinblatt (MG (1999)) that within-industry momentum is insignificant. In this chapter, I develop a direct empirical test of within-industry momentum that incorporates some factors overlooked by MG (1999), such as the potential effect of the bid-ask spread, and I find that momentum exists in most industries for most ranking/holding investment strategies. Furthermore, I report a positive relationship between momentum magnitudes and industry concentration levels.; In the third chapter, I propose a multi-period, multi-asset Rational Expectations Equilibrium (REE) model to explain within-industry momentum and momentum in general. The key feature of the model is a borrowing constraint. This model also addresses the cross-correlations of stock returns, which are ignored by most existing momentum models. In this model, investors are heterogeneously informed and some investors have borrowing constraints. Momentum is generated by gradual diffusion of information and investors' borrowing constraints. |
