Rational Expectations or Behaviorally Inefficient Markets

Abstract

In this dissertation, I examine the rational investment hypothesis, postulated by classical theories, in mutual fund and portfolio management settings. My first two essays focus on mutual fund investors. I study whether mutual fund investors display racial or ethnic prejudices, which can be observed by mutual fund flows. I hand-collect data on characteristics of mutual fund managers in addition to their names and photographs. I use machine learning algorithms from computer science literature to calculate the probabilistic race from photographs and the probabilistic ethnicity from names. In my third essay, I construct a portfolio comprising of the small growth firms and demonstrate that firms in this portfolio have heterogenous financial distress risk.

In the first essay, I examine investor rationality by examining the relation between ‘foreign-sounding’ names of the mutual fund managers and fund flows. I use MTurk survey as well as Name Prism to identify ‘foreign-sounding’ names of the mutual fund managers. I try to improve the accuracy of existing survey methodology by enforcing certain constraints from fields of sociology and onomastics. I also suggest an objective measure by identifying ethnicities using Name-Prism algorithm. I do not find any evidence of in-group bias in fund flows.

In the second essay, I examine investor rationality by investigating the racial (photo based) discrimination and the ethnic (name based) discrimination. I use machine learning algorithms to identify the probabilistic race (photo based) and the probabilistic ethnicity (name based) of the managers. After controlling for fund characteristics, flows, performance as well as other manager characteristics, I do not find any evidence of racial (photo based) or ethnic (name based) discrimination in fund flows.

In the third essay, I attempt to incorporate the financial distress risk as an explanatory measure only to correct any mis-estimation of returns by the Fama French 3-factor model. I demonstrate that the abnormal return, as represented by the intercept of the 3-factor model for the universe of stocks with lower market capitalization and low ratio of book equity to market equity, can be attributed to the financial distress risk, at least in part. Due to the heterogeneity of financial distress risk in this portfolio of small growth stocks, I demonstrate construction of profitable long-short portfolios by sorting on financial distress risk within the universe of small growth stocks.