Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Author ORCID Identifier


Campus-Only Access for Five (5) Years

Document Type


Degree Name

Doctor of Philosophy (PhD)

Degree Program


Year Degree Awarded


Month Degree Awarded


First Advisor

Nikunj Kapadia

Second Advisor

Nikolaos Artavanis

Subject Categories

Finance and Financial Management


This dissertation consists of three essays that examine investor behavior using mutual fund flows. Despite the ability to explain many puzzling phenomena in financial markets, direct tests of reference-dependent preferences of Kahneman and Tversky (1979, 1992) have mainly been conducted in experimental settings. In the first essay, we propose a novel test based on revealed preferences of real-world investors. Following Berk and van Binsbergen (2016) and Barber et al. (2016), we use the well-documented mutual fund flow-performance relationship to examine if investors have different preferences over gains and losses when they evaluate performance. We infer investors' utility function from mutual fund flows instead of structurally estimating it with experimental or field data. We find that investors place a greater emphasis on losses than gains. Moreover, institutional investors always penalize risk, yet retail investors are risk-averse only over gains but risk-seeking over losses, in line with reference-dependent preferences rather than global risk aversion.

Shumway (1997) and Ang, Chen, and Xing (2006) show that if investors place greater emphasis on losses than gains, they require a higher risk premium for downside risk rather than upside potential. Despite the intuitive appeal and theoretical justification, there is little empirical evidence regarding downside risk in asset pricing, mainly due to problems inherent in estimating downside risk. In the second essay, we argue that Berk and van Binsbergen (2016)'s approach to testing asset pricing models using the relation between investor flows and risk-adjusted returns is well suited for examining the merits of downside risk. We extend the analysis of Berk and van Binsbergen (2016) and Barber et al. (2016) by showing that investors care more about downside market risk than unconditional market risk when choosing mutual funds. We find that investors' sensitivity to downside risk increases following market crises and is more pronounced among funds with conservative investment objectives. Finally, we find that investors' response to downside risk is not subsumed by information/influence from Morningstar ratings.

Many mutual funds that employ hedge fund strategies with both long and short equity positions were launched after the 2007-2009 financial crisis and assets under their management grew rapidly over the past decade. In the third essay, we examine the performance of and flows to these funds relative to traditional mutual funds and hedge funds in order to understand the reasons underlying their recent popularity. Early studies on hedged mutual funds (e.g. Agarwal et al. (2009) and Chen et al. (2013)) document that they outperform traditional mutual funds due to superior investment strategies employed by skilled managers. However, we find that the outperformance vanishes over time as more funds compete to offer similar strategies in line with the findings of Wahal and Wang (2011) and Hoberg et al. (2018) that competition limits the ability to generate alpha. Finally, investors direct flows to hedged mutual funds, especially to market-neutral funds, when investor sentiment is low, suggesting that investors use these funds as a hedge against downside risk due to their low betas.