The three chapters in this thesis offer some novel explanations for stock return predictability, and examine firm managers' responses to the potential inefficiencies in the financial market.

In "A Flow-Based Explanation for Return Predictability," I propose and test a flow-based explanation for three important empirical findings on return predictability. Since mutual fund managers generally scale up or down their existing positions in response to investment flows, investment flows can cause significant demand shocks in individual stocks. Meanwhile, since fund flows are predictable from past fund performance and past fund flows, flow-induced price pressure is also predictable. This paper further shows that such flow-based return predictability can fully account for mutual fund performance persistence and the "smart money" effect, and can partially explain stock price momentum.

In "Attracting Investor Attention through Advertising," I provide empirical evidence that managers adjust firm advertising strategies in order to influence investor behavior and short-term stock prices. First, this paper shows that increased advertising spending is associated with a contemporaneous rise in abnormal stock returns, which is then reversed in subsequent years. Moreover, increased advertising is related to the stock purchases of individual investors. Finally, there is a significant rise in firm advertising expenditures prior to insider sales and seasoned equity offerings. This large increase is followed by an even larger decrease in advertising expenditures in the subsequent year. This pattern of advertising expenditures suggests that managers are exploiting the return effect induced by advertising to the benefit of the existing shareholders and themselves.

In my joint work with Darwin Choi, "A Test of the Self-Serving Attribution Bias: Evidence from Mutual Funds," We study the dynamics of investor overconfidence and find that even sophisticated investors like active mutual fund managers are susceptible to the self-serving attribution bias. Using Active Share as a proxy of confidence, we find that a typical fund manager tends to increase his self-perception of ability to a larger extent after good performance than to decrease it after poor performance. Moreover, overconfidence leads to poor future performance, most of which is driven by managers' sub-optimal portfolio choices rather than excessive trading.