I study various aspects of mutual funds in my thesis. These are divided over four chapters. The first chapter is an introduction to the thesis and sets out an executive summary of my research. The second to fourth chapters each deal with a new concept. The second chapter shows that the sensitivity of an investor's reaction to a mutual fund's recent performance increases with the fund's historical performance. Put differently, bad (good) performance combined with a good-history for a fund results in a greater fraction of capital outflows (inflows) relative to a fund with a poor past history. The evidence is puzzling as we would expect investors to stick with a fund having a good-history, even after a single bad performance. I solve this problem using a model with investors of differing attentiveness. In equilibrium, fund owner's attentiveness increases the historical record of a fund. With this mechanism, the model can explain the higher sensitivity of outflows for higher reputation funds. The chapter is important in that it shows that return-chasing behavior is not ubiquitous. It also provides a clear evidence where the market is slow to incorporate the new information into decision making. The third chapter studies the managerial side of the mutual funds industry regarding the risk-taking behavior of the mutual funds. Mutual fund managers are compared against a benchmark or with the peers. The employment, as well as investor's capital flows, depends on how the manager fares in the competition. I present new evidence in the chapter that the exposure of a manager to these risks is heterogeneous, and manager's historical performance governs it. The evidence implies that the risk-appetite and behavior of a manager depends on his historical performance. I find strong support in the data for this hypothesis. I show that funds with poor historical performance do not boost the portfolio risk to catch up with the peers if they are lagging at the interim date. In general, the risk appetite of the poor-history manager is less driven by their interim performance. But the good-history managers respond to their midyear position and more so during the bull years. The evidence on risk-shifting is consistent with the evidence on how each incentive behaves for good and poor history managers over bull and bear phases. The fourth chapter shows that capital movement in and out of a mutual fund is more sensitive to fund performance during periods of high market volatility. I explain this result using a model where the manager has picking as well as timing skill. A volatile market presents an opportunity to generate timing value and to that extent produces speedy learning about managerial timing ability. Persistence in volatility boosts the sensitivity of flows to performance during such times. Given the counter-cyclical nature of market volatility, the model predicts that the flow sensitivity is higher during the recessions. Data supports the model prediction. The chapter provides a clear example when the trade volume (here capital flows) is linked positively with the volatility. Usually, literature has shown how the volatile periods slows the learning and hence trade volumes too. But my model indicates that there could be substantial learning going on during volatile times about critical economics parameters, mainly because those parameters are revealed only during volatile times.