In the past, some of studies would like to research the mutual fund return persistence into the causes of performance and they find that short-term mutual fund persistence is largely driven by fund managers accidentally holding past winner/loser stocks (Carhart, 1997) and long-term mutual fund persistence is largely driven by persistence in expense ratios (Carhart, 1992). My sample of equity mutual fund is identified themselves by the Taiwan Economy Journal (TEJ) styles (e.g. closed-end, high tech, etc.) and is often confined to trading stocks within their style, style-adjusted fund returns would be a more appropriate measure of fund manager performance. we could find the performance of the portfolio of the style-adjusted mutual funds that truly and conspicuously has advantage over the comparative group consisted of the same funds in the same period, but without weighting. The value of alpha in the portfolio is much greater than the one of that comparative group.

none

This study examines several issues related to mutual fund herd behavior. First, a unifying and consistent framework for measuring herd behavior is developed. This framework generates portfolio-level measures for each fund manager over each quarter, and relates herd behavior to other aspects of portfolio dynamics. Simulations indicate significant and persistent non-random herd behavior. Second, mechanisms that potentially underly herd behavior are tested. Empirical results indicate that herding funds tend to i) change their holdings towards levels similar to peers, ii) have less experienced managers, and iii) underperform their peers. These results are consistent with a career concerns theory of herding. Third, the impact of mutual fund herding on stock liquidity is examined. Empirical results indicate that herd behavior can lead to correlation in stock-level liquidity. / text

In this dissertation I examine mutual fund performance from the points of view of three distinct, but interrelated parties: individual investors, financial advisors, and the boards of directors of mutual fund companies. In the first essay, "Comparing Fund Flow Sensitivity for Load and No-Load Funds Under Different Market States," I compare the flow-performance sensitivity of no-load funds and the three main classes of load fund shares, assuming investment advisors are more likely to guide the decision-making process of load fund investors. I find that load investors are more sensitive to raw fund return than are no-load investors. The flow to performance relation increases during good market states, but portfolios formed from the top performing funds after good market years actually tend to underperform during the following three years. In the second essay, "Mutual Fund Performance and Board Characteristics Relating to Manager Terminations," I examine the timing of the decision to replace fund managers. I find that while returns and flows improve for those funds which replace their managers, very similar improvement is found in funds which do not terminate their managers. However, for mutual fund boards which do choose to replace a poorly performing manager, stronger board governance characteristics are associated with a greater probability of early replacement. In the third essay, "Mutual Fund Performance in Extreme Market States," I examine performance of actively managed mutual funds separately for good and bad states of the market to test whether mutual funds perform differently under different market conditions. I find that the sample of funds performs 2.3 percentage points better in good states over bad on a risk adjusted basis. I also analyze the performance of mutual funds by assuming individual funds are part of a larger, more complete portfolio. The performance of the portfolios closely matches that of the individual funds.

none

This dissertation consists of three stand-alone but interrelated empirical studies investigating various aspects of the well-documented ??home bias anomaly?? in international investment. The findings help to understand the international investment allocation behaviour of mutual funds as well as their implications for asset pricing and mutual fund evaluation. The first study investigates the roles of various firm attributes that encapsulate the deadweight costs in determining firm level investment bias. The main findings suggest that firm characteristics related to transaction costs, corporate governance and information asymmetry create significant barriers for fund managers. In addition, foreign funds are more constrained than domestic funds by information asymmetry, even in developed and liberalized markets. Moreover, this study stylises the international investment allocation model in Cooper and Kaplanis (1986) with a quadratic cost function, which reveals the marginal influence of market level deadweight costs on the relationship between firm characteristics and investment bias. It is found that when market level cross-border barriers are exacerbated, as in the case of emerging and restricted financial markets, foreign fund managers become more sensitive to market level deadweight costs and ignore firm characteristics. In general, these findings imply that the market level ??home bias anomaly?? is an outcome of the complementary effects of investment barriers at both firm and market levels. The second study examines the role of firm level investment bias in predicting future stock returns. It is found that both firm level foreign and domestic biases contain valuable information with respect to firm prospects. However, domestic bias is more informative than foreign bias in terms of subsequent stock returns, partially because of information asymmetry. The third study explores the determinants of fund level investment bias and its ability to predict fund performance. It is found that fund portfolio attributes determine fund level investment biases after controlling for market and fund investment objective specific effects, and fund level investment bias is positively related to fund performance due to lower deadweight costs. Moreover, good macroeconomic environments foster the development of the mutual fund industry.

This dissertation contains two essays in empirical finance. The first essay studies the mutual fund industry, and the second essay looks into the stock market. Both studies provide insights in the underlying mechanism of some asset return patterns identified from the data currently available. The first essay investigates the sources of a recently identified performance pattern in mutual funds. Specifically, actively managed mutual funds, in general, underperform a passive benchmark; however, some recent studies find they, in fact, outperform the benchmark in bad economic states. I examine whether a state dependent risk shifting behavior of mutual fund managers contributes to this performance difference across states, and find supportive evidence. As shown in prior studies, the risk shifting behavior is motivated by a non-linear flow-performance relationship. Using a piece-wise linear regression, I demonstrate that the non-linearity exists mainly in good states; whereas in bad states, the flow-performance relationship is close to linear. Thus, non-zero risk shifting incentives are only expected in good states. I empirically measure these incentives in good states, and show that managers do react to the ``gambling'' (i.e., positive) incentives. In addition, higher ``gambling'' incentives are found to be associated with lower fund performance. The second essay, based on joint work with Hai Lu and Kevin Wang, examines how stock price shocks in the absence of public announcement of firm specific news affect future stock returns. We find that both large short term price drops and hikes are followed by negative abnormal returns over the subsequent twelve months. The pattern of asymmetric drifts, the return continuation for negative shocks versus the return reversal for positive shocks, is puzzling. We explore whether investor disagreement can explain the puzzle and find that the evidence is consistent with predictions of disagreement theory. Moreover, price shocks with public news disclosures are followed by weaker drifts, suggesting that reduction of information asymmetry from public disclosures mitigates disagreement-induced overpricing.

This dissertation contains two essays in empirical finance. The first essay studies the mutual fund industry, and the second essay looks into the stock market. Both studies provide insights in the underlying mechanism of some asset return patterns identified from the data currently available. The first essay investigates the sources of a recently identified performance pattern in mutual funds. Specifically, actively managed mutual funds, in general, underperform a passive benchmark; however, some recent studies find they, in fact, outperform the benchmark in bad economic states. I examine whether a state dependent risk shifting behavior of mutual fund managers contributes to this performance difference across states, and find supportive evidence. As shown in prior studies, the risk shifting behavior is motivated by a non-linear flow-performance relationship. Using a piece-wise linear regression, I demonstrate that the non-linearity exists mainly in good states; whereas in bad states, the flow-performance relationship is close to linear. Thus, non-zero risk shifting incentives are only expected in good states. I empirically measure these incentives in good states, and show that managers do react to the ``gambling'' (i.e., positive) incentives. In addition, higher ``gambling'' incentives are found to be associated with lower fund performance. The second essay, based on joint work with Hai Lu and Kevin Wang, examines how stock price shocks in the absence of public announcement of firm specific news affect future stock returns. We find that both large short term price drops and hikes are followed by negative abnormal returns over the subsequent twelve months. The pattern of asymmetric drifts, the return continuation for negative shocks versus the return reversal for positive shocks, is puzzling. We explore whether investor disagreement can explain the puzzle and find that the evidence is consistent with predictions of disagreement theory. Moreover, price shocks with public news disclosures are followed by weaker drifts, suggesting that reduction of information asymmetry from public disclosures mitigates disagreement-induced overpricing.

none

none