The investment management industry has grown substantially within the last ten years, and managed portfolios are currently over the $3 trillion asset level in the United States alone. In addition to traditional stock and bond funds, investors are placing funds in alternative asset classes such as commodity or managed futures funds and pools. A large number of academic studies exist that analyze the performance of managed portfolios. Even so, there are a number of issues that remain unresolved. The first essay focuses on determining the relationship between aggregate fund flows and securities returns. Previous studies on this topic do not incorporate competing asset classes, do not separately analyze equity sub-classifications, or use data during time periods when fund flows were a small fraction of what they are today. Time series and causality testwork are performed on the dynamic nature of the relationship between aggregate fund flows and market returns. We find evidence to support the hypothesis that fund flows tend to follow returns, but limited evidence to support the popular notion that fund flows "drive" markets. When aggregate equity flows are conditioned upon fund classifications, we find that the category flow determinants differ based upon risk classification. Finally, we find that the competing asset classes matter both statistically and economically in explaining aggregate fund flows for the majority of fund flow categories. The second essay concentrates on the performance of individual equity commodity trading advisors (CTAs), who focus on trading futures contracts for their clientele. To date, few studies have been performed that look formally at the performance abilities and sources of returns of individual equity CTAs. This essay uses a well-known econometric model and downside risk measures to evaluate the timing ability and performance of equity-based CTAs. In addition to examining stand-alone performance, CTA inclusion in broad-based equity portfolios is studied and compared to alternative downside risk reduction strategies, such as a protective put. Finally, performance persistence is evaluated for these individual CTAs. In general, we find equity-based Commodity Trading Advisors do not exhibit market timing or security selection skills; nor is performance consistent among investment vehicles. However, this investment class does exhibit characteristics consistent with downside risk protection, especially when compared with traditional portfolio insurance models.

This dissertation examines the impact of high frequency data in volatility measurement on the distributional properties and predictability of futures market volatility, direction-of-change probability forecasting using the dynamics of volatility and the presence of asymmetric volatility effects. Chapter 1 studies the distributional properties of returns and volatility in 33 futures markets and their implications for asset allocation, risk management and asset pricing. In particular, the focus is on realized volatility estimated from high frequency intraday returns and returns standardized by realized volatility. Returns standardized by realized volatility are approximately normal as is logarithmic realized volatility. Based on the statistical distribution analysis, time series forecasting models for logarithmic realized volatility are estimated and evaluated. In Chapter 2, I study the direction-of-change predictability in futures markets based on the dynamics of volatility and the economic value of such predictability. I extend previous research in a number of ways. First, I study direction-of-change predictability based on conditional volatility, skewness and kurtosis in a diverse selection of international currency, equity, financial and physical commodity futures markets. Second, I use the highly efficient measure of realized volatility based on high-frequency intraday returns. Third, I use the SEMIFAR model to forecast conditional volatility. Finally, I examine the economic value of sign predictability in market timing trading strategies. I find that volatility dynamics can successfully be used to forecast signs across a large number of futures markets. In low volatility periods, forecasting models using volatility dynamics generally outperform a baseline model which uses historical probability. Trading strategies based on sign forecasting through volatility dynamics produce performance comparable but not highly correlated to that of a common trendfollowing strategy. In Chapter 3, I study asymmetric volatility in physical commodity futures markets using high frequency data to construct efficient volatility measures at the daily and monthly horizons. The findings indicate that asymmetric volatility effects in futures markets are generally consistent with the size and the sign of net speculative interest in these markets. Volatility asymmetry is consistently present in certain markets, consistently absent from other markets and intermittently present and absent in a third group of markets.

The inability of traditional models to account for time-varying estimates has led to conditional models being adopted for performance evaluation. In this dissertation I use the conditional models of Ferson and Schadt [1996] and Christopherson, Ferson and Glassman [1998] and the CISDM Alternative Investments Database to evaluate the performance of hedge funds and managed futures. One of the advantages of the CISDM alternative investment database is its defunct funds component. I use both components to construct a dataset that is free of survivorship bias. I focus on four major issues related to the CISDM alternative investment database and hedge funds and managed futures. The first issue relates to the characteristics of the defunct funds component. It is well known that it consists of funds that have stopped reporting for reasons other than going out of business, although poor performance is the primary reason for disappearance. The CISDM database reports the reasons due to which funds stop reporting. I checked my quantitative results against this information and found consistency in most cases. The second issue I focus on is performance evaluation and market timing. I find that while portfolios of active funds exhibit significantly positive alphas, most dead fund portfolios do not. I also investigate the market timing ability of these portfolios. My results validate that hedge funds pursue short-volatility strategies. The third issue relates to the performance of managed futures. I use the models of Ferson and Schadt [1996] to estimate excess return alphas for 78 CTAs that had complete data for the period 1993–2003. I find that the MFSB indices that were used as proxies for the market were remarkably effective in evaluating performance of managed futures. Finally I examine out-of-sample performance. I evaluate the performance of hedge fund portfolios constructed by ranking commonly used risk measures. I find that in most cases performance of ranked portfolios vary considerable and conclude that investors should exercise caution when constructing portfolios based on the measures. I also conclude that standard deviation is remarkably consistent over time compared to other measures.

The purpose of this research is to investigate risk exposures and survival of individual hedge funds and managed futures funds both within and between strategy styles. There are two immediate new contributions: (1) a comprehensive database of hedge fund and managed futures funds comprised from the major database vendors; each database has been used singly or lesser combination by previous researchers, and (2) analysis of the variation of risk exposures of individual funds as opposed to portfolios or indices. The research focuses on three areas: (1) estimating statistical properties and survivorship bias, (2) estimating common economic/market risk factor exposures and variation, and, (3) providing evidence of survival patterns. The first part confirms previously documented properties of hedge fund returns. The results for survivorship bias estimation using the two main estimation methods, MaIkiel (1995) and Ackerman, McNally, and Ravenscraft (AMR) (1999), are problematic due to poor estimation of the magnitude of survivorship bias, and/or, inadequate value for adjusting actual portfolio returns. However, by employing a hybrid method of estimation, combining the AMR method and a fund-of-funds approach suggested by Fung & Hsieh (2000), survivorship bias estimates of approximately 2% annually are obtained. The second part utilizes multi-factor modeling with common macroeconomic and market risk factors. Hedge fund results are consistent with major strategy style; total risk exposure is greatest for Opportunistic strategies, followed by Event Driven, Relative Value, and Fund of Funds strategies. The source of return for Managed Futures funds derive from completely different patterns (and perhaps sources) of risk (consistent with Schneeweis & Spurgin (1998, 1999), and others). In addition, the risk of some strategies are clearly not well represented by the risk factor structure used for this research; the model is underspecified. The third part uses the technique of the second part on a sample of surviving and non-surviving funds during a time period of market distress. Results indicate that non-surviving funds have risk factor exposures that are about twice the magnitude of their surviving counterparts; funds with the greatest risk exposures are the funds most likely to dissolve.

This dissertation seeks to make a contribution to the recent discussions on the role of stock market development in economic growth. It conducts an empirical study of the economic impact of stock market expansion in China with a focus on the links between the stock market and firm investment. It is based on a panel data set constructed by the author of all Chinese listed firms for the time period from 1992 to 1999. This data set is modeled on and comparable to the COMPUSTAT database in the US. Utilizing this data set, I investigate the effects of the stock market on firm-level investment. The main findings are twofold. On the one hand, by applying both fixed effects and generalized method of moments (GMM) techniques to a modified Trobin's q theory of investment, I find that in China, stock market valuations have a highly significant and positive influence on listed firms' investment decisions, particularly during the stock market boom from 1996 to 1999. On the other hand, the results based on the present-value model show that valuations on the Chinese stock market do not correspond to underlying fundamentals. Stock valuations are especially out of line with firm fundamentals during the market expansion years from 1996 to 1999. These findings suggest that while the stock market in China has played an increasingly important role in guiding investment activities, the recent stock market expansion in China is likely to produce inefficient allocation of investable resources and cause detrimental effects on the real economy. As a result, this study supports the view that great caution must be exerted in relying on the stock market as an effective mechanism for directing investment funds in China. It proposes that a carefully thought out regulatory framework should be implemented in China in order to make best use of the stock market.

Firms that emerged from Chapter 11 as public companies have tons of characteristics. The first essay analyzes their post bankruptcy performance, duration effect, and the quality of their projection information. While the sample’s post bankruptcy performance does show improvement, their projections tend to be optimistic. Firms with shorter durations in Chapter 11generally achieve better performance than those with longer durations, in terms of Z-scores, but not in excess returns. Compared to firms that did not provide (complete) projection information, the sample firms generally exhibit better improvement, as measured by Z-scores and short term excess returns. The second essay tracks the holding period return in investing in bankrupt stocks using a buy-and-hold strategy. Holding period return using stock price alone cannot show the entire story, as when considering final distributions plus the stock price, we see a much severe loss. In the regression analysis, the results reveal that liquidity is always a key factor in explaining the returns. Profitability and information uncertainty plays a significant role in explaining the positive returns, while liquidity and (un)profitability are the two key issues in negative returns. In addition, the involvement of hedge funds does not show signs of better stock performance. The third essay explores the role hedge funds play as investors in bankrupt firms. The results show that their major contributions are to provide liquidity for and help the troubled firms improve their profitability. Compared the performances in post bankruptcy to pre-bankruptcy level, bankrupt firms with hedge funds involvement tend to be in better shape compared to the ones without any vulture investments, however, firms with hedge fund show comparable results with the ones with other vulture investors, such as private equities. In addition, the above improvements only appear in the short run, and the involvement of hedge funds does not guarantee a better stock performance. Therefore, hedge funds are more of financial players, rather than strategic players, as hedge funds do not help the troubled firms go through a systematic restructuring to achieve sustainable improvements.

This dissertation attempts to explore three new ways to understand credit spreads in credit default swaps. The first chapter investigates a hypothesis that the VIX in its role as a fear index impacts intermediary and arbitrageur capital, resulting in decreased market integration across credit and equity markets. Hedging of credit default swaps in the equity markets is found to be surprisingly ineffective. On average, hedging reduces the RMSE by 10% and the VaR by 12%. However, a passive hedge kept in place over a period as long as a month is (multifold) more effective than dynamic daily hedging. The VIX and market returns are demonstrated to predict the improvement that occurs over time. It suggests that frictionless structural models are of limited use in explaining changes in credit spreads. The second chapter attempts to quantify the extent of informed trading versus hedging trading in the CDS market. By examining the relationship between autocorrelation of CDS return and trading activities, I find that an intense trading shock predicts a large momentum in CDS spread for most of firms in my sample, which indicates that informed trading dominates hedging trading in the CDS market according to the theoretical model. The finding is not asymmetric for adverse and favorable news under my measure. However, I do find that the dominance increases ahead of the deteriorating credit conditions and with number of relationship banks in a manner consistent with the theoretical prediction. Weak condition of investment banks may limit their ability and willingness to take risky intermediary activities when firms try to access to public capital market. Credit spreads of firms may go up due to interruption of supply of credit when the market is not as perfect as indicated in theories. Empirical results in the third chapter show that the financial health condition of book runners in debt offerings has significant adverse impact on credit spreads of 133 non-financial client firms during the period from January 2002 to June 2008. Moreover, this effect is more significant for firms that have greater credit constraints and that have stronger relationship with underwriters.

This dissertation investigates the empirical behavior of the exchange rates, especially since the advent of the floating era in 1974. In a sequence of three essays, it seeks to verify whether exchange rate movements conform to the well-known international parity conditions such as Purchasing Power Parity and Covered and Uncovered Interest Rate Parity. It also tests the hypothesis that international capital flows and exchange rate volatility help to explain the forward forecast error (also called the risk premium in international finance literature). In all the three essays, data on exchange rates, interest rates and Consumer Price Indices beginning from 1957 (or from 1981, where appropriate) till the end of 1991 (and up to the end of 1992 in some cases) are used. Data are obtained from the International Monetary Fund monthly statistics and the Financial Times of London. This facilitates a more thorough and reliable empirical analysis. The first essay employs the variance ratio test for random-walk in real exchange rates. A version of the Variance ratio technique developed by Lo and Mackinlay (1988) is used to test for random walk in real exchange rates between the G-7 countries. Results remain unchanged when the samples are pooled. Maximum Likelihood procedures are employed to choose an appropriate random walk process for the real exchange rates. Results indicate that real exchange rates are better approximated by a jump-diffusion random walk process than a simple diffusion process. The second essay investigates the Covered and Uncovered Interest Rate Parity theorems involving the U.S. dollar, the Deutsche mark, the Swiss franc and the Japanese yen. The results offer strong support for the Covered Interest Parity Condition. The third essay involves modeling the risk-premium in forward exchange rates with the net foreign investment position of the U.S. investor and spot exchange rate volatility. Nominal interest differentials are chosen as the proxies for the net foreign investment position. Since under the maintained hypothesis of Covered Interest Parity, the forward premium (discount) reflects the nominal interest differential, they are used in empirical testing. Daily data on one monthly forward and spot rates are used. Implied volatility estimates from foreign currency options serve as instruments for exchange rate volatility. Generalized Method of Moments (GMM) estimation procedure is employed to handle serial correlation and heteroscedasticity in the error terms. (Abstract shortened by UMI.)

Research in market microstructure attempts to determine how differences among trading systems affect price formation in financial markets. In this dissertation, I investigate how large order sizes, and discrete order time/short selling constraints may affect price formation in an equity market. I model a trading system where there are three types of traders, pure information, pure liquidity, and information-liquidity traders. Information-liquidity traders use both private information regarding future value of securities and their liquidity needs to determine their trades. The information liquidity traders are institutions that frequently trade large blocks, and are not subject to short selling constraints. First, a one-shot game is developed as in Easley and O'Hara (1987) with three types of traders and multiple order sizes. Derived results from the model confirm observed price effect of block trades in general, and particularly of large order size on thinly traded stocks. Second, I derive a mixed model in which the arrival of each type of trader is assumed to be a poisson process with time-invariant parameters as in Easley, Kiefer, O'Hara, and Paperman (1996). I outline an MLE method to empirically estimate the parameters of the theoretical model in the first essay. Further using TORQ data, I provide empirical results confirming the role of institutional trading as a determinant of bid ask spread. Finally, I outline a repeated game with two kinds of shocks—private and public information shocks. Some agents receive both shocks, while others receive only a public information shock. I call them complete and incomplete information traders respectively. If there are conflicting signals from the two information sources a fraction of the complete information traders who do not have short selling constraints short sell; others do not trade at all. In this model several rounds of trades occur following a set of shocks. I follow the general framework of Easley and O'Hara (1992) and Diamond and Verecchhia (1987) to show that order time and short sell prohibitions have different price effects in bull (buy > sell) and bear (sell > buy) markets. In a bull market, adverse private information is readily impounded into prices.

This dissertation examines a number of empirical issues that arise in the trading of equity index futures and in research conducted using high frequency futures market data. Both essays benefit from a data set unique to futures market research. The dissertation consists of two essays. The Bid and Ask spread of the FTSE-100 futures contract, presents evidence that bid-ask spreads of the FTSE-100 index futures market are wider than microstructure theory would predict because full point price quotes are systematically preferred over half point price quotes by market makers. The findings are even more pronounced for electronic trading in the automated pit trading (APT) session than during open outcry pit trading. Intraday patterns of spreads are presented, as well as, a spread estimation model that decomposes spreads into its components. A simple model allows the calculation of the total spread related costs of operating the FTSE-100 futures trading pit. Market Structure and the performance of Spread Estimation Techniques in Futures Markets uses transaction data from the same data set to estimate bid-ask spreads with several different estimation techniques. The estimation techniques covered are covariance estimators, high frequency spread estimators, and structural models. The different spread estimates are compared to the actually observed spreads, and reasons for performance differences of the individual estimators are explored. If applied on high frequency data, spread estimation methods are found to be sensitive to the data collection procedure as well as to differences in market structure.