Doctor of Philosophy (PhD) / Allegations of market manipulation abound in the popular press, particularly during the recent financial turmoil. However, many aspects of manipulation are poorly understood. The purpose of this thesis is to enhance our understanding of market manipulation by providing empirical evidence on the prevalence, effects and determinants of closing price manipulation. The first issue examined in this thesis is the prevalence of closing price manipulation. This thesis uses a hand collected sample of prosecuted closing price manipulation cases from US and Canadian stock exchanges, and methods that explicitly model the incomplete and non-random detection of manipulation. The results suggest that approximately 1.1% of closing prices are manipulated. For every prosecuted closing price manipulation there are approximately 300 instances of manipulation that remain undetected or not prosecuted. Closing price manipulation is more prevalent on larger exchanges than smaller ones, but is detected at a higher rate on small exchanges. Second, this thesis examines the effects of closing price manipulation. Using a sample of prosecution cases, this thesis finds that closing price manipulation is associated with large day-end returns, subsequent return reversals, increases in day-end spreads and increases in day-end trading activity. At the broader level of market quality, this thesis provides evidence from a laboratory experiment that closing price manipulation decreases both price accuracy and liquidity. Even the mere possibility of manipulation decreases liquidity and increases trading costs. The third issue analysed in this thesis is the determinants of closing price manipulation and its detection. Estimating an empirical model of manipulation and detection, this thesis finds that the likelihood of closing price manipulation is increased by smaller regulatory budgets, greater information asymmetry, mid to low levels of liquidity, month-end days and lower volatility. Manipulation is more likely to be detected when regulatory budgets are larger and when the manipulation causes abnormal trading characteristics. Further evidence from laboratory experiments suggests that regulation helps restore price accuracy by deterring some manipulation and making remaining manipulation less aggressive. These experiments also show that regulation has an insignificant effect on liquidity because participants in regulated markets still face relatively high uncertainty about the presence of manipulators. This thesis also examines how closing price manipulation is conducted and how other market participants respond. It develops an index of closing price manipulation that can be used to study manipulation in markets or time periods in which prosecution data are not available. It also provides a tool for the detection of manipulation, which can be used by regulators in automated surveillance systems. Finally, this thesis has implications for economic efficiency and policy. Closing price manipulation is significantly more prevalent than the number of prosecution cases suggests. Further, it harms both pricing accuracy and liquidity and therefore undermines economic efficiency. The prevalence of closing price manipulation can be reduced by increasing regulatory budgets, improving the accuracy of market surveillance systems by using the detection tools developed in this thesis, structuring markets such that participants are better able to identify manipulation, and implementing closing mechanisms that are difficult to manipulate. These actions would enhance market integrity and economic efficiency.

Doctor of Philosophy (PhD) / Allegations of market manipulation abound in the popular press, particularly during the recent financial turmoil. However, many aspects of manipulation are poorly understood. The purpose of this thesis is to enhance our understanding of market manipulation by providing empirical evidence on the prevalence, effects and determinants of closing price manipulation. The first issue examined in this thesis is the prevalence of closing price manipulation. This thesis uses a hand collected sample of prosecuted closing price manipulation cases from US and Canadian stock exchanges, and methods that explicitly model the incomplete and non-random detection of manipulation. The results suggest that approximately 1.1% of closing prices are manipulated. For every prosecuted closing price manipulation there are approximately 300 instances of manipulation that remain undetected or not prosecuted. Closing price manipulation is more prevalent on larger exchanges than smaller ones, but is detected at a higher rate on small exchanges. Second, this thesis examines the effects of closing price manipulation. Using a sample of prosecution cases, this thesis finds that closing price manipulation is associated with large day-end returns, subsequent return reversals, increases in day-end spreads and increases in day-end trading activity. At the broader level of market quality, this thesis provides evidence from a laboratory experiment that closing price manipulation decreases both price accuracy and liquidity. Even the mere possibility of manipulation decreases liquidity and increases trading costs. The third issue analysed in this thesis is the determinants of closing price manipulation and its detection. Estimating an empirical model of manipulation and detection, this thesis finds that the likelihood of closing price manipulation is increased by smaller regulatory budgets, greater information asymmetry, mid to low levels of liquidity, month-end days and lower volatility. Manipulation is more likely to be detected when regulatory budgets are larger and when the manipulation causes abnormal trading characteristics. Further evidence from laboratory experiments suggests that regulation helps restore price accuracy by deterring some manipulation and making remaining manipulation less aggressive. These experiments also show that regulation has an insignificant effect on liquidity because participants in regulated markets still face relatively high uncertainty about the presence of manipulators. This thesis also examines how closing price manipulation is conducted and how other market participants respond. It develops an index of closing price manipulation that can be used to study manipulation in markets or time periods in which prosecution data are not available. It also provides a tool for the detection of manipulation, which can be used by regulators in automated surveillance systems. Finally, this thesis has implications for economic efficiency and policy. Closing price manipulation is significantly more prevalent than the number of prosecution cases suggests. Further, it harms both pricing accuracy and liquidity and therefore undermines economic efficiency. The prevalence of closing price manipulation can be reduced by increasing regulatory budgets, improving the accuracy of market surveillance systems by using the detection tools developed in this thesis, structuring markets such that participants are better able to identify manipulation, and implementing closing mechanisms that are difficult to manipulate. These actions would enhance market integrity and economic efficiency.

By using neural network, the relationship between the low price and the closing price in IMKB is developed by a fuzzy membership function.