An examination of three sources and impact of information asymmetry in financial markets

This dissertation presents three sets of empirical tests focused on sources of information asymmetry in financial markets. Specifically it determines the impact of CEO narcissism on earnings management, the factors which influence the release of privileged information by market intermediaries and asymmetrical information access related to technological enhancements. The results in this dissertation address a number of gaps in existing literature, which are of relevance to academics, and provide valuable insights for regulators and market participants. The first set of empirical tests examines the impact of firm CEO’s narcissism on earnings management. There is growing evidence to suggest that narcissistic leaders over-identify themselves with the organizations that they lead and expend considerable resources to achieve their goals, including engagements in unethical behaviour. Previous theoretical studies have conjectured the relation between leadership narcissism and accounting manipulation but lacking in empirical support. This dissertation examines this theoretical hypothesis adopting an unobtrusive yet psychologically supported proxy for narcissism based on the Narcissistic Personality Inventory (NPI) test. Results reported in this dissertation provides evidence that firms with more narcissistic CEOs are more likely to engage in such behaviour as evidence by the manipulation of accounts to present better earnings. The results highlight that information asymmetry caused by earnings management can be identified at an early stage by virtue of a CEO’s personality. The implications of these results should be of particular interest to regulators and shareholders as management personality is potentially a prescription to corporate fraud. The second set of empirical tests examines the extent of information leakage around analyst recommendations (‘tipping’), the propensity to act on this privileged information received under different circumstances and its resulting profitability. While the earliest evidence of tipping was only documented for upgrade recommendations, there has been increasing evidence of this phenomenon for downgrade recommendations. Using a unique dataset from the ASX, which contains broker identifiers, this study extends the empirical literature on tipping by examining factors where this phenomenon is most prevalent, and the corresponding profit associated with the phenomenon. Results reported in this dissertation demonstrate that irrespective of market conditions, recipients react predominantly to tips on downgrade recommendations. Further analysis by cross-section of firms indicates that leaks on smaller and mid-capitalization stocks exhibit higher abnormal trading volume and affords greater profitability. When negative returns precede the public release of downgrade reports, results demonstrate that recipients are less likely to act. On investigation of short-selling around analyst recommendations, results document no signs of institutional exodus around analyst releases, suggesting that institutional investors do not react to short-term price fluctuations. The implications of this chapter enable a better understanding of supplementary factors that contribute to the phenomenon of tipping. The final set of empirical tests investigates the impact of heightened levels of computerised trading on institutional execution costs. The advent of this new entrant into financial markets has attracted attention from regulators, exchanges, market participants and academics spurring empirical studies on the associated impacts on market quality. However, evidence on the impact to institutional investors has produced mixed results. This dissertation employs an exogenous event of co-location by 12 major exchanges worldwide to document the first empirical evidence on adverse effects to institutional trading costs as a result of intensified activity by computerised trading participants. Results of this dissertation provide that the reduction in exchange latency as a result of co-location attracts more activity by computerised traders and correspondingly, an increase in institutional execution costs.