This thesis examines the relationship between tax-loss selling (TLS), where investors with taxable gains sell stocks that have declined in value just before the fiscal year-end to generate offsetting tax losses, and managers?? incentives to influence stock prices, either through increased disclosure or by engaging in upwards earnings management. Firms whose stock prices represent greater potential tax losses in investors?? portfolios at year-end are predicted to increase their disclosure level in June to prevent further share price falls due to TLS, and have higher levels of accruals. Using the number of discretionary, market-sensitive news releases in the Signal G announcement database to measure disclosure frequency, this thesis finds that, for a sample of 14,713 firm-year observations drawn from all ASX firms for the years 1994 to 2007, stocks with larger negative returns have higher disclosure in June, after controlling for size, performance, risk and external financing dependence. This is particularly true of small mining and exploration companies that are more reliant on voluntary disclosure as a vehicle for lowering information asymmetry. This increased disclosure does not appear to contribute to the higher July returns earned by stocks that experienced significant TLS in June. Disclosure frequency is negatively associated with the magnitude of operating and total accruals, suggesting that earnings management is less likely for firms with higher disclosure. There is also evidence that smaller firms with poor stock price performance have higher levels of operating accruals and thus may be more likely to engage in earnings management.

This thesis examines the relationship between tax-loss selling (TLS), where investors with taxable gains sell stocks that have declined in value just before the fiscal year-end to generate offsetting tax losses, and managers?? incentives to influence stock prices, either through increased disclosure or by engaging in upwards earnings management. Firms whose stock prices represent greater potential tax losses in investors?? portfolios at year-end are predicted to increase their disclosure level in June to prevent further share price falls due to TLS, and have higher levels of accruals. Using the number of discretionary, market-sensitive news releases in the Signal G announcement database to measure disclosure frequency, this thesis finds that, for a sample of 14,713 firm-year observations drawn from all ASX firms for the years 1994 to 2007, stocks with larger negative returns have higher disclosure in June, after controlling for size, performance, risk and external financing dependence. This is particularly true of small mining and exploration companies that are more reliant on voluntary disclosure as a vehicle for lowering information asymmetry. This increased disclosure does not appear to contribute to the higher July returns earned by stocks that experienced significant TLS in June. Disclosure frequency is negatively associated with the magnitude of operating and total accruals, suggesting that earnings management is less likely for firms with higher disclosure. There is also evidence that smaller firms with poor stock price performance have higher levels of operating accruals and thus may be more likely to engage in earnings management.

It is well understood through previous literature that strategies, which buy past winning stocks and sell past losing stocks, can generate significant positive returns. This phenomenon is known as the momentum effect in the stock market. Furthermore, there is a common accounting practice used by portfolio managers called tax loss harvesting.Tax loss harvesting is the practice of selling a security in order to create a benefit for tax purposes. This paper attempts to build upon previous literature by explaining why the momentum effect is different at the beginning of the calendar year than in the middle and assessing whether or not tax loss harvesting may play a role. A trading strategy was created which calculates the returns of winning and losing portfolios intra industry groups, around different months of the year, in attempt to explain fluctuations in the momentum effect. Evidence in support of the hypothesis that tax loss harvesting played a role in impacting momentum strategies did not prove to be statistically significant.