AbstractThis study examines whether Hong Kong managers choose “benchmark” or “alternative” valuation method for investment securities, after the Hong Kong SSAP 24 became effective starting with fiscal‐year ending December 31, 1999. Tests are conducted on a sample of 292 firms, out of which 155 Hong Kong firms reported unrealized gains and losses and 128 firms that did not report holding gains/losses, but reported investment securities.The findings indicate that firms with strong relative performance, i.e. current year's EPS higher than that of the last year, chose the alternative valuation method when the investment securities had holding gains and recognized the unrealized holding gains in the equity section of the balance sheet. This finding is consistent with the Cookie Jar hypothesis because these holding gains would be used in the income statement in future periods, when needed. With regard to firms with strong relative performance and holding losses, the findings indicate that the benchmark valuation was used. The losses were reported in the income statement to the extent that they did not reduce the EPS below that of the last year. This finding is consistent with the Income Smoothing Hypothesis, because the use of benchmark valuation reduced EPS of the current year to bring it in line with that of the last year. Evidence on firms with weak economic performance and holding gains or losses provided weak support to the Income Smoothing Hypothesis and Big Bath Hypothesis. Additionally, the results indicate that the firms with high debt‐equity ratio preferred the benchmark method and recorded securities at cost. This treatment provided managers with an opportunity to liquidate or reclassify the securities in future periods and use the accrued gains, when needed. The findings are inconclusive with regard to the impact of bonus plan on the choice of valuation method.