DEPRECIATION: ACCOUNTING, TAXES, AND CAPITAL MARKET EQUILIBRIUM

This thesis examines the effect on capital market equilibrium of firms\u27 adoptions of accelerated methods of depreciation, including the double-declining-balance and sum-of-the-years digits methods, as permitted by the 1954 Tax Code. Stock returns are assessed at the time of firm adoption and a theoretical model of windfall gains is developed to stratify firms by relative impacts. The effect on reported earnings of the method selected for financial reporting purposes is also examined. This thesis also attempts to control for other intervening events at the time of adoption, specifically unexpected earnings changes. The empirical evidence supports the hypothesis that the windfall gains accruing to the firm translate into higher than expected returns on the firm\u27s stock. The evidence is generally consistent with market efficiency, since abnormal returns are positive and significant only prior to the adoption date. The strongest results are obtained for a sample of firms adopting accelerated depreciation not just prospectively but also retroactively for assets already in place. The driving force behind the theoretical windfall gains model is an expected capital expenditure growth rate variable. Firms\u27 abnormal returns are successfully stratified by one model of expected impact. The examination of samples by the choice of reporting method suggests that the market does not respond naively to the impact on reported earnings of the financial accounting choice. While the reported earnings impact for firms choosing accelerated is negative and for firms choosing either accelerated or straight-line deferral is non-positive, these firms experience positive abnormal returns. While the reported earnings impact for firms choosing straight-line flow-through is positive, these firms experience negative abnormal returns. There is some evidence that firms choosing straight-line flow-through may wish to disguise otherwise poor performance. While the earnings surprises in the adoption year were found to be substantially positive, these adopting firms outperformed control firms with similar earnings surprises. To the extent that unanticipated news is captured by unexpected changes in earnings per share, it appears that the observed windfall gains are due to the adoption decision itself and not other events correlated with the adoption decision