The role of the most recent prior period's price in value relevance studies : a thesis presented in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Finance at Massey University, Palmerston North, New Zealand

Abstract

Numerous value relevance investigations use the Ohlson (1995) model to empirically explore the value relevance of accounting variables such as earnings and goodwill amortisation by employing equity price as the dependent variable, but do not incorporate the most recent prior period’s equity price as an additional explanatory variable. The Ohlson (1995) model and the efficient market literature indicate that, since share prices represent the present value of future permanent earnings in an efficient market, the most recent prior period’s equity price should be a crucial variable for explaining the current price in value relevance models. This thesis therefore outlines how the Ohlson (1995) model incorporates the most recent prior period’s price as a potentially important value relevant explanatory variable, and reformulates the Ohlson (1995) model to demonstrate how the empirical specification of value relevance regression models can be greatly improved by including the most recent prior period’s price as an additional explanatory variable. We revisit the Jennings, LeClere, and Thompson (2001) empirical specification used to study whether goodwill amortisation is value relevant and potentially informative with respect to future earnings to illustrate the improvement to the Ohlson (1995) value relevance model empirical specification. When the model specification is improved by including the most recent prior period’s price as an additional explanatory variable, trailing earnings are shown, using time series, cross-sectional, and returns-based analysis, to be at best marginally value relevant when empirically explaining share prices in value relevance regression models. The thesis also indicates that goodwill amortisation should not be deducted from earnings in accounting statements because the presence of goodwill amortisation is significantly positively (not negatively) related to equity prices. This effect is eliminated when the most recent prior period’s price is included as an additional explanatory variable in the regression analysis, thus indicating that goodwill amortisation information as well as trailing earnings information have already been incorporated into the most recent prior period’s price. The thesis further indicates that value relevance studies that use the Ohlson (1995) model should use, for econometric reasons, change in price or else returns, not the price level, as the dependent variable. When returns are used to test the value relevance of goodwill amortisation, firms that report positive goodwill amortization actually have higher subsequent returns, a result that could possibly be due to the fact that growing firms tend to possess goodwill when they use acquisitions to expand. Results obtained when using returns to test whether goodwill amortisation is value relevant therefore extend the existing literature, since the prevailing expectation in the accounting literature is that goodwill amortization either represents a reduction in the value of goodwill over time or is not value relevant.