Longitudinal Earnings Value Relevance and Intangible Assets: Evidence from Australian Firms, 1975-99

Abstract

This paper investigates the longitudinal earnings value relevance of Australian firms for the period 1975 through 1999. There is evidence of a decline in earnings value relevance over this 25-year period, consistent with findings of US studies. However, only losses explain much of the decline, which is inconsistent with findings of US studies. In an attempt to reconcile this difference, intangible asset recognition is singled out as a likely discriminator since it is conjectured by some that immediate expensing of certain intangible assets significantly contributes to declining earnings value relevance. Results suggest earnings value relevance has declined for firms that don't recognise intangible assets (non-capitalizers) and there is weak evidence of decline for firms that recognise intangible assets (capitalizers). Additionally, after controlling for losses capitalizers have no change in (increasing) earnings value relevance and non-capitalizers have decreasing (no change in) earnings value relevance using R2 (earnings parameters) as a value relevance measure. Results are robust to various alternative explanations and different model specifications. For capitalizers in the 1990s, those that amortize generally have stronger value relevance than non-amortizers. Amortization and its change are most significantly associated with return for intangibles other than R&D and deferred costs. In some years, the level of amortization is significantly positive suggesting the market does not treat amortization as a period expense. The change is negative and significant in most years, suggesting amortization revisions reflect valuable information about future earnings.

Keywords: longitudinal earnings value relevance, intangible assets

JEL Classification: G12, M41