On the value-relevance of financial and non-financial information: the case of internally generated intangibles

Vidolovska, Ana (2006) On the value-relevance of financial and non-financial information: the case of internally generated intangibles. Doctoral thesis, University of London: London Business School.

Abstract

The debate about the conservative treatment of intemally developed intangibles has focused on the issue of their capitalization versus expensing. Given that a significant portion of intangible investments is not identified in the financial statements, a more basic question to explore is whether they should be disclosed separately. This dissertation examines the value-relevance of both financial and non-financial measures of intangible investments. It provides evidence that may be helpful to standard setters in assessing the need for regulation of the disclosure of intangibles captured by financial information, and for standardisation of the measurement and enhancement of the disclosure of non-financial indicators of intangible assets. The structure of the dissertation is as follows: Chapter 2 reviews prior literature and outlines the contribution. Part I (Chapter 3) focuses on financial measures and Part 11 (Chapters 4 and 5) explores non-financial indicators of intemally generated intangibles. The first study (Chapter 3) examines US firins' discretionary disclosures of advertising costs after 1994 as a function of its valuation benefits and proprietary costs experienced in the pre-1994 period when disclosure was mandatory. The association of advertising investments with stock prices and with future operating profit captures valuation benefits of advertising. The impact of firms' advertising on their competitors' perfon-nance is defined as the proprietary cost. The empirical findings show that firms whose advertising investments create firm value (benefit rivals) before 1994 tend to disclose advertising outlays more (less). The results also indicate that investors assign a higher value to voluntarily disclosed advertising expenses. The second study (Chapter 4) evaluates the value-relevance and predictive ability of customer relationship infon-nation for US-listed wireless firms over the period 1997- 2004. This part of my dissertation is based on a paper, co-authored with Prof. Eli Talmor and Dr Gilad Livne. Customer relationship is measured by subscriber acquisition cost, contractual arrangements, market share, and chum rate. The results reveal that most performance indicators of customer relationship exhibit predictive power for future revenues and operating profits. Collectively, the set of non-financial measures displays a significant association with stock prices and returns; this association is driven primarily by customer acquisition cost. The study also investigates the recoverability rate of investments in customer acquisition. Their return is found to exceed the cost, although it accrues over a relatively short period. The third dissertation study (Chapter 5) examines analysts' efficiency in utilising non-financial measures of intangibles. I compare the extent to which analysts rely on fundamentals constructed from audited items in the financial statements versus non- GAAP performance indicators. In addition, I examine whether investors follow analysts' interpretation of these items and whether they assess their implications correctly. The findings provide evidence that analysts tend to under-react to most non-GAAP measures, which affects the accuracy of their forecasts. There is some evidence that analysts overreact to certain GAAP proxies, consistent with analysts' differential treatment of audited versus non-standardised information. Investors rely on analysts' interpretations as indicated by zero abnormal returns to positions taken on analysts' forecasts. Investors fall to correctly assess the contribution of the same non-GAAP measures that analysts underutilise. This is demonstrated by non-zero abnormal returns to hedge strategies based on these indicators.