An experimental study of analysts' reactions to corporate R&D expenditure

Citation

Goodacre A & McGrath J (1997) An experimental study of analysts' reactions to corporate R&D expenditure. British Accounting Review, 29 (2), pp. 155-179. https://doi.org/10.1006/bare.1996.0041

Abstract
Company management have incentives to consider the impact of their decisions on short-term reported earnings if either investors adopt a mechanistic approach to earnings numbers or behave myopically. Previous research, mainly in the US, suggests that, while individuals may respond mechanistically to earnings numbers, investors in aggregate (i.e. the market) do not. Similarly, evidence suggests that company management may believe the market to be short-termist, whereas research has indicated that, at least in relation to R&D expenditure announcements, it is not. The present study investigates whether UK investment analysts exhibit mechanistic or myopic tendencies with respect to research and development expenditure. Two experimental studies using a postal questionnaire were undertaken in which investment analysts were asked to forecast earnings and market values of electronics companies based on simulated financial statements. The first experiment involved two companies which were identical in all respects except for the method used to account for R&D expenditure. The mean market value estimates for the companies were almost identical, a result which is inconsistent with a mechanistic approach. Analysts did not seem to be misled by the higher earnings reported for the company capitalizing R&D expenditure. In the second experiment, the only difference between the two companies was that one invested in tangible fixed assets rather than in R&D. In this case, the mean market value estimate for the R&D spender was significantly higher. This suggests that analysts recognize the importance of long-term R&D investment and are not myopic in this respect. Overall, the results imply that company management's pre-occupation with short-run earnings may be unnecessary.

Journal
British Accounting Review: Volume 29, Issue 2