قیمت سهام و اثر ریسک سیستماتیک قطع برنامه های شرکت های بزرگ R & D

We extend the evidence on whether investors impound efficiently into stock prices new disclosures about corporate R&D programs. We find that firms that disclose the discontinuation of some of their R&D programs experience a significant negative announcement-period stock price response which is worse for growth stocks, for small-size firms, and for firms with low operating cash flow. We find no evidence that R&D discontinuing firms experience an event-induced change in their systematic risk. We find evidence of a one-year-long price reversal; however, it is not robust to controlling for possible risk dimensions for firms with R&D capital that the three-factor model does not capture. Evidently, investors' initial response at disclosures of discontinuation of corporate R&D programs is efficient.

In this paper, we extend the empirical evidence on the question of whether investors in the U.S. capital markets value efficiently the R&D capital of firms in the science- and technology-based industries. This question is important because any evidence of under- or over-valuation would mean inefficient allocation of capital into corporate R&D. Yet, the empirical evidence provided by the studies of Chan et al. (2001), Chambers et al. (2002), and Eberhartet al. (2004) does not lead to a clear unanimous conclusion. We show that investors' initial response to announcements of discontinuation of corporate R&D programs is consistent with rationality. It is negative and worse for lower book-to-market ratio firms, whose growth opportunities constitute a more significant part of their market value. It is also more severe for smaller size firms, which due to their size must have fewer R&D programs. Moreover, it is worse for firms which may be compelled to eliminate R&D programs because of cash flow constraints as opposed to efficiency reasons. Furthermore, it is not influenced (or biased) by the condition of the stock market. In terms of long-term effects, we find no evidence of an event-induced change in the sample firms' systematic risk. Additionally, we show that any evidence of a post-event long-term abnormal return is fragile, in that it disappears when we apply reasonable changes to the method of estimating long-term abnormal returns, and therefore, not robust enough to suggest a pricing anomaly or reject the ‘rational expectations' hypothesis in our sample. These “reasonable” changes to the method of estimating long-term abnormal returns to firms with R&D capital are based on robust evidence provided by Lev and Sougiannis (1999) that indicates some deficiency in the Fama–French three-factor model in pricing firms with R&D capital. In fact, Fama and French (1993) and Mitchell and Stafford (2000) show that the three-factor model does not explain well the cross-sectional variation of stock returns for low book-to-market and small-size firms. This conclusion is consistent with the findings of Chambers et al. (2002) who report that a positive association between R&D investment levels and excess returns is more likely the result of failure to control adequately for risk than of mispricing. An important empirical implication of our study is that there is clear need to identify and validate one or more additional risk factor(s) that can complement the Fama–French risk factors in explaining the returns to firms in the science- and technology-based industries (i.e. firms with R&D capital and other intangible assets). We find that even the bootstrapping procedure prescribed by Mitchell and Stafford (2000) to adjust the abnormal returns or intercepts obtained from the Fama–French three-factor model, and which is expected to correct for any tilting in the sample toward characteristics that this model cannot price in the first place, does not adequately compensate for the inability of the size and book-to-market factors to fully capture the risks and growth opportunities of R&D-intensive firms. Based on our results and the evidence provided by Lev and Sougiannis (1999), the R&D to market value ratio may be a good starting point. We hope future research will address this important issue.