جبران خسارت صاحبان سهام و حساسیت تحقیق و توسعه به اصطکاک بازار مالی

When financial market frictions exist, executives may have to decide which investment activities to reduce when internal funds decrease. Expenditures on research and development (R&D) may be particularly vulnerable because of the long-term nature of innovative activity. We find that equity compensation is associated with lower levels of firm R&D expenditures. Rewarding executives to incur more risk has little effect on R&D expenditures, but rewarding executives for higher returns reduces R&D expenditures and makes R&D expenditures more sensitive to financial market frictions. In contrast, cash compensation reduces the sensitivity of R&D expenditures to financial market frictions.

This paper examines the incentive effects of equity compensation on firm-financed research and development (R&D) expenditures. In a world without any frictions, firms will pursue all investment projects that have a positive net present value so executive compensation should have no effect on the level of R&D activity. However, if asymmetric information, the financial friction we examine in this paper, exists in markets then firms may be unable or unwilling to pursue all positive net present value investment projects. We focus on agency problems caused by asymmetric information between shareholders and executives. However, we also discuss financial market frictions caused by asymmetric information between firms and creditors. In particular, we examine how the incentive effects imbedded in executive compensation packages influence financial market frictions. Shareholders are often risk neutral because they hold a large portfolio of assets which allow them to diversify away firm specific risks. However, much of the executive’s reputational and human capital is tied to the performance of the firm so executives are often risk averse. If it is difficult or costly to monitor executives then agency problems, exist which allow executives to pursue their own interests rather than the interests of the shareholders. One way executives may pursue their own interests at the expense of the shareholders is to reduce risky long-term investments such as R&D activity. Therefore, firms with agency problems may perform less R&D. The agency problem may also affect R&D expenditures by magnifying the influence of financial market frictions. When a firm faces binding financial market constraints, the firm cannot pursue all profitable investment opportunities so executives at the firm must decide which investment activities to reduce. If agency problems exist then executives may be more likely to reduce R&D expenditures than other types of investment projects. Therefore, agency problems may influence the sensitivity of R&D expenditures to financial market frictions. Equity compensation (both in the form of granted shares and stock options) may provide a solution to both financial market frictions and the agency problem. By making the executives part owner of the firm, equity compensation may align the interests of the executives with those of the owners. Therefore, equity compensation may make executives more willing to make R&D expenditures and, in addition, make executives less willing to reduce R&D expenditures when financial constraints bind. If equity compensation works as intended then an increased reliance on equity compensation should increase the average amount of R&D expenditures and reduce the sensitivity of R&D expenditures to financial market frictions. Alternatively, equity compensation may have perverse effects. Executives may seek to maximize the short-term value of their shares and options. If markets have a short-term bias then this could lead executives to reduce the average amount of R&D expenditures and may increase the sensitivity of R&D expenditures to financial frictions. For example, with options vesting or expiring in the near future a management team might cut R&D (to manage earnings), initiate a stock buyback plan, or announce an increase in dividends. Rather than solving the agency problem by aligning incentives of the executives and the owners of the firm, stock options and other forms of equity compensation might create an even larger gap between the incentives of executives and the owners of the firm. In this paper, we use delta (the sensitivity of executive compensation to stock price returns) and vega (the sensitivity of executive compensation to stock price volatility) to measure the incentive effects of executive compensation. We find essentially no relationship between vega and the level of R&D activity. However, we find that the greater the delta of the executive’s compensation package, the lower the level of R&D expenditures and the more sensitive R&D expenditures become to financial market frictions. Finally, we find that cash compensation reduces the sensitivity of R&D expenditures to financial market frictions. We believe that our paper is the first to show that executive compensation packages influence financial market frictions. We also use an econometric technique that overcomes several sources of bias that influence the results in the current literature. The paper proceeds as follows. In Section 2 we review the literature on R&D activity and the incentive effects of executive compensation. In Section 3 we modify a Tobin’s q model to allow for the possibility that executive compensation influences both the average level of R&D expenditures and the sensitivity of R&D to financial frictions. We also explain our empirical approach in this section. Section 4 describes the data and presents the major findings of the paper. We present results from a variety of estimation techniques in this section from ordinary least squares to generalized method of moments. We conclude in Section 5 and offer some policy recommendations to boards of directors.

To the best of our knowledge, this paper is the first to show that equity compensation can magnify the effect of financial market frictions. The delta of the compensation package is almost entirely responsible for this which makes sense if executives are driven by a desire to manage earnings to maximize the value of their in-the money options. We have also shown that the standard empirical models likely suffer from endogeneity bias even when executive compensation is a predetermined variable. Our results control for multiple sources of endogeneity bias. Moreover, the use of dynamic instruments provides a large number of over-identifying restrictions so, unlike Coles et al. (2006), we are able to test the validity of instrument set. In addition, the dynamic instruments allow us to avoid using firm risk and capital expenditures as instrumental variables for R&D. It is highly unlikely that these variables are actually exogenous so Coles et al. (2006) may be using instruments that are not valid. Comparison of the OLS and difference-GMM results in Table 2 suggest that exogenous increases in R&D expenditures lead firms to decrease the vega and increase the delta of executive compensation packages. This is inconsistent with the empirical findings of Coles et al. (2006), but consistent with the findings of Cheng (2004). After controlling for endogeneity bias, we find that the higher the delta of the compensation package the lower the level of R&D expenditures and the more sensitive R&D expenditures are to financial market frictions. This indicates that either equity compensation has a perverse effect on R&D expenditures or that executives tend to perform too much R&D expenditures. Since we also find that increasing delta of compensation packages increase the sensitivity of R&D expenditures to financial market frictions we believe that our results are most consistent with the view that executive compensation can have perverse effects on R&D expenditures. Our results on the moneyness of options show that the moneyness of options do matter and that the negative relationship between delta and R&D is driven entirely by in-the money options. The negative relationship between delta and R&D expenditures is robust to a variety of econometric techniques and control variables. In contrast, we find little to no evidence of a relationship between vega and R&D in most of our models indicating that rewarding executives for incurring risk has little effect on R&D expenditures. However, our results on the moneyness of options and vega are mixed. On the one hand, at-the money vega is associated with lower levels of R&D, but out-of-the money vega is associated with reduced sensitivity of R&D expenditures to internal funds. On balance, the evidence in favor of the view that increased equity compensation reduces the level of R&D expenditures is much stronger than the evidence that equity compensation increases R&D expenditures. Much like our results for delta, our results for cash compensation are robust across alternative specifications of the model. We consistently find that cash compensation either has no relationship with R&D expenditures or is associated with higher R&D expenditures. We also consistently find that increased cash compensation reduces the sensitivity of R&D expenditures to financial market frictions. This would occur if executives use cash compensation to diversify away firm specific risk so that they prefer a more risk-neutral corporate strategy. We believe that our results represent the exogenous effect of changes in executive compensation on R&D expenditures. If this is the case then boards that seek to encourage R&D expenditures should reduce the level of equity compensation. There appears to be no benefit to rewarding executives for risk while rewarding them for higher returns seems to encourage reductions in R&D expenditures. In contrast, boards should consider increasing the level of cash compensation which will allow the CEO to diversify away firm specific risk and make the CEO more risk neutral with respect to corporate strategy.