The Market Value of the Corporate Risk Management Function

Abstract

The Market Value of the Corporate Risk Management Function Introduction While the use of risk management strategies have increased dramatically over the last 30 years, measuring the value of risk management activities has been difficult. Much of this can be attributed to the fact that the impact of losses prevented or reduced through risk management practices cannot be easily measured. Schmit and Roth (1990) provide an explanation of the activities performed and tools used by practicing risk managers. The current study examines the issue of whether shareholders value the risk management function. Previous Work Valuation of the firm can be done through several theoretical models such as the Gordon Dividend Growth Model, the Capital Asset Pricing Model (CAPM), and the Arbitrage Pricing Theory model (APT). Cho (1988) explores the relationship between firm value and risk management activities using Gordon's Constant Dividend Growth Model. With the assumption that risk management activities affect the firm's cost of capital, he shows that under certain conditions, risk management activities lower the cost of capital, thus raising the present value of the firm to investors. This suggests that the investors might place some positive value on the risk management process. If the CAPM is valid, rational investors eliminate firm-specific risk through the process of holding a well diversified portfolio of assets, leaving only systematic or market risk. Other authors have examined the issue of whether risk management can be reconciled with the CAPM (see Diallo and Kim, 1989; Cho, 1988; Cross, Davidson, and Thornton, 1986; Cummins, 1976 and 1983; Doherty, 1984 and 1985; Doherty and Tinic, 1981; Hiebert, 1983; Main, 1983a and 1983b; Mayers and Smith, 1982; MacMinn, 1987; and Spreecher and Pertl, 1983), an issue on which the current article provides empirical evidence. Sprecher and Pertl (1983) provide the only other empirical work that is related to this question of the value of risk management activities. Using an event study methodology, they find that large, firm-specific losses create negative abnormal returns. Further, they argue that because the risk management techniques of loss prevention and control can reduce the negative impact of large losses, risk management activities should be positively valued by stockholders. Research Methods An event study methodology similar to that used by Cross, Davidson, and Thornton (1986 and 1989) and Diallo and Kim (1989) is used in this study. The events of interest are defined as published announcements of the formation or expansion of a risk management department in the sample firms. These announcements appeared in various weekly issues of Business Insurance. Investor reaction to these announcements should indicate the value placed on risk management activities by financial market participants. There were 117 announcements of this type identified during the period from 1980 through 1986. To be included in the final sample, the announcements had to refer to firms for which the Center For Research in Security Prices Daily Stock Return contained the necessary stock return information. The final sample consisted of 80 announcements relating to firms whose common stocks are traded on either the New York Stock Exchange or the American Exchange. Unlike other daily event studies in which events are usually identified as having been reported in a daily publication such as the Wall Street Journal, this study uses announcement dates associated with a weekly publication. Furthermore, the information appearing in the published announcements arrives at the publishers in the form of a press release two to five weeks prior to the publication date. Sometime during this pre-event period the original announcement is released and investors are exposed to this new information. This means that any shareholder wealth effect associated with this new information would be expected to occur sometime during this five-week period prior to the identified announcement date rather than on the actual publication date. …