Description
To become a dictionary definition of managerial effectiveness, you must be able to lead a group of people. Leaders lead in different ways, but getting results usually involves motivational techniques.
Managerial Motives and Corporate Use of Derivatives: Some Evidence
Peter D. Wysocki* University of Michigan Business School 701 Tappan Street, Ann Arbor, MI 48109-1234, USA Telephone: (734) 763-6820, E-mail: [email protected]
July 1998
Abstract
This paper provides empirical evidence on the link between a firm's use of derivatives and its compensation policy, ownership structure and organizational structure. Agency theory indicates that top management may have incentives to use derivatives to reduce personal risk if CEO cash compensation is risky, insiders have high levels of wealth vested in firm equity, the CEO is about to retire, or equity ownership is diffuse. In addition, division managers have incentives to use derivatives to reduce risk in measures of division performance. I investigate these determinants of derivatives use in a 1994 sample of 403 U.S. firms, including 215 derivatives users, and a 1993 sample of 807 U.S. firms, including 234 foreign exchange derivatives users. Consistent with the division manager hypothesis, I find that derivatives use is increasing in the number of lines of business and the number of overseas operations. Derivatives use is found to be decreasing in inside ownership, but is unaffected by the riskiness of CEO compensation, the level of insider wealth vested in equity, or CEO retirement. These results are consistent with the notion that shareholders write optimal contracts with top management to mitigate opportunistic use of derivatives. Similar to previous studies, I find that derivatives use is increasing in firm size and decreasing in regulation.
? I wish to thank Ray Ball, Dan Gode, Christoph Hinkelmann, S.P. Kothari, D.J. Nanda, Jerry Warner, and the participants of the Simon School Doctoral Seminar Series for their helpful comments and suggestions. Please note that this is a draft doucment reporting preliminary results of on-going research.
1. Introduction
There is a growing debate about how and why firms use derivative securities to alter firm risk. This debate is of particular relevance given the explosive growth in the use of derivatives by corporations. The preliminary results of the October 1995 Wharton/CIBC Wood Gundy Survey of Derivatives Usage indicate that derivatives use among respondents increased from 35% to 41% between 1994 and 1995. In addition, the results of the 1995 survey indicate that 91% of derivativesusers cite volatility in cash flows or earnings as the most important reason for using derivatives.
Much of the recent debate on derivatives use has focused on whether firms use these instruments for hedging purposes to maximize shareholder wealth or for opportunistic speculation. Finance theory indicates that hedging with derivatives can increase firm value by reducing expected taxes, expected costs of financial distress, underinvestment costs associated with investment opportunities in the presence of financial constraints, and agency costs. Several recent studies including Nance, Smith and Smithson (1993) have found mixed empirical results on these hedging-related determinants of derivatives use1. Hentschel and Kothari (1995) find that derivatives users and non-users exhibit few measurable differences in risk and conclude that this is consistent with firms using derivatives to hedge rather than speculate. However, these investigations fail to address a key factor that may influence the use of derivatives. In particular, past studies do not examine the risk characteristics of managers and their incentives to use derivatives. Therefore, while firms use derivatives for hedging purposes, this activity may benefit risk-averse managers at the expense of diversified shareholders.
It may be argued that managerial risk, compensation policies, ownership and organizational structure are not important ex post determinants of derivatives use if, ex ante, shareholders choose optimal management contracts that remove incentives for suboptimal reduction in firm risk. However, May (1995) finds empirical evidence consistent with the hypothesis that managerial motives influence a firm's risk reduction strategies. In particular, he finds that CEOs with more personal wealth vested in
1
Recent cross-sectional studies on derivatives use among non-financial firms include Nance, Smith and Smithson (1993), Francis and Stephan (1993), Mian (1994), and Geczy, Minton, and Schrand (1995).
equity tend to diversify through acquisitions, and that there is a negative relation between CEO human capital invested in the firm and the firm's equity variance and debt to equity ratio. These results appear to indicate that shareholder contracts with top management may not remove all personal incentives for managers to reduce firm risk.
Jensen and Meckling (1976) argue that agency conflicts may arise when managers hold an undiversified portfolio of equity and human capital invested in a single firm. Therefore, risk-averse managers may be motivated to reduce firm risk for personal gain at the expense of diversified shareholders. There are several cases where top managers would have the highest incentives to use derivatives to reduce risk. First, if CEO cash compensation contains a large risk component that is outside his control, then he may wish to use derivatives to hedge this risk. In particular, a compensation plan that includes a large bonus component based on accounting measures of performance may induce a manager to hedge accounting numbers. Second, insiders and top managers with high levels of personal wealth vested in firm equity may use derivatives to hedge firm value to reduce the risk of their undiversified investment portfolio. Third, if a CEO receives a post-retirement compensation package based on the firm's accounting performance prior to the CEO's retirement, then he may use derivatives to hedge accounting numbers to reduce the risk of this compensation. In theory, the preceding incentives can be eliminated if shareholders can write optimal compensation contracts. However, such contracts may be costly to negotiate and implement. Moreover, if equity ownership is diffuse, then there may be reduced incentives for individual shareholders to write and enforce optimal management contracts.
Jensen and Meckling also provide a framework for analyzing the organizational structure of a firm. In a decentralized firm, division managers may have incentives to reduce risk in measures of division performance if their compensation is based on division operating performance. It should be noted that division managers are further removed from shareholders and it may be difficult to write and implement contracts which limit opportunistic use of derivatives by division managers. This would imply that greater the number of divisions, the greater the likelihood that a firm will use derivatives.
The goal of this paper is to provide new empirical evidence on the relation between derivatives use and a firm's compensation policy, ownership structure and organizational structure. My empirical analysis focuses on two samples of firms that include derivatives users and non-users. The first sample is based on fiscal year 1994 financial statement information collected for 403 non-financial firms that trade on the New York (NYSE) and American (AMEX) stock exchanges. The analysis uses data collected from financial footnotes in annual reports to determine use of any type of derivative security at the end of fiscal 1994. Over 46% of the firms in this sample report using derivatives. CEO salary and bonus compensation, insider holdings of stock, line of business classifications, and other financial information are collected from electronic filings of proxy statements with the SEC and from the Standard and Poor's (S&P) Disclosure and Compustat databases.
The second sample is based on fiscal year 1993 financial statement information collected for 807 non-financial firms that trade on the NASDAQ, NYSE and AMEX stock exchanges. Corporate use of foreign exchange or currency derivatives at the end of fiscal 1993 is determined from information contained in annual report financial footnotes. Segmented data on foreign sales activities is obtained from the S&P Compustat Geographic Segment database. Additional financial data is gathered from the S&P Compustat database.
Using the two samples, I apply empirical tests of the relation between a firm's compensation policy, ownership and organizational structure and its use of derivatives. After controlling for firm size and other firm attributes, I find that use of derivatives is increasing in the number of divisions which is consistent with the hypothesis that division managers hedge division performance. In particular, derivatives use is found to be increasing in the number of lines of business and the number of overseas operations. The results of tests of risk characteristics of top management indicate the derivatives use is unaffected by the riskiness of CEO compensation, the level of insider wealth vested in equity, or CEO retirement. However, derivatives use is decreasing in the inside ownership of the firm, Although preliminary, these results for top management are consistent with the notion that shareholders write optimal management contracts to mitigate opportunistic use of derivatives. Consistent with previous
studies, the likelihood of derivatives use is found to be increasing in firm size and decreasing in regulation.
The paper is organized as follows: Section 2 provides a brief overview of current theory on corporate use of derivatives and summarizes previous empirical work. Section 3 outlines the hypotheses to be tested. Section 4 describes the sample and data. The results of the empirical tests are presented in section 5. Finally, section 7 concludes and proposes issues to be considered in future research.
2. Theories of Derivatives Use
This study attempts to extend previous empirical investigations on the determinants of corporate use of derivatives. As background, this section outlines current theory and summarizes the results of previous empirical work on the determinants of corporate use of derivatives. First, the extant theories for corporate derivatives are outlined. These determinants will be considered as cross-sectional control factors in tests of the marginal impact of managerial motives on derivatives use. In developing a framework for analyzing corporate use of derivative instruments, it is generally assumed that shareholders maximize expected wealth. Therefore, shareholders will try to optimally contract with management and bondholders. Therefore, managerial motives for derivatives use must be take into account shareholder attempts to write optimal contracts that maximize shareholder wealth.
Use of derivatives will not alter firm value if there are no taxes, no contracting costs, no costs to financial distress, no information costs and capital market imperfections do not exist. Therefore, use of derivative securities can alter firm value if they affect expected taxes; if contracting, financial distress or information costs exist; or if there are imperfections in the capital markets. The discussion below examines these factors and how the use of derivatives for hedging purposes may increase shareholder wealth.
Taxes are Convex in Net Income:
The presence of convexities in the tax code imply that a firm's profit function may be concave in taxable income. Smith and Stulz (1985) outline how hedging with derivatives can decrease a firm's expected tax liability if it faces a progressive corporate tax schedule over a range of taxable income. This relation is derived from the convexity of the corporate tax schedule at low levels of taxable income. Therefore, the use of derivatives to hedge variability in the firm's expected stream of taxable income can reduce expected taxes. Although the progressivity in the tax schedule applies over a small range of taxable income, generous provisions for tax loss carry forwards and investment tax credits reinforce convexities over a larger range of taxable income. The use of derivatives to hedge taxable income can increase the present value of these tax shields by reducing variability in taxable income.
Nance, Smith and Smithson (1993) find weak support for the hypothesis that firms use derivatives to hedge taxable income. However, Francis and Stephan (1993), Mian (1994) and Geczy, Minton, and Schrand (1995) find no support for this hypothesis 2.
Financial Distress Costs: If a firm enters financial distress then it will face costs of default on debt obligations, costs of filing for bankruptcy, and costs related to reorganization and liquidation. Given these costs, firms have incentives to reduce the probability of financial distress. Smith and Stulz argue that firms can reduce the likelihood of financial distress by hedging variability in earnings. Empirical studies typically use a firm's leverage ratio and interest coverage ratio to proxy for firms with higher probability of incurring the costs of financial distress. Using these proxies for financial distress, Nance et al, Francis and Stephan, Mian, and Geczy et al find mixed or no support for the financial distress rationale for derivatives use.
Debtholder Contracting Costs: Debtholders take into account the incentives of equityholders to opportunistically transfer wealth in the future from debtholders. Therefore, debtholders will discount the value of such actions in the issue
2
The mixed results on the tax reasons for derivatives use may be due to misspecification of a firm’s true marginal tax rate. Graham (1995) develops an algorithm to calculate simulated marginal tax rates based on current and expected taxable income, net operating loss carryforwards and investment tax credits. Application of such an algorithm may provide a better measure of the covexities in the tax schedule faced by a firm.
price of debt. To minimize these agency costs, equityholders can bond against these actions by instituting restrictive debt covenants or use convertible debt. Conflict with debtholders can also arise if fixed claims in a firm's capital structure induces equityholders to forgo positive NPV projects if the benefits of these projects primarily accrue to debtholders. This underinvestment problem can be mitigated if derivatives are used to reduce the likelihood of states in which wealth does not accrue to equityholders. Nance, Smith and Smithson posit that firms with more growth opportunities in their investment set, proxied for by high levels of R&D and a low book-to-market ratio, have a higher likelihood of using derivatives to hedge variance in firm value. In addition, they suggest that this underinvestment problem will be most severe in firms with high debt to equity ratios and, therefore, these firms are more likely to use derivatives to hedge firm value. Their empirical results are found to be inconclusive on these determinants of derivatives use.
Imperfection in Capital Markets: Froot, Scharfstein, and Stein (1993) draw on previous work to develop a model of optimal hedging to offset inefficient investment. They hypothesize that if external sources of funds are more costly to a firm than internally generated funds then the firm could benefit from using derivatives. In particular, firms can hedge cash flows to avoid a shortfall in funds that may require a costly visit to the capital markets. Geczy et al find that use of derivatives is positively related to measures of the firm's investment opportunity set proxied for by R&D expenditures and related to the firms coverage ratio. They claim that these findings are consistent with the Froot, Scharfstein and Stein underinvestment explanation for corporate use of derivatives.
Regulation: Firms in regulated industries provide top management with few opportunities for discretion in corporate investment and financing decisions. Smith and Watts (1992) show that regulation is a key determinant of a firm's corporate financial policy. Therefore, if regulated firms face tighter scrutiny and face lower contracting costs, then they are less likely to use derivatives to hedge firm risk.
Substitutes for Derivatives: Nance, Smith and Smithson note that corporate use of derivatives is also influenced by other financial and operating policies. For example, instead of using off balance sheet instruments like derivatives to hedge, a firm could manage financial price risk on its balance sheet by structuring its assets and liabilities to reduce its exposure to movements in financial prices. Nance et al propose that a firm could use a) convertible debt to help control agency conflicts with debtholders, b) use preferred stock instead of debt to provide greater flexibility in situations of financial distress, or c) have a lower dividend yield to assure bondholders that sufficient funds are available to pay fixed claims. They find empirical results consistent with these hypotheses.
Firm Size: A cursory analysis of the relation between firm size and derivatives may indicate that small firms have a greater demand for derivatives to hedge their more volatile cashflows and equity prices. Moreover, large firms may be able to find cost-effective substitutes for derivatives by hedging risk through line-of-business and geographic diversification. Prior to current empirical work on derivatives use, Smith and Stulz posited that small firms would be more likely to hedge than large firms if a) bankruptcy costs are less than proportional to firm size and b) the cost of using of derivatives for hedging purposes is proportional to firm size. Clearly these are valid arguments supporting the hypothesis the derivatives use should be more prevalent among small firms.
However, all cross-sectional empirical studies on the determinants of derivatives use find that derivatives use is increasing in firm size. There are various theories to support the scale effect of derivatives. In their paper on accounting standards, Watts and Zimmerman (1978) suggest that firms
subject to high levels of outside scrutiny from investors or regulators face higher political costs. They posit that these firms may take actions to reduce these costs by decreasing firm visibility. Francis and Stephan (1993) suggest that more visible firms will wish to reduce scrutiny of variable earnings by using derivatives to hedge accounting earnings. They suggest that firm size proxies for actual firm visibility and that use of derivatives should therefore be increasing in firm size.
Nance et al and Geczy et al suggest that size is related to significant information and transaction cost scale economies in establishing the expertise to use derivatives for hedging or speculation. This suggests that derivatives use should again be increasing in firm size.
The proper specification of firm size impacts any empirical tests of derivatives use. Functional form tests performed on the data samples in this paper indicate that a logarithmic transformation of firm size is the proper functional form for this proxy3.
2.1 Managerial Motives and Derivatives Use
Divisional Incentives: Derivatives use within firms is not restricted to central treasury operations. Firms with overseas operating divisions and autonomous lines of business may use derivatives at the division level for risk management. The compensation of division managers is both directly and indirectly related to measures of division operating performance. A division manager's compensation may be directly tied to accounting measures of division performance and promotion and firing decisions are indirectly impacted by division operating performance. A risk-averse division manager may have incentives to use derivatives to hedge accounting measures of operating performance. Given that division managers are farther removed from shareholders, it may be more costly to write and implement optimal compensation contracts which mitigate suboptimal use of derivatives. For example, an upstream overseas operating division that is evaluated as a cost center cause the manager to bear all of the risk of exchange rate
3
Both Francis and Stephan (1993) and Nace, Smith and Smithson (1993) use untransformed assets levels as proxies for firm size. It is unclear whether this functional form is the proper specification of size across firms.
fluctuations on prices of manufactured products. Therefore, even though it may not be optimal for the firm to hedge these exchange rate exposures, the performance measurement system induces the division manager to use derivatives to hedge exchange rate risk.
Therefore, division managers may have more opportunities to use derivatives for personal risk reduction and it would be predicted that derivatives use will be increasing in the number of operating divisions within a firm. Given the widespread use of foreign exchange derivatives that can be traced to exposures of overseas divisions, it is predicted that use of foreign exchange derivatives use will also be increasing in the number of overseas divisions.
It should be noted there is an opposite hypothesis regarding the relation between derivatives use and organizational structure. In particular, firms with many distinct lines of business and overseas divisions may be naturally hedged. Therefore, a greater number of distinct operating divisions is seen as a substitute for hedging with derivatives. This leads to the alternate prediction that firms with a larger number of lines-of-business and a larger number of overseas operating divisions would be less likely to use derivatives.
Undiversified Managerial Capital Vested in the Firm : Relative to shareholders, managers have a higher demand for firm-specific risk reduction because a manager's portfolio of human and investment capital is mostly tied to firm performance while shareholders' investments are well diversified. Therefore, managers who have more human capital invested in the firm have higher demand for risk reduction. May (1995) proxies for underdiversified CEO human capital by measuring the number of years a CEO has been employed at various levels of executive responsibility within the firm. May suggests that CEO managerial skills become more firm-specific as time spent with the firm increases. Given that hedging reduces human capital risk, firms whose CEOs have more years vested are more likely to use derivatives. CEO years with the firm also proxies for other factors that may affect the decision to use derivatives. For example, it may capture
managerial entrenchment, and more entrenched managers may have greater flexibility in using derivatives for managerial motives vs. stockholder wealth maximization4.
CEO and Insider Wealth Vested in Firm Equity: Jensen and Meckling (1976) argue that agency investment conflicts may result when managers hold an undiversified portfolio that consists primarily of firm equity. The problem occurs because managers cannot easily diversify their wealth invested in firm equity. In such a case, managers may undertake actions may reduce firm value if they reduce firm-specific return variance, while equityholders who have perfectly hedged away firm-specific risk through diversification are concerned only with maximizing firm value. Under this hypothesis we would expect that CEOs (and more generally insiders) with higher levels of personal wealth vested in firm equity would have a higher demand for firm-specific risk reduction. As May (1995) notes, the problem with testing this theory is that while one can measure insider wealth vested in firm equity, one cannot easily determine other elements of insiders' personal portfolios. One could use the ratio of CEO equity wealth to current cash compensation to proxy for wealth vested in firm equity. While this ratio is easy to measure, it will overstate CEO wealth vested in firm equity because it captures the accumulation of equity wealth but not the accumulation of potentially well-diversified cash compensation. Therefore, measures of insider wealth vested in equity may be biased measures of undiversified equity investments in the firm.
CEO Compensation Tied to Firm Performance: If CEO cash compensation contains a large risk component that is outside his control, then he may wish to use derivatives to hedge this risk. In particular, a compensation plan that includes a large bonus component based on accounting measures of performance may induce a manager to hedge accounting numbers5. In addition, if a CEO has contracted to receive a post retirement compensation package based on the firm's accounting performance prior to the CEO's retirement, then he may use derivatives to hedge accounting numbers to reduce the risk of this compensation. It should be noted that, in theory, the incentives for top management to engage in suboptimal use of derivatives can be
4
Years of actual service with the firm must be hand collected from various sources. I expect to collect this data to include in future versions of this paper.
eliminated if shareholders can write optimal compensation contracts. However, such contracts may be costly to negotiate and implement. In addition, diffuse ownership of equity may reduce the incentives for individual shareholders to write and enforce optimal management contracts.
Firms with greater growth opportunities have higher levels of performance-rated compensation. Smith and Watts (1992) document the relationship between a firm's investment opportunity set and its financing and compensation policies. Given that CEO compensation is endogenously determined, then the impact of performance-based compensation on the managerial incentives to use of derivatives may be difficult to measure.
3. Empirical Hypotheses
The preceding discussion on the determinants of derivatives use provides a framework for empirical tests of the relation between corporate use of derivatives and a firm's compensation policy, ownership and organizational structure. This section presents the main empirical hypotheses to be tested.
3.1 CEO Compensation
As outlined in section 2, if CEO cash compensation contains a large risk component that is outside the CEO's control, then he may wish to use derivatives to hedge this risk. In particular, a compensation plan that includes a large bonus component based on accounting measures of performance may induce a manager to hedge accounting numbers. In addition, if a CEO has contracted to receive a post retirement compensation package based on the firm's accounting performance prior to his retirement, then he may use derivatives to hedge accounting numbers to reduce the risk of this compensation. These theories lead to the following empirical hypotheses:
5
See Healy (1985) for a discussion of CEO bonus plans.
H1: Other things equal, the higher the ratio of bonus compensation to total CEO cash compensation, the greater the likelihood that the firm will use derivatives. H2: Other things equal, the closer a CEO is to retirement, the greater the likelihood that the firm will use derivatives.
3.2 Insider Wealth Vested in Firm Equity
Insiders and top managers with high levels of personal wealth vested in firm equity may use derivatives to hedge firm value to reduce the risk of their undiversified investment portfolios. This leads to hypothesis 3:
H3: Other things equal, the higher the level of wealth insiders have vested infirm equity, the greater the likelihood that the firm will use derivatives.
An underlying assumption of hypotheses H1, H2, and H3 is that shareholders cannot write and implement optimal contracts with top management to mitigate suboptimal use of derivatives. Rejection of these hypotheses would be consistent with the theory that shareholders can effectively implement contracts with top management that remove opportunities for value-reducing investments in derivatives. Therefore, these contracts have aligned managerial incentives with the interests of shareholders and any remaining opportunities to hedge managerial risk with derivatives are too costly. If equity ownership in the firm is diffuse, then there may be weak incentives for individual shareholders to write and implement optimal managerial contracts. As a proxy for ownership concentration, I use the percent of total shares outstanding held by insiders6. This leads to the fourth hypothesis:
H4: Other things equal, the greater the concentration of inside ownership in the firm's equity, the lower the likelihood that the firm will use derivatives.
6
It could also be argued that concentrated inside ownership would give insiders greater latitude to take actions at the expense of outside shareholders. Therefore, this proxy may have uncertain implications for derivatives use.
3.3 Divisional Incentives
In a decentralized firm, division managers have may incentives to reduce risk in measures of division performance if their compensation is based on these division operating performance. Given that division managers are further removed from shareholders, it may be difficult for shareholders to write and implement contracts which limit opportunistic use of derivatives by division managers. This leads to the following predictions:
H5: Holding firm size constant, the larger the number of lines of business within the firm, the greater the likelihood that the firm will use derivatives. H6: Holding firm size and relative level of foreign sales constant, the larger the number of overseas operations, the greater the likelihood that the firm will use derivatives.
As outlined in section 2, the existence of numerous distinct lines of business and overseas operations may provide a firm with a natural hedge. This result would suggest that diversified firms are less likely to use derivatives. Therefore, failure to reject hypotheses H5 and H6 should be considered strong support for the division manager theory of derivatives use.
4. Data and Sample Selection
The empirical analysis focuses on two samples of firms that include derivatives users and non-users. The first sample is based on fiscal year 1994 financial statement information collected for 403 non-financial firms that trade on the New York (NYSE) and American (AMEX) stock exchanges. The analysis uses data collected from financial footnotes in annual reports to determine use of any type of derivative security at the end of fiscal 1994. Over 46% of the firms in this sample report using derivatives. CEO salary and bonus compensation, insider holdings of stock, line of business classifications, and other financial information are collected from electronic filings of proxy statements with the SEC and from the Standard and Poor's (S&P) Disclosure and Compustat databases. The sample selection procedure for the first sample is outlined in Table 1A.
Recent changes to disclosure requirements for compensation of top executives have provided a valuable source of information about CEO pay. CEO salary and bonus compensation data for fiscal years 1992, 1993 and 1994 were hand collected for the 403 firms in the sample from electronic filings of proxy statements with the SEC.
The second sample is based on data collected for a paper investigating the determinants of foreign exchange derivatives (FXD) use by U.S. corporations 7. The data covers fiscal year 1993 financial statement information collected for 807 non-financial firms that trade on the NASDAQ, NYSE and AMEX stock exchanges. Corporate use of foreign exchange or currency derivatives at the end of fiscal 1993 is determined from information contained in annual report financial footnotes. Segmented data on foreign sales activities is obtained from the S&P Compustat Geographic Segment database. Additional financial data is gathered from the S&P Compustat database. The sample selection procedure for the FXD sample is presented in Table 1A.
The definitions of and construction procedure for the proxy variables used in the empirical tests are outlined in Table 2.
5. Empirical Evidence
5.1 Univariate tests
Univariate tests of differences means of the variables hypothesized to impact a firm's use of derivatives are presented in Table 3. Definitions of the variables can be found in Table 2. The sample is comprised of 403 firm-year observations for the various variables including 215 observations for firms reporting the use of any type of derivative at the end of fiscal 1994. Of the "managerial motive" variables of interest, all of are differences in means between derivatives users and non-users have predicted sign and only the difference in the means of A_RETIRE is statistically insignificant.
The univariate results indicate that derivatives users have significantly higher levels of CEO cash compensation risk, higher levels of insider wealth vested in firm equity, a larger number of lines-of business and lower levels of insider ownership. The results of the univariate tests for the tax, financial distress, derivatives substitutes and other control variables are summarized in Table 1.
5.2 Multivariate Logit Regression Results
To control for interactions among the hypothesize determinants of derivatives use, we estimate logistic regressions using a categorical dependent variable describing derivatives use. Table 5 outlines the results of various specifications of the logistic model. Model 1 is estimated as a benchmark case that does not include CEO and insider determinants of derivatives use. It is found that only S_DIVYLD, O_SIZE and A_DIVISION are significant determinants of derivatives use. It should be noted that S_DIVYLD has the opposite sign to that predicted by Nance, Smith and Smithson. An examination of the correlation between S_DIVYLD and O_REGLTL indicates that they are highly correlated ( see Table 4). Therefore, model 3 is estimated without S_DIVYLD and it is found that O_REGUL comes in strongly significant and of the predicted sign. It appears that S_DIVYLD may just be a proxy for regulated firms in this sample. A_DIVISION is found to be of the predicted sign and is consistent with the division manager hypothesis for derivatives use.
In model 3, all hypothesized determinants of derivatives use are included. Again, S_DIVYLD, O_SIZE and A_DIVISION are found to be significant and have the same sign as models 1 and 2. In addition, A_EQUITYINSIDE is found to be significant and of the predicted sign. This result is consistent with the diffuse ownership hypothesis. It should be noted that A_RETIRE and A_COMPRISK are found to be insignificant determinants of derivatives use8. This result is consistent
7 8
See Appendix A for a summary of the results of Wysocki (1995). Alternate specifications of A_COMPRISK were also tested. The average ration of bonus to total compensation over the three year period of 1992-94 and the ratio of the maximum bonus over the past three years to current total cash compensation in 1994 were tested. These variables also proved to be insignificant determinants of derivatives use.
with the notion that shareholders write and implement optimal contracts that remove incentives for top management to engage in suboptimal use of derivatives. Model 4 is presented as a specification test excluding the A-RETIRE dummy. The results are remain unchanged as compared to model 3.
To obtain a better intuition on the impact of the determinants an OLS regression of the binary dependent variable on the explanatory variables is estimated. The coefficients from this linear probability model are unbiased and can be directly interpreted as marginal probabilities. The results of this regression are presented in Table 6. The p-values of the OLS regression are presented for comparative purposes only as they are biased estimates of the variance of the coefficient estimates. The linear probability model is found to explain 33% of variation in derivatives use in this sample of firms.
Table 6 and 7 present the results of the multivariate logit regression of foreign exchange derivatives use on various explanatory variables. After controlling for firm size, relative levels of foreign sales activity, foreign tax liability and other factors, it is found that the number of countries/regions in which a firm has foreign sales operations is found to positively and significantly related to foreign exchange derivatives use. This is consistent with the division manager hypothesis for derivatives use. A complete discussion of the other variables in this regression can be found in Appendix A.
6. Conclusions and Future Research
In this paper, I include corporate compensation policies, ownership structure and organizational structure as possible determinants of derivatives use. First, I examine the relation between compensation of top management and the use of derivatives by managers to reduce personal risk exposure. The results of the multivariate tests indicate that firms with high risk in CEO cash compensation, high levels of insider wealth vested in firm equity, or high probability of CEO retirement are not more likely to use derivatives. However, the likelihood of derivatives use is found to be decreasing in inside ownership. Although these results are preliminary, they are consistent with the view that shareholders account for managerial risk aversion when writing optimal compensation contracts for top executives. Second, I
investigate whether firms with a larger number of divisions are more likely to hedge. The empirical results are consistent with the hypothesis that division managers that are far removed from shareholders have incentives to use derivatives to reduce risk in measures of division performance. I find that derivatives use is increasing in both the number of lines of business and the number of overseas operations of a firm.
Consistent with previous studies, I find that the likelihood of derivatives use is found to be increasing in firm size and decreasing in regulation. In addition, I find that derivatives users and non-users are not different with respect to proxies of the firm's tax position.
Although this paper provides new empirical evidence on link between managerial incentives and derivatives use, there are a number of issues that should be explored. In particular, I hope to implement better proxies of managerial compensation including CEO holdings of firm equity, CEO compensation based on stock options, and measures of CEO human capital invested in the firm. In addition, I plan to develop a better model and gather additional data on a firm's divisional characteristics division manager compensation and division level use of derivatives. I hope to draw on the work of Bushman, Indejejikian and Smith (1994 and 1995) and Christie, Joye and Watts (1993) and Baiman, Larcker and Rajan 9195) to develop a more complete model of why derivatives use is increasing in the number of divisions within the firm.
Data Sources:
•
Standard and Poor's Disclosure Database (SEC Edition)- December 1995 release (Financial footnotes and fulltext of SEC filings used to determine derivatives holdings of firms with December 31, 1994 fiscal year end. Also used to determine CEO age, and insider stock holdings.
• •
Standard and Poor's CONTUSTAT Database - Annual Industrial File - 1994 Edition SEC EDGAR Database - Proxy statements filings. Used to determine salary and bonus compensation of Chief Executive Officers of firms in the sample.
References Baiman, S., D.F. Larcker, and M.V. Rajan, 1995, Organizational Design of Business Units, Journal of Accounting Research 33, 205-229. Bushman, R., R. Indejejikian, and A. Smith, 1994, Compensation contracts in hierarchies: determinants of incentive pat for business unit managers", University of Chicago working paper. Bushman, R., R. Indejejikian, and A. Smith, 1995, Aggregate performance measures in business unit manager compensation: the role of intra-firm interdependencies", University of Chicago working paper. Christie, A., M. Joye and R. Watts, 1993, Decentralization of the firm: theory and evidence", University of Rochester working paper. DeMarzo, P.M., and D. Duffie, 1991, Corporate Financial Hedging with Proprietary Information, Journal of Economic Theory 53, 261-286. DeMarzo, P.M., and D. Duffie, 1995, Corporate Incentives for Hedging and Hedge Accounting, Review of Financial Studies, Vol. 8, No. 3, 743 -77 1. Francis, J., and J. Stephan, 1993, Characteristics of Hedging Firms: An Empirical Investigation, in Advanced Strategies in Financial Risk Management, R.J. Swartz and CW. Smith eds., 615 -65 8. Froot, K.A., D.S. Scharfstein, and J.C. Stein, 1993, Risk Management: Coordinating Corporate Investment and Financing Policies, Journal of Finance, Vol XLVIII, No. 5, 1629-1658. Geczy, C., B.A. Minton, and C. Schrand, 1995, Why Firms Use Derivatives: Distinguishing Among Existing Theories, Working Paper, Fisher College of Business, The Ohio State University. Graham, JR., 1995, Debt and the marginal tax rate, Working Paper, David Eccles School of Business, The University of Utah. Healy, P.M., 1985, The effect of bonus schemes on accounting decisions, Journal of Accounting and Economics 7, 85-107. Hentschel, L., and S.P. Kothari, 1995, Life Insurance or Lottery: Are Corporations Managing or Taking Risks with Derivatives?, Working Paper, William E. Simon School of Business Administration, University of Rochester.
Jensen, M.C., and W.H. Meckling, 1976, Theory of the firm: Managerial behavior, agency costs and ownership structure, Journal of Financial Economics 3, 305-360. May, D.O., 1995, Do Managerial Motives Influence Firm Risk Reduction Strategies, Journal of Finance, Vol. L, No. 4, 1291-1308. Mian, S.L., 1994, Evidence on the determinants of corporate hedging policy, Working Paper, Emory Business School, Emory University. Nance, DR., CW. Smith, and CW. Smithson, 1993, On the Determinants of Corporate Hedging, Journal of Finance, Vol. XLVIII, No. 1, 267-284. Smith, C.W. and R.M. Stulz, 1985, The Determinants of Firms' Hedging Policies, Journal of Financial and Quantitative Analysis, Vol. 20, No. 4, 391405, Smith, C.W., and R.L. Watts, 1992, The investment opportunity set and corporate financing, dividend and compensation policy, Journal of Financial Economics 3 2, 263 -292. Smithson, CW., 1996, Managing Financial Risk: 1996 Yearbook (CIBC/Wood Gundy Publications). Stulz, R.M., 1984, Optimal Hedging Policies, Journal of Financial and Quantitative Analysis, Vol. 19, 127-140. Warfield, T.D., J.J. Wild, and K.L. Wild, 1995, Managerial ownership, accounting choices and informativeness of earnings, Journal of Accounting and Economics 20, 61-9 1. Watts, R.L. and J.L. Zimmerman, 1978, Towards a positive theory of the determination of accounting standards, Accounting Review 53, 112-134. Wysocki, P.D., 1995, Determinants of Foreign Exchange Derivatives Use by U.S. Corporations: An Empirical Investigation, Simon School of Business Summer Working Paper, University of Rochester.
Appendix A: Summary of "Determinants of Foreign Exchange Derivatives Use by U.S. Corporations: An Empirical Investigation" Peter Wysocki, 1995 Working Paper
This paper extends the previous literature by focusing on the specific determinants of firms’ foreign exchange hedging behavior. I apply current theory on foreign exchange risk management to develop a set of hypotheses on the determinants of firms' use of foreign exchange derivatives (FXDs). These hypotheses are tested on a large sample of firms using newly available data. In addition, I make novel use of geographic segment information to obtain a detailed measure of the extent and character of foreign operations of U.S. firms. The results of my hypothesis tests indicate that in 1993 U.S. firms with FXD holdings were typically large; had high levels of transaction and translation exposures as measured by relative levels of foreign sales and taxes; tended to operate in numerous countries; and, were likely to have foreign sales revenue from Japan. Inconclusive results are found for the hypothesis that FXD hedgers are more likely to have indirect economic exposures as measured by industry-level foreign import competition. We believe that these results help provide a more complete understanding of the general determinants of corporate hedging as well as the specific determinants of FXD hedging behavior of firms.
The following hypotheses were tested in the paper:
Transaction Exposure: Accounting earnings of U.S. firms can be affected by foreign exchange fluctuations by changing the expected results of transactions in non-local currencies. This is known as transaction exposure. A U.S. firm that derives a substantial fraction of its sales revenues from foreign sources would likely be impacted by this type of exposure. This argument leads to the hypothesis:
FXD HI: Foreign exchange hedging firms derive a larger proportion of sales revenues from foreign sources than nonhedging firms.
Translation Exposure: Translation gains and losses arise when a company translates financial statements from its overseas operations from a foreign currency to the U.S. dollar. As specified under SFAS No.52, an overseas subsidiary that designates the U.S. dollar as its functional currency must include translation gains and losses in the determination of income. Therefore, firms that have high levels of overseas assets are more likely to have a accounting translation exposure. Under SFAS 14, firms are required to report geographically segmented information on significant foreign sales activity and foreign assets. Data on foreign assets was unavailable for this study. Consequently, foreign taxes, a common proxy for foreign assets, are used as an indicator of possible translation exposures9.
FXD H2: Foreign exchange hedging firms have a higher fraction of foreign tax incidence than nonhedging firms.
Economic Exposure: Economic exposures impact a firm's competitive position and are caused by exchange rate induced changes in input and output prices. Many of these exposures are proxied for by transaction and translation exposures. However, firms without direct foreign activities, sales or assets can also be impacted by changes in exchange rates. These exposures can result from a number of factors including foreign competitors that benefit from exchange rate changes, customers whose activities are impacted by foreign exchange factors, or suppliers that use foreign-sourced inputs. Firms that are in domestic markets with meaningful foreign import competition can be adversely affected by an appreciation in the U.S. dollar. It is posited that foreign competitors with meaningful U.S. market shares would be in a strong position to benefit from an appreciation in the U.S. dollar. Also, larger firms are more likely to have a wide range of suppliers and customers, increasing the likelihood that the firms costs or revenues may be impacted by foreign exchange fluctuations.
FXD H3: Foreign exchange hedging firms face higher levels on industry-level foreign import competition. FXD H4: Foreign exchange hedging firms are larger than nonhedging firms.
9
Lee and Kwok (1998) and Teerathorn, Alcerraca-Joaquin and Seigel (1992) use foreign tax ratios to proxy for the level of foreign assets. Lee and Kwok examine a sample of firms and find significant correlations between foreign
Information and Transaction Scale Economies: A firm must incur substantial fixed costs to either contract with outside specialists or employ skilled and informed managers to manage a foreign exchange hedging program. Nance, Smith, and Smithson note that a hedging program exhibits informational and transaction scale economies. Clearly, larger firms can spread these fixed costs over a larger set of foreign activities and exposures.
FXD H5: Foreign Exchange hedging firms are larger than nonhedging firms.
Developed Foreign Exchange Markets: The existence of liquid and well-developed markets for trading foreign exchange contracts will reduce that cost of implementing a foreign exchange hedging program. Organized derivatives exchanges provide unencumbered trading opportunities, high volumes, and decreased risk of counterparty default. If such markets do not exist, then a firm may have to pay a substantial premium to purchase foreign exchange derivatives contracts in the over-the-counter market. Liquid public U.S. dollar foreign exchange markets exist for major currencies such as the Gennan mark, French franc, British pound, Australian and Canadian dollars, and the Japanese yen. On the other hand, liquid markets do not exist for currencies of developing economies. Therefore, FXD hedging may not be a cost-effective option for hedging exposures in these currencies.
FXD H6: Foreign exchange hedging firms are more likely to derive sales from Europe, Japan, Canada and Australia. FXD H7: Foreign exchange hedging firms have equal probability of deriving sales from South America, Africa, and Mexico than nonhedging firms.
asset ratios and foreign tax ratios.
Table 1A: Sample Selection Procedures for Main Sample
Criteria Population of firms on Disclosure Database Firm’s shares trade on NYSE or AMEX AND Firm is incorporated in the U.S. AND Primary SIC outside range 6000-6999 AND Financial data in U.S. dollars AND Financial footnote information provided AND December 31 fiscal year-end AND Financial footnotes with unambiguous classification of derivatives use Total Number of Firms in Disclosure Sample Firms without matching SEC proxy statement data or financial data on Compustat Firms remaining in Main Sample Derivatives user subsample Firms contained in Main Sample Firms reporting the following in financial footnotes: • hedge(s), hedged, hedger, hedging • swap(s), swaption(s) • forward • futures • cap, collar, floor • option(s) • foreign exchange contract • currency contract derivative Firms in Main Sample reporting using derivatives at fiscal year end (adjustment for misclassified firms in keyword search) Number of firm observations
2452 2375 2047 1907 1724 1045 980 980 577
403 403 209
188
Table 1B: Sample Selection Procedure for Foreign Exchange Derivatives Sample
Criteria Firm’s shares trade on NYSE, AMEX or NASDAQ stock exchange AND fiscal 1993 financial footnote information contained in Disclosure database AND firm is active in fiscal 1993 reporting period AND firm is incorporated in the US AND primary SIC outside the range 6000-6999 Number of observations not on Compustat “ALL FIRMS” FXD Hedger Subsample Firms with “ALL FIRMS” reporting keywords in full text of Disclosure database: • currency(s) • OR exchange rate(s) • OR foreign exchange AND • derivative(s) • OR hedge(s) OR hedged • OR forward • OR futures • OR option(s) • OR risk management • OR swap(tions) Number of firms without FXD holdings “FXD HEDGER” (MAIN SAMPLE) Subsample A (Include only firms with SIC in range 0-3999) Hedgers in Subsample A Subsample B (Include only firms with (a) SIC in range 0-3999 and (b) report data for foreign and domestic taxes Hedgers in Subsample B Number of firm observations 6613 2334 2314 2163 1902 1902 143 1759
417
417 106 311 1004 248 807 234
Table 2: Definition of Variables
This table summarizes all variables used in the analysis including a detailed description of the method of calculation
Variable Name Tax Variables T_NOL T_ITC Bankruptcy Variables B_DEBTEQ B_COVERAGE Derivatives Substitutes S_CONVERT S_PREFSHR S_DIVYLD Organizational Form O_REGUL O_SIZE Description
Net operating losses applied as a reduction of taxable income in 1994 (Compustat item 52) Dummy variable equal to 1 if firm has investment tax credits on its balance sheet (derived from Compustat item 208) Debt-equity ratio: ratio of 1994 book value of LT debt (Compustat items 35 and 9) to firm size (see below) Interest coverage ratio: 1994 ratio of pretax income (item 170) plus interest expense (item 5) to interest expense Ratio of total convertible debt (Compustat item 79) to firm size Ratio of book-value of total preferred stack (Compustat item 130) to firm size Dividend yield: ratio of cash dividend per share (Compustat item 26) to closing price per share at end of fiscal year 1994 (Compustat item 199) Dummy variable equal to 1 if firm’s primary 2-digit SIC is 49 Market value of the firm at end of fiscal 1994. Defined as the sum of mkt value of equity (Compustat item 199 times item 25) plus book value of LT debt (item numbers 9 and 34) plus book value of preferred stock (item number 130). O_SIZE is constructed using the logarithm of the market value of the firm. Ratio of book value to market value of the firm. Book value = Total assets minus total liabilities less preferred stock (Compustat items 6-181-130). Market value = firm size (Actual CEO bonus in 1994) / (Actual CEO salary plus bonus) Number of held shares held by insiders / Total number of shares outstanding Total number of different 2 digit SIC classifications for the firm Number of shares held by insiders times market value of shares at fiscal year end (Compustat item 24) Dummy variable equal to 1if CEO is age 64 or 65 at fiscal year end 1994
Agency Costs A_BKMKT A_COMPRISK A_EQUITYINSIDE A_DIVISION A_WEATHEQUITY A_RETIRE FXD Sample Multi-country Sales Foreign Imports Country-specific Sales Foreign Tax Ratio Foreign Sales Ratio
Number of countries in which the firm reports segmented sales operations U.S. Domestic market share held foreign companies in the firm’s primary 4 digit SIC in 1993. A dummy variable equal to 1 if a firm has foreign sales operations in the identified country/region in 1993. Ratio of foreign taxes paid to total taxes paid in fiscal 1993 Ratio of total foreign sales/revenues to total firm sales/revenues in 1993
Table 3: Summary Statistics for Characteristics of Derivative Users and Non-Users
Sample of 403 firms including 215 firms reporting the use of any type of derivative at the end of fiscal yr 1994. Data measure at fiscal year-ends. Derivative Users (N=215) Variable Mean Tax Variables T_NOL ($MM) T_ITC (dummy) 36.857 0.404 9.507 0.046 7.731 0.365 6.843 0.033 3.063 0.861 0.002 0.389 Std. Dev. Mean Std. Dev. Derivative Nonusers (N=188) Test of Difference in Means (Users – Nonusers) t-stat p-value
Bankruptcy Variables B_DEBTEQ B_COVERAGE Derivatives Substitutes S_CONVERT S_PREFSHR S_DIVYLD Organizational Form O_REGUL (dummy) O_SIZE (log($MM)) 0.078 7.632 0.029 0.165 0.136 5.681 0.021 0.118 -1.963 11.848 0.050 0.000 0.017 0.011 0.020 0.006 0.004 0.002 0.021 0.018 0.019 0.005 0.003 0.002 -0.757 -1.972 0.187 0.449 0.049 0.852 0.852 11.121 0.859 7.153 5.673 19.073 0.620 5.243 -5.613 -1.111 0.000 0.267
Agency Costs A_BKMKT A_COMPRISK A_EQUITYINSIDE A_DIVISION A_WEALTHEQUITY (log $MM) A_RETIRE (dummy) 0.335 0.365 0.094 2.609 3.734 0.052 0.033 0.020 0.019 0.118 0.179 0.021 0.448 0.245 0.189 1.935 2.436 0.051 0.023 0.014 0.014 0.085 0.130 0.015 -3.458 6.089 -5.073 5.706 7.243 0.037 0.001 0.000 0.000 0.000 0.000 0.971
TABLE 4: Pearson Correlation Coefficients for Variables
Sample of 403 firms including 215 firms reporting the use of any type of derivative security at the end of fiscal year 1994. Data measured at fiscal year-ends.
A_EQUITYINSIDE
A_COMPRISK
A_DIVISION
A_WEALTH
S_CONVERT
A_BKMKT
B_DEBTEQ
DER_USER
A_RETIRE
S_DIVYLD
B_COVER
O_SIZE
S_PREF
DER_USER A_BKMKT A_COMPRISK A_EQUITYINSIDE A_RETIRE A_WEALTH B_COVER B_DEBTEQ A_DIVISION O_REG O_SIZE S_CONVERT S_DIVYLD S_PREF T_ITC T_NOL
1.000 -0.152 0.288 -0.254 -.005 0.325 0.006 -0.188 0.267 -0.112 0.490 -0.043 -0.017 -0.078 0.044 0.144 1.000 -.0168 0.120 -0.001 -0.228 0.013 0.217 -0.072 0.021 -0.363 -0.085 -0.007 -0.131 0.026 -0.055 1.000 -0.132 0.007 0.368 0.041 -0.127 0.179 -0.184 0.430 -0.013 -0.043 -0.050 0.007 0.092 1.000 -0.006 0.272 -0.012 0.095 -0.124 -0.193 -0.419 0.076 -0.269 -0.110 0.042 -0.054 1.000 0.006 -0.016 0.025 -0.045 0.027 -0.002 -0.002 0.167 0.015 0.044 0.015 1.000 0.071 -0.230 0.187 -0.263 0.521 -0.043 -0.123 -0.123 -0.030 0.150 =5% signif 1.000 -0.019 -0.017 -0.025 0.019 -0.022 -0.011 -0.027 0.024 -0.017 1.000 -0.097 -0.055 -0.321 -0.002 -0.097 -0.054 -0.029 -0.029 1.000 -0.104 0.318 -0.081 0.129 -0.016 0.032 0.087 1.000 0.128 -0.076 0.815 0.294 0.008 -0.078 1.000 -.0114 0.313 0.064 -0.025 0.156 1.000 -0.139 0.041 0.010 -0.002 1.000 0.162 0.048 -0.064 1.000 -0.046 -0.016 1.000 -0.030 1.000
=10% signif
=1% signif
T_NOL
O-REG
T_ITC
Table 5: Estimates of logistic models relating probability of derivatives use to firm-specific characteristics (Main sample of derivatives users and non-users)
Sample of 403 firms including 215 firms reporting the use of any type of derivative at the end of fiscal year 1994. Independent Variable Model 1 Model 2 Model 3 Model 4 Model 5 (OLS) R = 0.33 -0.199 (0.961)
Intercept
-4.790 (<0.001)
-4.561 (<0.001)
-4.059 (<0.001)
-4.316 (<0.001)
Tax Variables T_NOL T_ITC 0.003 (0.194) 0.397 (0.112) 0.003 (0.176) 0.340 (0.167) 0.003 (0.261) 0.492 (0.062) 0.003 (0.259) 0.498 (0.059) 0.0002 (0.314) 0.054 (0.217)
Bankruptcy Variables B_DEBTEQ B_COVERAGE -0.050 (0.083) 0.000001 (0.734) -0.050 (0.083) 0.000001 (0.748) -0.048 (0.097) 0.000002 (0.620) -0.050 (0.079) 0.000002 (0.686) -0.0007 (0.515) 0.0000001 (0.887)
Derivatives Substitutes S_CONVERT S_PREFSHR S_DIVYLD 0.054 (0.976) -3.689 (0.296) -15.625 (0.038) 0.385 (0.828) -4.076 (0.255) -0.190 (0.919) -4.542 (0.211) -18.762 (0.023) -0.209 (0.911) -4.767 (0.191) -19.607 (0.017) 0.097 (0.764) -0.805 (0.195) -2.787 (0.021)
Organizational Form O_REGUL O_SIZE -0.486 (0.338) 0.681 (<0.001) -1.153 (0.004) 0.632 (<0.001) -0.419 (0.451) 0.554 (<0.001) -0.536 (0.322) 0.631 (<0.001) -0.055 (0.556) 0.091 (<0.001)
Agency Costs A_BKMKT A_COMPRISK A_EQUITYINSIDE A_WEALTHEQUITY A_DIVISION A-RETIRE 0.212 (0.037) 0.186 (0.064) 0.521 (0.233) 0.419 (0.338) 0.406 (0.373) 0.275 (0.681) -2.095 (0.034) 0.098 (0.392) 0.265 (0.014) 0.308 (0.618) 0.415 (0.362) 0.311 (0.641) -1.528 (0.034) 0.069 (0.306) -0.307 (0.102) -0.440 (0.005) 0.028 (0.128) 0.043 (0.015)
0.266 (0.014)
Table 6: Pearson Correlation Coefficients Between Proxy Variables for 807 Firms (1993 Data)
Sample of 807 firms including 234 firms reporting foreign exchange derivatives holdings at the end of fiscal year 1993. For. Tax Ration For. Tax Ratio Africa Sales Austral. Sales Canada Sales Europe Sales Foreign Imports Japan Sales Mexico Sales MultiCountry Sales South Am. Sales For. Sales Ratio Africa Sales Austral. Sales Canada Sales Europe Sales For. Import Japan Sales Mex. Sales Divers. For. Sales South Amer. Sales For. Sales Ratio Firm Size
1
0.269 0.255 0.270 0.415 0.041 0.167 0.212 0.507
1 0.215 0.187 0.235 -0.027 0.059 0.156 0.457 1 0.339 0.347 -0.028 0.254 0.309 0.537 1 0.394 -0.097 0.151 0.360 0.562 1 0.029 0.269 0.274 0.775 1 0.015 -0.060 -0.026 1 0.173 0.389 1 0.504 1
0.324 0.666
0.402 0.309
0.378 0.323
0.295 0.342
0.331 0.592
-0.061 0.060
0.185 0.322
0.452 0.290
0.144 0.127
1 0.400 1
Firm Size 0.175 0.197 0.093 0.111 0.107 -0.001 0.023 0.161 0.071 0.188 0.198 1
Table 7: Estimates of logistic model relation probability of foreign exchange derivatives use to firm-specific characteristics
Notes: N=Number of firms in sample, D=Number of foreign exchange derivatives users is sample, FXD=foreign exchange derivatives Sample Statistics in table are: Mean / Standard Error / Probability Value Theory Predicted Sign Subsample B (N=807, D=234) Transaction and Economic Exposure 1.378 + 0.751 0.067 • Information & Transaction 0.336 Scale Economies + 0.061 • Political Costs 0.000 • Increased Economic Exposure 0.325 Division Manager Hedging Risk in + 0.098 Measures of Division Performance 0.001 0.375 Developed FXD Market + 0.360 0.297 0.751 Developed FXD Market + 0.358 0.036 -0.609 Developed FXD Market + 0.255 0.017 0.197 Developed FXD Market + 0.318 0.536 0.061 Undeveloped FXD Market 0 0.295 0.838 -0.203 Undeveloped FXD Market 0 0.317 0.522 0.274 Undeveloped FXD Market 0 0.472 0.561 0.0070 Hedging Indirect Economic + 0.0067 Exposure 0.296 0.891 Hedging Translation Exposure + 0.389 0.022
Proxy Variable Foreign Sales Ratio Firm Size
# of Regions with Foreign Sales
Sales in Europe
Sales in Japan
Sales in Canada Sales in Australia Sales in Mexico
Sales in South America Sales in Africa
Degree of Foreign Import Penetration Foreign Tax Ratio
Table 8: Results of Compensation Policy Regressions on Firm Size, Dividend Yield, Regulation, Book to Market, and Derivative User Variables
Sample of 403 firms including 188 firms reporting the use of derivatives of fiscal year-end 1994 (All data in 1994)
Dependent Variable
Regression R2
Intercept
O_SIZE
S_DIVYLD
O_REG
A_BKMKT
DER_USER
A_WEALTHEQUITY
0.393
-0.3528 (0.327)
0.676 (<0.001)
-11.962 (0.004)
-1.483 (<0.001)
-0.125 (0.589)
0.0354 (0.841)
A_COMPRISK
0.253
-0.249 (0.5352)
0.0519 (<0.001)
-0.596 (0.234)
-0.149 (<0.001)
0.015 (0.594)
0.009 (0.665)
A_EQUITYINSIDE
0.209
0.393 (<0.001)
-0.031 (<0.001)
-0.958 (0.043)
-0.056 (0.108)
-0.005 (0.846)
-0.043 (0.033)
doc_999049877.pdf
To become a dictionary definition of managerial effectiveness, you must be able to lead a group of people. Leaders lead in different ways, but getting results usually involves motivational techniques.
Managerial Motives and Corporate Use of Derivatives: Some Evidence
Peter D. Wysocki* University of Michigan Business School 701 Tappan Street, Ann Arbor, MI 48109-1234, USA Telephone: (734) 763-6820, E-mail: [email protected]
July 1998
Abstract
This paper provides empirical evidence on the link between a firm's use of derivatives and its compensation policy, ownership structure and organizational structure. Agency theory indicates that top management may have incentives to use derivatives to reduce personal risk if CEO cash compensation is risky, insiders have high levels of wealth vested in firm equity, the CEO is about to retire, or equity ownership is diffuse. In addition, division managers have incentives to use derivatives to reduce risk in measures of division performance. I investigate these determinants of derivatives use in a 1994 sample of 403 U.S. firms, including 215 derivatives users, and a 1993 sample of 807 U.S. firms, including 234 foreign exchange derivatives users. Consistent with the division manager hypothesis, I find that derivatives use is increasing in the number of lines of business and the number of overseas operations. Derivatives use is found to be decreasing in inside ownership, but is unaffected by the riskiness of CEO compensation, the level of insider wealth vested in equity, or CEO retirement. These results are consistent with the notion that shareholders write optimal contracts with top management to mitigate opportunistic use of derivatives. Similar to previous studies, I find that derivatives use is increasing in firm size and decreasing in regulation.
? I wish to thank Ray Ball, Dan Gode, Christoph Hinkelmann, S.P. Kothari, D.J. Nanda, Jerry Warner, and the participants of the Simon School Doctoral Seminar Series for their helpful comments and suggestions. Please note that this is a draft doucment reporting preliminary results of on-going research.
1. Introduction
There is a growing debate about how and why firms use derivative securities to alter firm risk. This debate is of particular relevance given the explosive growth in the use of derivatives by corporations. The preliminary results of the October 1995 Wharton/CIBC Wood Gundy Survey of Derivatives Usage indicate that derivatives use among respondents increased from 35% to 41% between 1994 and 1995. In addition, the results of the 1995 survey indicate that 91% of derivativesusers cite volatility in cash flows or earnings as the most important reason for using derivatives.
Much of the recent debate on derivatives use has focused on whether firms use these instruments for hedging purposes to maximize shareholder wealth or for opportunistic speculation. Finance theory indicates that hedging with derivatives can increase firm value by reducing expected taxes, expected costs of financial distress, underinvestment costs associated with investment opportunities in the presence of financial constraints, and agency costs. Several recent studies including Nance, Smith and Smithson (1993) have found mixed empirical results on these hedging-related determinants of derivatives use1. Hentschel and Kothari (1995) find that derivatives users and non-users exhibit few measurable differences in risk and conclude that this is consistent with firms using derivatives to hedge rather than speculate. However, these investigations fail to address a key factor that may influence the use of derivatives. In particular, past studies do not examine the risk characteristics of managers and their incentives to use derivatives. Therefore, while firms use derivatives for hedging purposes, this activity may benefit risk-averse managers at the expense of diversified shareholders.
It may be argued that managerial risk, compensation policies, ownership and organizational structure are not important ex post determinants of derivatives use if, ex ante, shareholders choose optimal management contracts that remove incentives for suboptimal reduction in firm risk. However, May (1995) finds empirical evidence consistent with the hypothesis that managerial motives influence a firm's risk reduction strategies. In particular, he finds that CEOs with more personal wealth vested in
1
Recent cross-sectional studies on derivatives use among non-financial firms include Nance, Smith and Smithson (1993), Francis and Stephan (1993), Mian (1994), and Geczy, Minton, and Schrand (1995).
equity tend to diversify through acquisitions, and that there is a negative relation between CEO human capital invested in the firm and the firm's equity variance and debt to equity ratio. These results appear to indicate that shareholder contracts with top management may not remove all personal incentives for managers to reduce firm risk.
Jensen and Meckling (1976) argue that agency conflicts may arise when managers hold an undiversified portfolio of equity and human capital invested in a single firm. Therefore, risk-averse managers may be motivated to reduce firm risk for personal gain at the expense of diversified shareholders. There are several cases where top managers would have the highest incentives to use derivatives to reduce risk. First, if CEO cash compensation contains a large risk component that is outside his control, then he may wish to use derivatives to hedge this risk. In particular, a compensation plan that includes a large bonus component based on accounting measures of performance may induce a manager to hedge accounting numbers. Second, insiders and top managers with high levels of personal wealth vested in firm equity may use derivatives to hedge firm value to reduce the risk of their undiversified investment portfolio. Third, if a CEO receives a post-retirement compensation package based on the firm's accounting performance prior to the CEO's retirement, then he may use derivatives to hedge accounting numbers to reduce the risk of this compensation. In theory, the preceding incentives can be eliminated if shareholders can write optimal compensation contracts. However, such contracts may be costly to negotiate and implement. Moreover, if equity ownership is diffuse, then there may be reduced incentives for individual shareholders to write and enforce optimal management contracts.
Jensen and Meckling also provide a framework for analyzing the organizational structure of a firm. In a decentralized firm, division managers may have incentives to reduce risk in measures of division performance if their compensation is based on division operating performance. It should be noted that division managers are further removed from shareholders and it may be difficult to write and implement contracts which limit opportunistic use of derivatives by division managers. This would imply that greater the number of divisions, the greater the likelihood that a firm will use derivatives.
The goal of this paper is to provide new empirical evidence on the relation between derivatives use and a firm's compensation policy, ownership structure and organizational structure. My empirical analysis focuses on two samples of firms that include derivatives users and non-users. The first sample is based on fiscal year 1994 financial statement information collected for 403 non-financial firms that trade on the New York (NYSE) and American (AMEX) stock exchanges. The analysis uses data collected from financial footnotes in annual reports to determine use of any type of derivative security at the end of fiscal 1994. Over 46% of the firms in this sample report using derivatives. CEO salary and bonus compensation, insider holdings of stock, line of business classifications, and other financial information are collected from electronic filings of proxy statements with the SEC and from the Standard and Poor's (S&P) Disclosure and Compustat databases.
The second sample is based on fiscal year 1993 financial statement information collected for 807 non-financial firms that trade on the NASDAQ, NYSE and AMEX stock exchanges. Corporate use of foreign exchange or currency derivatives at the end of fiscal 1993 is determined from information contained in annual report financial footnotes. Segmented data on foreign sales activities is obtained from the S&P Compustat Geographic Segment database. Additional financial data is gathered from the S&P Compustat database.
Using the two samples, I apply empirical tests of the relation between a firm's compensation policy, ownership and organizational structure and its use of derivatives. After controlling for firm size and other firm attributes, I find that use of derivatives is increasing in the number of divisions which is consistent with the hypothesis that division managers hedge division performance. In particular, derivatives use is found to be increasing in the number of lines of business and the number of overseas operations. The results of tests of risk characteristics of top management indicate the derivatives use is unaffected by the riskiness of CEO compensation, the level of insider wealth vested in equity, or CEO retirement. However, derivatives use is decreasing in the inside ownership of the firm, Although preliminary, these results for top management are consistent with the notion that shareholders write optimal management contracts to mitigate opportunistic use of derivatives. Consistent with previous
studies, the likelihood of derivatives use is found to be increasing in firm size and decreasing in regulation.
The paper is organized as follows: Section 2 provides a brief overview of current theory on corporate use of derivatives and summarizes previous empirical work. Section 3 outlines the hypotheses to be tested. Section 4 describes the sample and data. The results of the empirical tests are presented in section 5. Finally, section 7 concludes and proposes issues to be considered in future research.
2. Theories of Derivatives Use
This study attempts to extend previous empirical investigations on the determinants of corporate use of derivatives. As background, this section outlines current theory and summarizes the results of previous empirical work on the determinants of corporate use of derivatives. First, the extant theories for corporate derivatives are outlined. These determinants will be considered as cross-sectional control factors in tests of the marginal impact of managerial motives on derivatives use. In developing a framework for analyzing corporate use of derivative instruments, it is generally assumed that shareholders maximize expected wealth. Therefore, shareholders will try to optimally contract with management and bondholders. Therefore, managerial motives for derivatives use must be take into account shareholder attempts to write optimal contracts that maximize shareholder wealth.
Use of derivatives will not alter firm value if there are no taxes, no contracting costs, no costs to financial distress, no information costs and capital market imperfections do not exist. Therefore, use of derivative securities can alter firm value if they affect expected taxes; if contracting, financial distress or information costs exist; or if there are imperfections in the capital markets. The discussion below examines these factors and how the use of derivatives for hedging purposes may increase shareholder wealth.
Taxes are Convex in Net Income:
The presence of convexities in the tax code imply that a firm's profit function may be concave in taxable income. Smith and Stulz (1985) outline how hedging with derivatives can decrease a firm's expected tax liability if it faces a progressive corporate tax schedule over a range of taxable income. This relation is derived from the convexity of the corporate tax schedule at low levels of taxable income. Therefore, the use of derivatives to hedge variability in the firm's expected stream of taxable income can reduce expected taxes. Although the progressivity in the tax schedule applies over a small range of taxable income, generous provisions for tax loss carry forwards and investment tax credits reinforce convexities over a larger range of taxable income. The use of derivatives to hedge taxable income can increase the present value of these tax shields by reducing variability in taxable income.
Nance, Smith and Smithson (1993) find weak support for the hypothesis that firms use derivatives to hedge taxable income. However, Francis and Stephan (1993), Mian (1994) and Geczy, Minton, and Schrand (1995) find no support for this hypothesis 2.
Financial Distress Costs: If a firm enters financial distress then it will face costs of default on debt obligations, costs of filing for bankruptcy, and costs related to reorganization and liquidation. Given these costs, firms have incentives to reduce the probability of financial distress. Smith and Stulz argue that firms can reduce the likelihood of financial distress by hedging variability in earnings. Empirical studies typically use a firm's leverage ratio and interest coverage ratio to proxy for firms with higher probability of incurring the costs of financial distress. Using these proxies for financial distress, Nance et al, Francis and Stephan, Mian, and Geczy et al find mixed or no support for the financial distress rationale for derivatives use.
Debtholder Contracting Costs: Debtholders take into account the incentives of equityholders to opportunistically transfer wealth in the future from debtholders. Therefore, debtholders will discount the value of such actions in the issue
2
The mixed results on the tax reasons for derivatives use may be due to misspecification of a firm’s true marginal tax rate. Graham (1995) develops an algorithm to calculate simulated marginal tax rates based on current and expected taxable income, net operating loss carryforwards and investment tax credits. Application of such an algorithm may provide a better measure of the covexities in the tax schedule faced by a firm.
price of debt. To minimize these agency costs, equityholders can bond against these actions by instituting restrictive debt covenants or use convertible debt. Conflict with debtholders can also arise if fixed claims in a firm's capital structure induces equityholders to forgo positive NPV projects if the benefits of these projects primarily accrue to debtholders. This underinvestment problem can be mitigated if derivatives are used to reduce the likelihood of states in which wealth does not accrue to equityholders. Nance, Smith and Smithson posit that firms with more growth opportunities in their investment set, proxied for by high levels of R&D and a low book-to-market ratio, have a higher likelihood of using derivatives to hedge variance in firm value. In addition, they suggest that this underinvestment problem will be most severe in firms with high debt to equity ratios and, therefore, these firms are more likely to use derivatives to hedge firm value. Their empirical results are found to be inconclusive on these determinants of derivatives use.
Imperfection in Capital Markets: Froot, Scharfstein, and Stein (1993) draw on previous work to develop a model of optimal hedging to offset inefficient investment. They hypothesize that if external sources of funds are more costly to a firm than internally generated funds then the firm could benefit from using derivatives. In particular, firms can hedge cash flows to avoid a shortfall in funds that may require a costly visit to the capital markets. Geczy et al find that use of derivatives is positively related to measures of the firm's investment opportunity set proxied for by R&D expenditures and related to the firms coverage ratio. They claim that these findings are consistent with the Froot, Scharfstein and Stein underinvestment explanation for corporate use of derivatives.
Regulation: Firms in regulated industries provide top management with few opportunities for discretion in corporate investment and financing decisions. Smith and Watts (1992) show that regulation is a key determinant of a firm's corporate financial policy. Therefore, if regulated firms face tighter scrutiny and face lower contracting costs, then they are less likely to use derivatives to hedge firm risk.
Substitutes for Derivatives: Nance, Smith and Smithson note that corporate use of derivatives is also influenced by other financial and operating policies. For example, instead of using off balance sheet instruments like derivatives to hedge, a firm could manage financial price risk on its balance sheet by structuring its assets and liabilities to reduce its exposure to movements in financial prices. Nance et al propose that a firm could use a) convertible debt to help control agency conflicts with debtholders, b) use preferred stock instead of debt to provide greater flexibility in situations of financial distress, or c) have a lower dividend yield to assure bondholders that sufficient funds are available to pay fixed claims. They find empirical results consistent with these hypotheses.
Firm Size: A cursory analysis of the relation between firm size and derivatives may indicate that small firms have a greater demand for derivatives to hedge their more volatile cashflows and equity prices. Moreover, large firms may be able to find cost-effective substitutes for derivatives by hedging risk through line-of-business and geographic diversification. Prior to current empirical work on derivatives use, Smith and Stulz posited that small firms would be more likely to hedge than large firms if a) bankruptcy costs are less than proportional to firm size and b) the cost of using of derivatives for hedging purposes is proportional to firm size. Clearly these are valid arguments supporting the hypothesis the derivatives use should be more prevalent among small firms.
However, all cross-sectional empirical studies on the determinants of derivatives use find that derivatives use is increasing in firm size. There are various theories to support the scale effect of derivatives. In their paper on accounting standards, Watts and Zimmerman (1978) suggest that firms
subject to high levels of outside scrutiny from investors or regulators face higher political costs. They posit that these firms may take actions to reduce these costs by decreasing firm visibility. Francis and Stephan (1993) suggest that more visible firms will wish to reduce scrutiny of variable earnings by using derivatives to hedge accounting earnings. They suggest that firm size proxies for actual firm visibility and that use of derivatives should therefore be increasing in firm size.
Nance et al and Geczy et al suggest that size is related to significant information and transaction cost scale economies in establishing the expertise to use derivatives for hedging or speculation. This suggests that derivatives use should again be increasing in firm size.
The proper specification of firm size impacts any empirical tests of derivatives use. Functional form tests performed on the data samples in this paper indicate that a logarithmic transformation of firm size is the proper functional form for this proxy3.
2.1 Managerial Motives and Derivatives Use
Divisional Incentives: Derivatives use within firms is not restricted to central treasury operations. Firms with overseas operating divisions and autonomous lines of business may use derivatives at the division level for risk management. The compensation of division managers is both directly and indirectly related to measures of division operating performance. A division manager's compensation may be directly tied to accounting measures of division performance and promotion and firing decisions are indirectly impacted by division operating performance. A risk-averse division manager may have incentives to use derivatives to hedge accounting measures of operating performance. Given that division managers are farther removed from shareholders, it may be more costly to write and implement optimal compensation contracts which mitigate suboptimal use of derivatives. For example, an upstream overseas operating division that is evaluated as a cost center cause the manager to bear all of the risk of exchange rate
3
Both Francis and Stephan (1993) and Nace, Smith and Smithson (1993) use untransformed assets levels as proxies for firm size. It is unclear whether this functional form is the proper specification of size across firms.
fluctuations on prices of manufactured products. Therefore, even though it may not be optimal for the firm to hedge these exchange rate exposures, the performance measurement system induces the division manager to use derivatives to hedge exchange rate risk.
Therefore, division managers may have more opportunities to use derivatives for personal risk reduction and it would be predicted that derivatives use will be increasing in the number of operating divisions within a firm. Given the widespread use of foreign exchange derivatives that can be traced to exposures of overseas divisions, it is predicted that use of foreign exchange derivatives use will also be increasing in the number of overseas divisions.
It should be noted there is an opposite hypothesis regarding the relation between derivatives use and organizational structure. In particular, firms with many distinct lines of business and overseas divisions may be naturally hedged. Therefore, a greater number of distinct operating divisions is seen as a substitute for hedging with derivatives. This leads to the alternate prediction that firms with a larger number of lines-of-business and a larger number of overseas operating divisions would be less likely to use derivatives.
Undiversified Managerial Capital Vested in the Firm : Relative to shareholders, managers have a higher demand for firm-specific risk reduction because a manager's portfolio of human and investment capital is mostly tied to firm performance while shareholders' investments are well diversified. Therefore, managers who have more human capital invested in the firm have higher demand for risk reduction. May (1995) proxies for underdiversified CEO human capital by measuring the number of years a CEO has been employed at various levels of executive responsibility within the firm. May suggests that CEO managerial skills become more firm-specific as time spent with the firm increases. Given that hedging reduces human capital risk, firms whose CEOs have more years vested are more likely to use derivatives. CEO years with the firm also proxies for other factors that may affect the decision to use derivatives. For example, it may capture
managerial entrenchment, and more entrenched managers may have greater flexibility in using derivatives for managerial motives vs. stockholder wealth maximization4.
CEO and Insider Wealth Vested in Firm Equity: Jensen and Meckling (1976) argue that agency investment conflicts may result when managers hold an undiversified portfolio that consists primarily of firm equity. The problem occurs because managers cannot easily diversify their wealth invested in firm equity. In such a case, managers may undertake actions may reduce firm value if they reduce firm-specific return variance, while equityholders who have perfectly hedged away firm-specific risk through diversification are concerned only with maximizing firm value. Under this hypothesis we would expect that CEOs (and more generally insiders) with higher levels of personal wealth vested in firm equity would have a higher demand for firm-specific risk reduction. As May (1995) notes, the problem with testing this theory is that while one can measure insider wealth vested in firm equity, one cannot easily determine other elements of insiders' personal portfolios. One could use the ratio of CEO equity wealth to current cash compensation to proxy for wealth vested in firm equity. While this ratio is easy to measure, it will overstate CEO wealth vested in firm equity because it captures the accumulation of equity wealth but not the accumulation of potentially well-diversified cash compensation. Therefore, measures of insider wealth vested in equity may be biased measures of undiversified equity investments in the firm.
CEO Compensation Tied to Firm Performance: If CEO cash compensation contains a large risk component that is outside his control, then he may wish to use derivatives to hedge this risk. In particular, a compensation plan that includes a large bonus component based on accounting measures of performance may induce a manager to hedge accounting numbers5. In addition, if a CEO has contracted to receive a post retirement compensation package based on the firm's accounting performance prior to the CEO's retirement, then he may use derivatives to hedge accounting numbers to reduce the risk of this compensation. It should be noted that, in theory, the incentives for top management to engage in suboptimal use of derivatives can be
4
Years of actual service with the firm must be hand collected from various sources. I expect to collect this data to include in future versions of this paper.
eliminated if shareholders can write optimal compensation contracts. However, such contracts may be costly to negotiate and implement. In addition, diffuse ownership of equity may reduce the incentives for individual shareholders to write and enforce optimal management contracts.
Firms with greater growth opportunities have higher levels of performance-rated compensation. Smith and Watts (1992) document the relationship between a firm's investment opportunity set and its financing and compensation policies. Given that CEO compensation is endogenously determined, then the impact of performance-based compensation on the managerial incentives to use of derivatives may be difficult to measure.
3. Empirical Hypotheses
The preceding discussion on the determinants of derivatives use provides a framework for empirical tests of the relation between corporate use of derivatives and a firm's compensation policy, ownership and organizational structure. This section presents the main empirical hypotheses to be tested.
3.1 CEO Compensation
As outlined in section 2, if CEO cash compensation contains a large risk component that is outside the CEO's control, then he may wish to use derivatives to hedge this risk. In particular, a compensation plan that includes a large bonus component based on accounting measures of performance may induce a manager to hedge accounting numbers. In addition, if a CEO has contracted to receive a post retirement compensation package based on the firm's accounting performance prior to his retirement, then he may use derivatives to hedge accounting numbers to reduce the risk of this compensation. These theories lead to the following empirical hypotheses:
5
See Healy (1985) for a discussion of CEO bonus plans.
H1: Other things equal, the higher the ratio of bonus compensation to total CEO cash compensation, the greater the likelihood that the firm will use derivatives. H2: Other things equal, the closer a CEO is to retirement, the greater the likelihood that the firm will use derivatives.
3.2 Insider Wealth Vested in Firm Equity
Insiders and top managers with high levels of personal wealth vested in firm equity may use derivatives to hedge firm value to reduce the risk of their undiversified investment portfolios. This leads to hypothesis 3:
H3: Other things equal, the higher the level of wealth insiders have vested infirm equity, the greater the likelihood that the firm will use derivatives.
An underlying assumption of hypotheses H1, H2, and H3 is that shareholders cannot write and implement optimal contracts with top management to mitigate suboptimal use of derivatives. Rejection of these hypotheses would be consistent with the theory that shareholders can effectively implement contracts with top management that remove opportunities for value-reducing investments in derivatives. Therefore, these contracts have aligned managerial incentives with the interests of shareholders and any remaining opportunities to hedge managerial risk with derivatives are too costly. If equity ownership in the firm is diffuse, then there may be weak incentives for individual shareholders to write and implement optimal managerial contracts. As a proxy for ownership concentration, I use the percent of total shares outstanding held by insiders6. This leads to the fourth hypothesis:
H4: Other things equal, the greater the concentration of inside ownership in the firm's equity, the lower the likelihood that the firm will use derivatives.
6
It could also be argued that concentrated inside ownership would give insiders greater latitude to take actions at the expense of outside shareholders. Therefore, this proxy may have uncertain implications for derivatives use.
3.3 Divisional Incentives
In a decentralized firm, division managers have may incentives to reduce risk in measures of division performance if their compensation is based on these division operating performance. Given that division managers are further removed from shareholders, it may be difficult for shareholders to write and implement contracts which limit opportunistic use of derivatives by division managers. This leads to the following predictions:
H5: Holding firm size constant, the larger the number of lines of business within the firm, the greater the likelihood that the firm will use derivatives. H6: Holding firm size and relative level of foreign sales constant, the larger the number of overseas operations, the greater the likelihood that the firm will use derivatives.
As outlined in section 2, the existence of numerous distinct lines of business and overseas operations may provide a firm with a natural hedge. This result would suggest that diversified firms are less likely to use derivatives. Therefore, failure to reject hypotheses H5 and H6 should be considered strong support for the division manager theory of derivatives use.
4. Data and Sample Selection
The empirical analysis focuses on two samples of firms that include derivatives users and non-users. The first sample is based on fiscal year 1994 financial statement information collected for 403 non-financial firms that trade on the New York (NYSE) and American (AMEX) stock exchanges. The analysis uses data collected from financial footnotes in annual reports to determine use of any type of derivative security at the end of fiscal 1994. Over 46% of the firms in this sample report using derivatives. CEO salary and bonus compensation, insider holdings of stock, line of business classifications, and other financial information are collected from electronic filings of proxy statements with the SEC and from the Standard and Poor's (S&P) Disclosure and Compustat databases. The sample selection procedure for the first sample is outlined in Table 1A.
Recent changes to disclosure requirements for compensation of top executives have provided a valuable source of information about CEO pay. CEO salary and bonus compensation data for fiscal years 1992, 1993 and 1994 were hand collected for the 403 firms in the sample from electronic filings of proxy statements with the SEC.
The second sample is based on data collected for a paper investigating the determinants of foreign exchange derivatives (FXD) use by U.S. corporations 7. The data covers fiscal year 1993 financial statement information collected for 807 non-financial firms that trade on the NASDAQ, NYSE and AMEX stock exchanges. Corporate use of foreign exchange or currency derivatives at the end of fiscal 1993 is determined from information contained in annual report financial footnotes. Segmented data on foreign sales activities is obtained from the S&P Compustat Geographic Segment database. Additional financial data is gathered from the S&P Compustat database. The sample selection procedure for the FXD sample is presented in Table 1A.
The definitions of and construction procedure for the proxy variables used in the empirical tests are outlined in Table 2.
5. Empirical Evidence
5.1 Univariate tests
Univariate tests of differences means of the variables hypothesized to impact a firm's use of derivatives are presented in Table 3. Definitions of the variables can be found in Table 2. The sample is comprised of 403 firm-year observations for the various variables including 215 observations for firms reporting the use of any type of derivative at the end of fiscal 1994. Of the "managerial motive" variables of interest, all of are differences in means between derivatives users and non-users have predicted sign and only the difference in the means of A_RETIRE is statistically insignificant.
The univariate results indicate that derivatives users have significantly higher levels of CEO cash compensation risk, higher levels of insider wealth vested in firm equity, a larger number of lines-of business and lower levels of insider ownership. The results of the univariate tests for the tax, financial distress, derivatives substitutes and other control variables are summarized in Table 1.
5.2 Multivariate Logit Regression Results
To control for interactions among the hypothesize determinants of derivatives use, we estimate logistic regressions using a categorical dependent variable describing derivatives use. Table 5 outlines the results of various specifications of the logistic model. Model 1 is estimated as a benchmark case that does not include CEO and insider determinants of derivatives use. It is found that only S_DIVYLD, O_SIZE and A_DIVISION are significant determinants of derivatives use. It should be noted that S_DIVYLD has the opposite sign to that predicted by Nance, Smith and Smithson. An examination of the correlation between S_DIVYLD and O_REGLTL indicates that they are highly correlated ( see Table 4). Therefore, model 3 is estimated without S_DIVYLD and it is found that O_REGUL comes in strongly significant and of the predicted sign. It appears that S_DIVYLD may just be a proxy for regulated firms in this sample. A_DIVISION is found to be of the predicted sign and is consistent with the division manager hypothesis for derivatives use.
In model 3, all hypothesized determinants of derivatives use are included. Again, S_DIVYLD, O_SIZE and A_DIVISION are found to be significant and have the same sign as models 1 and 2. In addition, A_EQUITYINSIDE is found to be significant and of the predicted sign. This result is consistent with the diffuse ownership hypothesis. It should be noted that A_RETIRE and A_COMPRISK are found to be insignificant determinants of derivatives use8. This result is consistent
7 8
See Appendix A for a summary of the results of Wysocki (1995). Alternate specifications of A_COMPRISK were also tested. The average ration of bonus to total compensation over the three year period of 1992-94 and the ratio of the maximum bonus over the past three years to current total cash compensation in 1994 were tested. These variables also proved to be insignificant determinants of derivatives use.
with the notion that shareholders write and implement optimal contracts that remove incentives for top management to engage in suboptimal use of derivatives. Model 4 is presented as a specification test excluding the A-RETIRE dummy. The results are remain unchanged as compared to model 3.
To obtain a better intuition on the impact of the determinants an OLS regression of the binary dependent variable on the explanatory variables is estimated. The coefficients from this linear probability model are unbiased and can be directly interpreted as marginal probabilities. The results of this regression are presented in Table 6. The p-values of the OLS regression are presented for comparative purposes only as they are biased estimates of the variance of the coefficient estimates. The linear probability model is found to explain 33% of variation in derivatives use in this sample of firms.
Table 6 and 7 present the results of the multivariate logit regression of foreign exchange derivatives use on various explanatory variables. After controlling for firm size, relative levels of foreign sales activity, foreign tax liability and other factors, it is found that the number of countries/regions in which a firm has foreign sales operations is found to positively and significantly related to foreign exchange derivatives use. This is consistent with the division manager hypothesis for derivatives use. A complete discussion of the other variables in this regression can be found in Appendix A.
6. Conclusions and Future Research
In this paper, I include corporate compensation policies, ownership structure and organizational structure as possible determinants of derivatives use. First, I examine the relation between compensation of top management and the use of derivatives by managers to reduce personal risk exposure. The results of the multivariate tests indicate that firms with high risk in CEO cash compensation, high levels of insider wealth vested in firm equity, or high probability of CEO retirement are not more likely to use derivatives. However, the likelihood of derivatives use is found to be decreasing in inside ownership. Although these results are preliminary, they are consistent with the view that shareholders account for managerial risk aversion when writing optimal compensation contracts for top executives. Second, I
investigate whether firms with a larger number of divisions are more likely to hedge. The empirical results are consistent with the hypothesis that division managers that are far removed from shareholders have incentives to use derivatives to reduce risk in measures of division performance. I find that derivatives use is increasing in both the number of lines of business and the number of overseas operations of a firm.
Consistent with previous studies, I find that the likelihood of derivatives use is found to be increasing in firm size and decreasing in regulation. In addition, I find that derivatives users and non-users are not different with respect to proxies of the firm's tax position.
Although this paper provides new empirical evidence on link between managerial incentives and derivatives use, there are a number of issues that should be explored. In particular, I hope to implement better proxies of managerial compensation including CEO holdings of firm equity, CEO compensation based on stock options, and measures of CEO human capital invested in the firm. In addition, I plan to develop a better model and gather additional data on a firm's divisional characteristics division manager compensation and division level use of derivatives. I hope to draw on the work of Bushman, Indejejikian and Smith (1994 and 1995) and Christie, Joye and Watts (1993) and Baiman, Larcker and Rajan 9195) to develop a more complete model of why derivatives use is increasing in the number of divisions within the firm.
Data Sources:
•
Standard and Poor's Disclosure Database (SEC Edition)- December 1995 release (Financial footnotes and fulltext of SEC filings used to determine derivatives holdings of firms with December 31, 1994 fiscal year end. Also used to determine CEO age, and insider stock holdings.
• •
Standard and Poor's CONTUSTAT Database - Annual Industrial File - 1994 Edition SEC EDGAR Database - Proxy statements filings. Used to determine salary and bonus compensation of Chief Executive Officers of firms in the sample.
References Baiman, S., D.F. Larcker, and M.V. Rajan, 1995, Organizational Design of Business Units, Journal of Accounting Research 33, 205-229. Bushman, R., R. Indejejikian, and A. Smith, 1994, Compensation contracts in hierarchies: determinants of incentive pat for business unit managers", University of Chicago working paper. Bushman, R., R. Indejejikian, and A. Smith, 1995, Aggregate performance measures in business unit manager compensation: the role of intra-firm interdependencies", University of Chicago working paper. Christie, A., M. Joye and R. Watts, 1993, Decentralization of the firm: theory and evidence", University of Rochester working paper. DeMarzo, P.M., and D. Duffie, 1991, Corporate Financial Hedging with Proprietary Information, Journal of Economic Theory 53, 261-286. DeMarzo, P.M., and D. Duffie, 1995, Corporate Incentives for Hedging and Hedge Accounting, Review of Financial Studies, Vol. 8, No. 3, 743 -77 1. Francis, J., and J. Stephan, 1993, Characteristics of Hedging Firms: An Empirical Investigation, in Advanced Strategies in Financial Risk Management, R.J. Swartz and CW. Smith eds., 615 -65 8. Froot, K.A., D.S. Scharfstein, and J.C. Stein, 1993, Risk Management: Coordinating Corporate Investment and Financing Policies, Journal of Finance, Vol XLVIII, No. 5, 1629-1658. Geczy, C., B.A. Minton, and C. Schrand, 1995, Why Firms Use Derivatives: Distinguishing Among Existing Theories, Working Paper, Fisher College of Business, The Ohio State University. Graham, JR., 1995, Debt and the marginal tax rate, Working Paper, David Eccles School of Business, The University of Utah. Healy, P.M., 1985, The effect of bonus schemes on accounting decisions, Journal of Accounting and Economics 7, 85-107. Hentschel, L., and S.P. Kothari, 1995, Life Insurance or Lottery: Are Corporations Managing or Taking Risks with Derivatives?, Working Paper, William E. Simon School of Business Administration, University of Rochester.
Jensen, M.C., and W.H. Meckling, 1976, Theory of the firm: Managerial behavior, agency costs and ownership structure, Journal of Financial Economics 3, 305-360. May, D.O., 1995, Do Managerial Motives Influence Firm Risk Reduction Strategies, Journal of Finance, Vol. L, No. 4, 1291-1308. Mian, S.L., 1994, Evidence on the determinants of corporate hedging policy, Working Paper, Emory Business School, Emory University. Nance, DR., CW. Smith, and CW. Smithson, 1993, On the Determinants of Corporate Hedging, Journal of Finance, Vol. XLVIII, No. 1, 267-284. Smith, C.W. and R.M. Stulz, 1985, The Determinants of Firms' Hedging Policies, Journal of Financial and Quantitative Analysis, Vol. 20, No. 4, 391405, Smith, C.W., and R.L. Watts, 1992, The investment opportunity set and corporate financing, dividend and compensation policy, Journal of Financial Economics 3 2, 263 -292. Smithson, CW., 1996, Managing Financial Risk: 1996 Yearbook (CIBC/Wood Gundy Publications). Stulz, R.M., 1984, Optimal Hedging Policies, Journal of Financial and Quantitative Analysis, Vol. 19, 127-140. Warfield, T.D., J.J. Wild, and K.L. Wild, 1995, Managerial ownership, accounting choices and informativeness of earnings, Journal of Accounting and Economics 20, 61-9 1. Watts, R.L. and J.L. Zimmerman, 1978, Towards a positive theory of the determination of accounting standards, Accounting Review 53, 112-134. Wysocki, P.D., 1995, Determinants of Foreign Exchange Derivatives Use by U.S. Corporations: An Empirical Investigation, Simon School of Business Summer Working Paper, University of Rochester.
Appendix A: Summary of "Determinants of Foreign Exchange Derivatives Use by U.S. Corporations: An Empirical Investigation" Peter Wysocki, 1995 Working Paper
This paper extends the previous literature by focusing on the specific determinants of firms’ foreign exchange hedging behavior. I apply current theory on foreign exchange risk management to develop a set of hypotheses on the determinants of firms' use of foreign exchange derivatives (FXDs). These hypotheses are tested on a large sample of firms using newly available data. In addition, I make novel use of geographic segment information to obtain a detailed measure of the extent and character of foreign operations of U.S. firms. The results of my hypothesis tests indicate that in 1993 U.S. firms with FXD holdings were typically large; had high levels of transaction and translation exposures as measured by relative levels of foreign sales and taxes; tended to operate in numerous countries; and, were likely to have foreign sales revenue from Japan. Inconclusive results are found for the hypothesis that FXD hedgers are more likely to have indirect economic exposures as measured by industry-level foreign import competition. We believe that these results help provide a more complete understanding of the general determinants of corporate hedging as well as the specific determinants of FXD hedging behavior of firms.
The following hypotheses were tested in the paper:
Transaction Exposure: Accounting earnings of U.S. firms can be affected by foreign exchange fluctuations by changing the expected results of transactions in non-local currencies. This is known as transaction exposure. A U.S. firm that derives a substantial fraction of its sales revenues from foreign sources would likely be impacted by this type of exposure. This argument leads to the hypothesis:
FXD HI: Foreign exchange hedging firms derive a larger proportion of sales revenues from foreign sources than nonhedging firms.
Translation Exposure: Translation gains and losses arise when a company translates financial statements from its overseas operations from a foreign currency to the U.S. dollar. As specified under SFAS No.52, an overseas subsidiary that designates the U.S. dollar as its functional currency must include translation gains and losses in the determination of income. Therefore, firms that have high levels of overseas assets are more likely to have a accounting translation exposure. Under SFAS 14, firms are required to report geographically segmented information on significant foreign sales activity and foreign assets. Data on foreign assets was unavailable for this study. Consequently, foreign taxes, a common proxy for foreign assets, are used as an indicator of possible translation exposures9.
FXD H2: Foreign exchange hedging firms have a higher fraction of foreign tax incidence than nonhedging firms.
Economic Exposure: Economic exposures impact a firm's competitive position and are caused by exchange rate induced changes in input and output prices. Many of these exposures are proxied for by transaction and translation exposures. However, firms without direct foreign activities, sales or assets can also be impacted by changes in exchange rates. These exposures can result from a number of factors including foreign competitors that benefit from exchange rate changes, customers whose activities are impacted by foreign exchange factors, or suppliers that use foreign-sourced inputs. Firms that are in domestic markets with meaningful foreign import competition can be adversely affected by an appreciation in the U.S. dollar. It is posited that foreign competitors with meaningful U.S. market shares would be in a strong position to benefit from an appreciation in the U.S. dollar. Also, larger firms are more likely to have a wide range of suppliers and customers, increasing the likelihood that the firms costs or revenues may be impacted by foreign exchange fluctuations.
FXD H3: Foreign exchange hedging firms face higher levels on industry-level foreign import competition. FXD H4: Foreign exchange hedging firms are larger than nonhedging firms.
9
Lee and Kwok (1998) and Teerathorn, Alcerraca-Joaquin and Seigel (1992) use foreign tax ratios to proxy for the level of foreign assets. Lee and Kwok examine a sample of firms and find significant correlations between foreign
Information and Transaction Scale Economies: A firm must incur substantial fixed costs to either contract with outside specialists or employ skilled and informed managers to manage a foreign exchange hedging program. Nance, Smith, and Smithson note that a hedging program exhibits informational and transaction scale economies. Clearly, larger firms can spread these fixed costs over a larger set of foreign activities and exposures.
FXD H5: Foreign Exchange hedging firms are larger than nonhedging firms.
Developed Foreign Exchange Markets: The existence of liquid and well-developed markets for trading foreign exchange contracts will reduce that cost of implementing a foreign exchange hedging program. Organized derivatives exchanges provide unencumbered trading opportunities, high volumes, and decreased risk of counterparty default. If such markets do not exist, then a firm may have to pay a substantial premium to purchase foreign exchange derivatives contracts in the over-the-counter market. Liquid public U.S. dollar foreign exchange markets exist for major currencies such as the Gennan mark, French franc, British pound, Australian and Canadian dollars, and the Japanese yen. On the other hand, liquid markets do not exist for currencies of developing economies. Therefore, FXD hedging may not be a cost-effective option for hedging exposures in these currencies.
FXD H6: Foreign exchange hedging firms are more likely to derive sales from Europe, Japan, Canada and Australia. FXD H7: Foreign exchange hedging firms have equal probability of deriving sales from South America, Africa, and Mexico than nonhedging firms.
asset ratios and foreign tax ratios.
Table 1A: Sample Selection Procedures for Main Sample
Criteria Population of firms on Disclosure Database Firm’s shares trade on NYSE or AMEX AND Firm is incorporated in the U.S. AND Primary SIC outside range 6000-6999 AND Financial data in U.S. dollars AND Financial footnote information provided AND December 31 fiscal year-end AND Financial footnotes with unambiguous classification of derivatives use Total Number of Firms in Disclosure Sample Firms without matching SEC proxy statement data or financial data on Compustat Firms remaining in Main Sample Derivatives user subsample Firms contained in Main Sample Firms reporting the following in financial footnotes: • hedge(s), hedged, hedger, hedging • swap(s), swaption(s) • forward • futures • cap, collar, floor • option(s) • foreign exchange contract • currency contract derivative Firms in Main Sample reporting using derivatives at fiscal year end (adjustment for misclassified firms in keyword search) Number of firm observations
2452 2375 2047 1907 1724 1045 980 980 577
403 403 209
188
Table 1B: Sample Selection Procedure for Foreign Exchange Derivatives Sample
Criteria Firm’s shares trade on NYSE, AMEX or NASDAQ stock exchange AND fiscal 1993 financial footnote information contained in Disclosure database AND firm is active in fiscal 1993 reporting period AND firm is incorporated in the US AND primary SIC outside the range 6000-6999 Number of observations not on Compustat “ALL FIRMS” FXD Hedger Subsample Firms with “ALL FIRMS” reporting keywords in full text of Disclosure database: • currency(s) • OR exchange rate(s) • OR foreign exchange AND • derivative(s) • OR hedge(s) OR hedged • OR forward • OR futures • OR option(s) • OR risk management • OR swap(tions) Number of firms without FXD holdings “FXD HEDGER” (MAIN SAMPLE) Subsample A (Include only firms with SIC in range 0-3999) Hedgers in Subsample A Subsample B (Include only firms with (a) SIC in range 0-3999 and (b) report data for foreign and domestic taxes Hedgers in Subsample B Number of firm observations 6613 2334 2314 2163 1902 1902 143 1759
417
417 106 311 1004 248 807 234
Table 2: Definition of Variables
This table summarizes all variables used in the analysis including a detailed description of the method of calculation
Variable Name Tax Variables T_NOL T_ITC Bankruptcy Variables B_DEBTEQ B_COVERAGE Derivatives Substitutes S_CONVERT S_PREFSHR S_DIVYLD Organizational Form O_REGUL O_SIZE Description
Net operating losses applied as a reduction of taxable income in 1994 (Compustat item 52) Dummy variable equal to 1 if firm has investment tax credits on its balance sheet (derived from Compustat item 208) Debt-equity ratio: ratio of 1994 book value of LT debt (Compustat items 35 and 9) to firm size (see below) Interest coverage ratio: 1994 ratio of pretax income (item 170) plus interest expense (item 5) to interest expense Ratio of total convertible debt (Compustat item 79) to firm size Ratio of book-value of total preferred stack (Compustat item 130) to firm size Dividend yield: ratio of cash dividend per share (Compustat item 26) to closing price per share at end of fiscal year 1994 (Compustat item 199) Dummy variable equal to 1 if firm’s primary 2-digit SIC is 49 Market value of the firm at end of fiscal 1994. Defined as the sum of mkt value of equity (Compustat item 199 times item 25) plus book value of LT debt (item numbers 9 and 34) plus book value of preferred stock (item number 130). O_SIZE is constructed using the logarithm of the market value of the firm. Ratio of book value to market value of the firm. Book value = Total assets minus total liabilities less preferred stock (Compustat items 6-181-130). Market value = firm size (Actual CEO bonus in 1994) / (Actual CEO salary plus bonus) Number of held shares held by insiders / Total number of shares outstanding Total number of different 2 digit SIC classifications for the firm Number of shares held by insiders times market value of shares at fiscal year end (Compustat item 24) Dummy variable equal to 1if CEO is age 64 or 65 at fiscal year end 1994
Agency Costs A_BKMKT A_COMPRISK A_EQUITYINSIDE A_DIVISION A_WEATHEQUITY A_RETIRE FXD Sample Multi-country Sales Foreign Imports Country-specific Sales Foreign Tax Ratio Foreign Sales Ratio
Number of countries in which the firm reports segmented sales operations U.S. Domestic market share held foreign companies in the firm’s primary 4 digit SIC in 1993. A dummy variable equal to 1 if a firm has foreign sales operations in the identified country/region in 1993. Ratio of foreign taxes paid to total taxes paid in fiscal 1993 Ratio of total foreign sales/revenues to total firm sales/revenues in 1993
Table 3: Summary Statistics for Characteristics of Derivative Users and Non-Users
Sample of 403 firms including 215 firms reporting the use of any type of derivative at the end of fiscal yr 1994. Data measure at fiscal year-ends. Derivative Users (N=215) Variable Mean Tax Variables T_NOL ($MM) T_ITC (dummy) 36.857 0.404 9.507 0.046 7.731 0.365 6.843 0.033 3.063 0.861 0.002 0.389 Std. Dev. Mean Std. Dev. Derivative Nonusers (N=188) Test of Difference in Means (Users – Nonusers) t-stat p-value
Bankruptcy Variables B_DEBTEQ B_COVERAGE Derivatives Substitutes S_CONVERT S_PREFSHR S_DIVYLD Organizational Form O_REGUL (dummy) O_SIZE (log($MM)) 0.078 7.632 0.029 0.165 0.136 5.681 0.021 0.118 -1.963 11.848 0.050 0.000 0.017 0.011 0.020 0.006 0.004 0.002 0.021 0.018 0.019 0.005 0.003 0.002 -0.757 -1.972 0.187 0.449 0.049 0.852 0.852 11.121 0.859 7.153 5.673 19.073 0.620 5.243 -5.613 -1.111 0.000 0.267
Agency Costs A_BKMKT A_COMPRISK A_EQUITYINSIDE A_DIVISION A_WEALTHEQUITY (log $MM) A_RETIRE (dummy) 0.335 0.365 0.094 2.609 3.734 0.052 0.033 0.020 0.019 0.118 0.179 0.021 0.448 0.245 0.189 1.935 2.436 0.051 0.023 0.014 0.014 0.085 0.130 0.015 -3.458 6.089 -5.073 5.706 7.243 0.037 0.001 0.000 0.000 0.000 0.000 0.971
TABLE 4: Pearson Correlation Coefficients for Variables
Sample of 403 firms including 215 firms reporting the use of any type of derivative security at the end of fiscal year 1994. Data measured at fiscal year-ends.
A_EQUITYINSIDE
A_COMPRISK
A_DIVISION
A_WEALTH
S_CONVERT
A_BKMKT
B_DEBTEQ
DER_USER
A_RETIRE
S_DIVYLD
B_COVER
O_SIZE
S_PREF
DER_USER A_BKMKT A_COMPRISK A_EQUITYINSIDE A_RETIRE A_WEALTH B_COVER B_DEBTEQ A_DIVISION O_REG O_SIZE S_CONVERT S_DIVYLD S_PREF T_ITC T_NOL
1.000 -0.152 0.288 -0.254 -.005 0.325 0.006 -0.188 0.267 -0.112 0.490 -0.043 -0.017 -0.078 0.044 0.144 1.000 -.0168 0.120 -0.001 -0.228 0.013 0.217 -0.072 0.021 -0.363 -0.085 -0.007 -0.131 0.026 -0.055 1.000 -0.132 0.007 0.368 0.041 -0.127 0.179 -0.184 0.430 -0.013 -0.043 -0.050 0.007 0.092 1.000 -0.006 0.272 -0.012 0.095 -0.124 -0.193 -0.419 0.076 -0.269 -0.110 0.042 -0.054 1.000 0.006 -0.016 0.025 -0.045 0.027 -0.002 -0.002 0.167 0.015 0.044 0.015 1.000 0.071 -0.230 0.187 -0.263 0.521 -0.043 -0.123 -0.123 -0.030 0.150 =5% signif 1.000 -0.019 -0.017 -0.025 0.019 -0.022 -0.011 -0.027 0.024 -0.017 1.000 -0.097 -0.055 -0.321 -0.002 -0.097 -0.054 -0.029 -0.029 1.000 -0.104 0.318 -0.081 0.129 -0.016 0.032 0.087 1.000 0.128 -0.076 0.815 0.294 0.008 -0.078 1.000 -.0114 0.313 0.064 -0.025 0.156 1.000 -0.139 0.041 0.010 -0.002 1.000 0.162 0.048 -0.064 1.000 -0.046 -0.016 1.000 -0.030 1.000
=10% signif
=1% signif
T_NOL
O-REG
T_ITC
Table 5: Estimates of logistic models relating probability of derivatives use to firm-specific characteristics (Main sample of derivatives users and non-users)
Sample of 403 firms including 215 firms reporting the use of any type of derivative at the end of fiscal year 1994. Independent Variable Model 1 Model 2 Model 3 Model 4 Model 5 (OLS) R = 0.33 -0.199 (0.961)
Intercept
-4.790 (<0.001)
-4.561 (<0.001)
-4.059 (<0.001)
-4.316 (<0.001)
Tax Variables T_NOL T_ITC 0.003 (0.194) 0.397 (0.112) 0.003 (0.176) 0.340 (0.167) 0.003 (0.261) 0.492 (0.062) 0.003 (0.259) 0.498 (0.059) 0.0002 (0.314) 0.054 (0.217)
Bankruptcy Variables B_DEBTEQ B_COVERAGE -0.050 (0.083) 0.000001 (0.734) -0.050 (0.083) 0.000001 (0.748) -0.048 (0.097) 0.000002 (0.620) -0.050 (0.079) 0.000002 (0.686) -0.0007 (0.515) 0.0000001 (0.887)
Derivatives Substitutes S_CONVERT S_PREFSHR S_DIVYLD 0.054 (0.976) -3.689 (0.296) -15.625 (0.038) 0.385 (0.828) -4.076 (0.255) -0.190 (0.919) -4.542 (0.211) -18.762 (0.023) -0.209 (0.911) -4.767 (0.191) -19.607 (0.017) 0.097 (0.764) -0.805 (0.195) -2.787 (0.021)
Organizational Form O_REGUL O_SIZE -0.486 (0.338) 0.681 (<0.001) -1.153 (0.004) 0.632 (<0.001) -0.419 (0.451) 0.554 (<0.001) -0.536 (0.322) 0.631 (<0.001) -0.055 (0.556) 0.091 (<0.001)
Agency Costs A_BKMKT A_COMPRISK A_EQUITYINSIDE A_WEALTHEQUITY A_DIVISION A-RETIRE 0.212 (0.037) 0.186 (0.064) 0.521 (0.233) 0.419 (0.338) 0.406 (0.373) 0.275 (0.681) -2.095 (0.034) 0.098 (0.392) 0.265 (0.014) 0.308 (0.618) 0.415 (0.362) 0.311 (0.641) -1.528 (0.034) 0.069 (0.306) -0.307 (0.102) -0.440 (0.005) 0.028 (0.128) 0.043 (0.015)
0.266 (0.014)
Table 6: Pearson Correlation Coefficients Between Proxy Variables for 807 Firms (1993 Data)
Sample of 807 firms including 234 firms reporting foreign exchange derivatives holdings at the end of fiscal year 1993. For. Tax Ration For. Tax Ratio Africa Sales Austral. Sales Canada Sales Europe Sales Foreign Imports Japan Sales Mexico Sales MultiCountry Sales South Am. Sales For. Sales Ratio Africa Sales Austral. Sales Canada Sales Europe Sales For. Import Japan Sales Mex. Sales Divers. For. Sales South Amer. Sales For. Sales Ratio Firm Size
1
0.269 0.255 0.270 0.415 0.041 0.167 0.212 0.507
1 0.215 0.187 0.235 -0.027 0.059 0.156 0.457 1 0.339 0.347 -0.028 0.254 0.309 0.537 1 0.394 -0.097 0.151 0.360 0.562 1 0.029 0.269 0.274 0.775 1 0.015 -0.060 -0.026 1 0.173 0.389 1 0.504 1
0.324 0.666
0.402 0.309
0.378 0.323
0.295 0.342
0.331 0.592
-0.061 0.060
0.185 0.322
0.452 0.290
0.144 0.127
1 0.400 1
Firm Size 0.175 0.197 0.093 0.111 0.107 -0.001 0.023 0.161 0.071 0.188 0.198 1
Table 7: Estimates of logistic model relation probability of foreign exchange derivatives use to firm-specific characteristics
Notes: N=Number of firms in sample, D=Number of foreign exchange derivatives users is sample, FXD=foreign exchange derivatives Sample Statistics in table are: Mean / Standard Error / Probability Value Theory Predicted Sign Subsample B (N=807, D=234) Transaction and Economic Exposure 1.378 + 0.751 0.067 • Information & Transaction 0.336 Scale Economies + 0.061 • Political Costs 0.000 • Increased Economic Exposure 0.325 Division Manager Hedging Risk in + 0.098 Measures of Division Performance 0.001 0.375 Developed FXD Market + 0.360 0.297 0.751 Developed FXD Market + 0.358 0.036 -0.609 Developed FXD Market + 0.255 0.017 0.197 Developed FXD Market + 0.318 0.536 0.061 Undeveloped FXD Market 0 0.295 0.838 -0.203 Undeveloped FXD Market 0 0.317 0.522 0.274 Undeveloped FXD Market 0 0.472 0.561 0.0070 Hedging Indirect Economic + 0.0067 Exposure 0.296 0.891 Hedging Translation Exposure + 0.389 0.022
Proxy Variable Foreign Sales Ratio Firm Size
# of Regions with Foreign Sales
Sales in Europe
Sales in Japan
Sales in Canada Sales in Australia Sales in Mexico
Sales in South America Sales in Africa
Degree of Foreign Import Penetration Foreign Tax Ratio
Table 8: Results of Compensation Policy Regressions on Firm Size, Dividend Yield, Regulation, Book to Market, and Derivative User Variables
Sample of 403 firms including 188 firms reporting the use of derivatives of fiscal year-end 1994 (All data in 1994)
Dependent Variable
Regression R2
Intercept
O_SIZE
S_DIVYLD
O_REG
A_BKMKT
DER_USER
A_WEALTHEQUITY
0.393
-0.3528 (0.327)
0.676 (<0.001)
-11.962 (0.004)
-1.483 (<0.001)
-0.125 (0.589)
0.0354 (0.841)
A_COMPRISK
0.253
-0.249 (0.5352)
0.0519 (<0.001)
-0.596 (0.234)
-0.149 (<0.001)
0.015 (0.594)
0.009 (0.665)
A_EQUITYINSIDE
0.209
0.393 (<0.001)
-0.031 (<0.001)
-0.958 (0.043)
-0.056 (0.108)
-0.005 (0.846)
-0.043 (0.033)
doc_999049877.pdf