Growth opportunities and corporate debt policy: the case of the U.S. defense industry

Growth opportunities and corporate debt policy: the case of the U.S. defense industry

Journal of Financial Economics 64 (2002) 35–59 Growth opportunities and corporate debt policy: the case of the U.S. defense industry$ Vidhan K. Goyal...

204KB Sizes 0 Downloads 8 Views

Journal of Financial Economics 64 (2002) 35–59

Growth opportunities and corporate debt policy: the case of the U.S. defense industry$ Vidhan K. Goyala, Kenneth Lehnb,*, Stanko Racicb b

a Hong Kong University of Science and Technology, Hong Kong Katz School of Business, University of Pittsburgh, Pittsburgh, PA 15260, USA

Received 1 October 1998; received in revised form 4 May 2001

Abstract The U.S. defense industry provides a natural experiment for examining how changes in growth opportunities affect the level and structure of corporate debt. The growth opportunities of defense firms, compared with other firms, increased substantially during the Reagan defense buildup of the early 1980s but then declined significantly with the end of the cold war and associated defense budget cuts in the late 1980s and early 1990s. We examine how the level and structure of corporate debt changed for a sample of 61 defense firms and a benchmark sample of 61 manufacturing firms during 1980–95, a period spanning the changes in growth opportunities. The debt levels of weapons manufacturers, which were most affected by the changes in growth opportunities, increased significantly as their growth opportunities declined. In addition, these firms lengthened the maturity structure of their debt, decreased the ratio of private to public debt, and decreased the use of senior debt as their growth opportunities declined. The results complement other studies that have found cross-sectional relations between proxies for growth opportunities and leverage variables and validate the prominent role played by growth opportunities in the theory of corporate finance. r 2002 Elsevier Science B.V. All rights reserved. JEL classification: G32 Keywords: Capital structure; Debt policy; Payout policy; Growth opportunities

$ We gratefully acknowledge comments from Michael Barclay (the referee), Gershon Mandelker, Frederik Schlingemann, William Schwert (the editor), Shawn Thomas, and seminar participants at Arizona State University, Boston College, Michigan State University, University of North Carolina, Northwestern University, University of Oregon, and the University of Pennsylvania. This work has been supported by the Research Grants Council of the Hong Kong SAR (Project No. HKUST6184/98H). *Corresponding author. E-mail address: [email protected] (K. Lehn).

0304-405X/02/$ - see front matter r 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 4 0 5 X ( 0 2 ) 0 0 0 7 0 - 3

36

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

1. Introduction This paper examines how the debt policy of U.S. defense firms changed during 1980–95, a period spanning dramatic changes in growth opportunities for the U.S. defense industry. In the early 1980s, growth opportunities expanded as the Reagan administration increased U.S. defense spending substantially. In the late 1980s and early 1990s, growth opportunities dwindled as the cold war ended and U.S. defense spending fell substantially. These changes in the growth opportunities of U.S. defense firms provide a natural experiment for testing theories on the relation between a firm’s growth opportunities and its debt policy. Growth opportunities play a prominent role in the theory of corporate finance. It is generally believed that growth opportunities play an especially important role in determining a firm’s debt policy. Firms with good growth opportunities are expected to have little debt, and a high proportion of their debt is expected to be short-term rather than long-term, private instead of public, and senior (e.g., secured debt) instead of junior. Several papers have established cross-sectional relations between proxies for growth opportunities and debt variables. For example, Smith and Watts (1992) and Gaver and Gaver (1993) find that debt ratios vary inversely with proxies for growth opportunities. Barclay and Smith (1995a) find an inverse relation between proxies for growth opportunities and the maturity of corporate debt, while Stohs and Mauer (1996) find only mixed support for this relation. Houston and James (1996) find a direct relation between a firm’s market-to-book ratio, used as a proxy for its growth opportunities, and its use of private debt. Barclay and Smith (1995b) find that proxies for growth opportunities are directly related to the proportion of a firm’s debt accounted for by senior claims. Notwithstanding the available evidence, it is difficult to test hypotheses about the causal relation between growth opportunities and corporate debt policy with crosssectional data, because growth opportunities and corporate debt policy may be jointly determined. Baker (1993) states that this issue ‘‘plagues all of the papers in this recently burgeoning area. ... [T]here is a fundamental simultaneity in this problem that is extremely difficult to sort out’’ (p. 164). The changes in growth opportunities experienced by the U.S. defense industry permit a study of the relation between a firm’s growth opportunities and its debt policy that is free of this problem. The growth opportunities of U.S. defense firms fell substantially from the early 1980s to the mid-1990s as the U.S. defense budget declined more than 50% in real terms over this period. For purposes of this study, we can treat this decline in growth opportunities as exogenous to a firm’s debt policy. By examining how the debt policy of defense firms changed as their growth opportunities changed, we can infer a causal relation between a firm’s growth opportunities and its debt policy. Using a sample of 61 large defense firms (28 weapons manufacturers and 33 other defense firms) and a benchmark sample of 61 large manufacturing firms, we document that five commonly used proxies for growth opportunities decline significantly as real defense spending declines. The results are most pronounced for

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

37

weapons firms, which were most affected by the defense budget cuts. We then examine how several debt policy variables change as the growth opportunities of defense contractors change. First, we find significant changes in the level of corporate debt as the growth opportunities of defense firms change. Weapons manufacturers increase their debt sharply during the early years of reduced growth opportunities and then gradually reduce it to a level that is nonetheless higher than it was during the higher growth period. This result complements cross-sectional studies that find an inverse relation between growth opportunities and the level of corporate debt. It also suggests that debt plays an important role in the transitional period from a high to low growth regime, which is consistent with Jensen (1986) free cash flow theory. Second, we find several significant changes in the structure of defense firms’ debt as their growth opportunities decline. The maturity of debt issued after the contraction of growth opportunities is significantly longer than the maturity of debt issued in the higher growth period, a result consistent with Barclay and Smith (1995a). Weapons manufacturers significantly reduce their use of bank debt and increase their use of public debt in the lower growth environment, which is consistent with Houston and James (1996) and with the argument that the value of bank monitoring is inversely related to a firm’s growth opportunities. Finally, weapons firms reduce their use of secured debt in the low growth environment, which is consistent with results in Barclay and Smith (1995b). Section 2 describes and documents the changes in growth opportunities that occurred in the U.S. defense industry over the sample period. Section 3 presents predictions and evidence on the relation between growth opportunities and the level of debt in the defense industry. Section 4 includes predictions and evidence on the relation between growth opportunities and the structure of debt in the defense industry. Concluding comments are contained in Section 5.

2. Changes in growth opportunities in the U.S. defense industry, 1980–95 2.1. Changes in defense spending U.S. defense spending changed dramatically during the period of 1980–95. In the 1980 presidential campaign, Ronald Reagan promised to increase U.S. defense spending, which had fallen substantially after the Vietnam War. In November 1980, Reagan was elected president and Republicans gained control of the Senate, which ushered in a period of large increases in real defense spending. As Fig. 1 shows, real defense procurement (constant 1995 dollars) increased steadily from $53.6 billion in 1980 to more than $108.3 billion in 1987. Real outlays on defense procurement declined to roughly $100 billion in both 1988 and 1989, largely due to federal budgetary reforms. After the fall of the Berlin Wall in 1989, real defense procurements fell steadily and substantially to $46.7 billion in 1996. Real defense procurements in 1996 were less than they were in 1980, before the Reagan defense buildup.

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

38

120.00

350.00 100.00

300.00 80.00

250.00

200.00

60.00

150.00 40.00

100.00 20.00

50.00

0.00

Procurement expenditures (billions of 1995 dollars)

Defense expenditures (billions of 1995 dollars)

400.00

0.00

1980

1982

1984

1986

1988

1990

1992

1994

1996

Year Defense expenditures

Procurement expenditures

Fig. 1. Plot of total defense expenditures and procurement expenditures, 1980–96.

To examine how the changes in defense spending affected the revenues of individual defense contractors, we collected data on defense awards for a sample of 61 large defense contractors: 28 weapons and 33 nonweapons firms. This sample consists of all firms that appeared on the Department of Defense’s list of Top 100 Defense Contractors at least once during each of three periods: 1980–85 (the high growth period), 1986–88 (the no growth period), and 1989–95 (the negative growth period); the Center for Research in Security Prices (CRSP) tapes, and the Compustat (annual, research and full coverage) tapes. We further required that each firm in the sample received annual defense contract awards in excess of 5% of sales during the sample period. A list of the defense firms in the sample, denoted by whether they are a weapons or nonweapons firm, is contained in Appendix A. Fig. 2 plots the total real sales to the U.S. government for weapons and nonweapons manufacturers in the sample (in constant 1995 dollars). The figure illustrates two relevant points for the tests that follow. First, the value of real sales to the U.S. government is much higher for weapons firms. The real value of contract awards ranges from approximately $50 billion to approximately $125 billion over the sample period for weapons firms, versus a range of approximately $7 billion to $20 billion for nonweapons firms. In almost every year of the sample period, the value of contract awards is more than five times higher for weapons firms. To gauge the relative importance of defense contract awards for weapons versus nonweapons firms, we computed the ratio of contract award to sales for each firm in each year. These data show that the fortunes of weapons firms are much more dependent on defense contracts than those of nonweapons firms. The median value of this ratio is 0.31 during 1980–85, 0.30 during 1986–88, and 0.25 during 1989–95 for weapons

39

140 120 100 80 60 40 20

85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95

84

19

83

19

82

19

81

19

19

80

0 19

Sales to US Government (billions of constant 1995 dollars)

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

Year Nonweapons manufacturers

Weapons manufacturers

Fig. 2. Plot of total sales to U.S. government for weapons and nonweapons manufacturers.

firms compared with 0.09, 0.12, and 0.10, respectively, for nonweapons firms. In each subperiod, the difference in contract awards to sales ratio is significant at the 0.01 level. Fig. 2 also shows that weapons firms experienced a substantially larger decline in defense contract awards than did the nonweapons firms. The real value of sales to the U.S. government fell by roughly $64 billion from 1986 to 1995 (from $125 billion to $61 billion for weapons firms), or approximately $2 billion on average for the 28 weapons firms in our sample. The corresponding decline for nonweapons firms is roughly $13 billion (from $20 billion to $7 billion), or about $400 million on average for the 33 nonweapons firms in the sample. Hence Fig. 2 establishes that both the Reagan buildup and the end of the cold war had larger effects on weapons firms than on nonweapons firms. To further examine how the changes in defense spending affected the fortunes of weapons firms versus other defense firms, we conducted an event study around two announcements that crystallize the changes in growth opportunities for U.S. defense firms. The first event is the election of Ronald Reagan as president and a Republican Senate on November 5, 1980. The second event is an announcement on November 20, 1989 by Defense Secretary Richard Cheney that substantial cuts would be made in the U.S. defense budget following the dismantling of the Berlin Wall and the demise of the Soviet Union. Over a window of one day before through one day after the election of Reagan and a Republican Senate, the average cumulative abnormal returns (CARs) for weapons manufacturers is 4.84% and significant at the 1% level.

40

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

Furthermore, 81.5% of the weapons firms experienced a positive CAR over this window. The average CAR for other defense contractors (nonweapons firms) is 0.34% and not significant. Over the corresponding window surrounding Secretary Cheney’s announcement, the average CAR for weapons firms is 4:5% and highly significant. Only 18.5% of weapons firms experienced positive CARs around this announcement. The average CAR for nonweapons firms around this announcement is 0:27% and not significant. The event study results show that the fortunes of U.S. defense contractors were expected to increase under the Reagan administration in the early 1980s and to decline sharply in the late 1980s with the end of the cold war. They also reveal that this effect is confined largely to manufacturers of weapons and not to other defense firms, a result that conforms with intuition. The correlation coefficient between the CARs for the two events is 0:56 and significant at the 0.01 level for weapons firms, indicating that firms benefiting the most from the Reagan buildup suffered the most with the end of the cold war. The corresponding correlation coefficient for nonweapons firms is 0:17 and not significant. 2.2. Changes in proxies for growth opportunities The decline in defense spending presumably reduced the value of defense contractors in two ways. First, it reduced the value of the industry’s assets in place, because defense contractors would be expected to sell fewer of their existing products. Second, it reduced the value of the industry’s growth options, because the expected profits from new products declined. Because we are interested in the relation between growth opportunities and debt policy, it is important to document that the decline in defense spending reduced the value of defense firms’ growth options and not simply the value of their assets in place. To do this, we examine changes in several commonly used proxies for growth opportunities. Specifically, we examine five widely used proxies for growth opportunities: 1. the ratio of the market value of a firm’s assets to the book value of its assets (MBA), measured as the ratio of the sum of the book value of debt, the book value of preferred stock, and the market value of equity to the book value of assets at year-end; 2. the ratio of the market value of equity to the book value of equity (MBE), measured as the ratio of the market value of equity to the book value of equity at year-end (Collins and Kothari, 1989); 3. the earnings-to-price ratio (EPR), computed as the ratio of earnings per share divided by closing stock price at year-end (Chung and Charoenwong, 1991); 4. the ratio of capital expenditures to the book value of assets at year-end (CAPEX); and 5. the ratio of research and development (R&D) expenditures to the book value of assets at year-end.

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

41

To examine whether the proxies for growth opportunities change in the anticipated direction for defense firms over the sample period, we use the sample of defense firms described above. To control for more general changes in the value of these ratios over time, we construct a benchmark sample of manufacturing firms. The benchmark sample is identified from among firms that are listed on both Compustat (annual, research and full coverage tapes) and CRSP; have no missing data at the end of fiscal year 1979; and belong to Compustat SIC codes 2000-3999 in 1979, which qualifies them as a manufacturing firm. In addition, we exclude foreign firms. For each defense firm, we identify a set of potential benchmark firms whose market value of assets at the end of 1979 is within 50% and 150% of the market value of the corresponding defense firm. From among this set of potential benchmark firms, we identify the firm that is closest to the corresponding defense firm in terms of its market-to-book ratio at the end of 1979. That firm is then included in the benchmark sample. If that firm is delisted over the sample period, we replace it with the next best match. This frees the benchmark sample of any survivorship bias. Hence, the benchmark sample is designed to resemble the defense sample in terms of both size and market-to-book ratio at the end of 1979. A list of the benchmark sample is contained in the appendix, alongside its corresponding defense firm. Fig. 3 plots the average annual value of the five growth proxies for the weapons, nonweapons, and benchmark samples. The graphs generally show that, during the early 1980s, the average value of each proxy for growth opportunities was roughly the same for the three samples. From the mid-1980s through the early 1990s, each of the five proxies suggests declining growth opportunities for weapons versus other firms. Since the early 1990s, the average values of the proxies based on stock prices (i.e., the market-to-book ratios and the earnings-to-price ratio) changed in ways consistent with an increase in growth opportunities for weapons firms, whereas the two proxies not based on stock prices (i.e., the ratios of capital expenditures to assets and R&D expenditures to assets) reveal a steady decline in growth opportunities in the 1990s. To test whether the proxies for growth opportunities changed for defense firms, we estimate a fixed effects regression on a panel database consisting of annual data for the 122 firms (61 defense, 61 benchmark) over the period of 1980 through 1995. The natural log of the proxy for growth opportunities serves as the dependent variable in these equations. The independent variables include the natural log of asset value in the contemporaneous year, dummy variables for weapons and nonweapons defense firms, dummy variables for the 1986–88 and 1989–95 periods, and interaction terms for weapons and nonweapons firms in each of the two subperiods. The 1986–88 subperiod is when real defense spending had leveled off after the Reagan buildup, and the 1989–95 subperiod is when real defense spending fell substantially after the end of the cold war. Our principal interest is whether the coefficient on the variable interacting the defense firms with the two dummy variables for 1986–88 and 1989–95 reveals changes in growth opportunities for defense firms during these two periods. The results, contained in Table 1, demonstrate that all five measures reveal changes in growth opportunities for weapons firms vis-a" -vis the benchmark firms.

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

42

1.50

Market-to-book assets

1.40 1.30 1.20 1.10 1.00 0.90 0.80 0.70 0.60 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Year

(a)

4.10

Market-to-book equity

3.60 3.10 2.60 2.10 1.60 1.10 0.60 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Year

(b)

Earnings-to-price ratio

0.23 0.19 0.15 0.11 0.07 0.03 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Year

(c)

0.07

Benchmarks 0.08

Weapons 0.07

Nonweapons

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

43

0.1300

Capital expenditure/total assets

0.1200 0.1100 0.1000 0.0900 0.0800 0.0700 0.0600 0.0500 0.0400 0.0300 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Year

(d) 0.0600 0.0550 R&D expenditure/total assets

0.0500 0.0450 0.0400 0.0350 0.0300 0.0250 0.0200 0.0150 0.0100 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995

(e)

Year Benchmarks

Weapons

Nonweapons

Fig. 3. (Continued).

The coefficient on the variable that interacts the dummy for weapons firms with the 1989–95 dummy has the expected sign and is highly significant in all five equations. The corresponding coefficient on the variable interacting weapons firms with the 1986–88 dummy also has the expected sign in four of the equations (all except the 3 Fig. 3. Average annual value of five growth proxies for benchmark, weapons, and nonweapons samples. Panel A plots the average market-to-book assets ratio for the benchmark, weapons, and nonweapons samples for the sample period 1980–95; panel B, the average market-to-book equity; panel C, the average earnings-to-price ratio; panel D, the capital expenditure to assets ratio; and panel E, the research and development expenditure to assets ratio.

44

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

R&D equation), and it is significant at the 1% level in the MBA and MBE equations. The coefficient on the 1989–95 interaction term reveals a more substantial decline in growth opportunities during this period than in the 1986–88 period, which conforms with the fact that real defense spending fell more substantially during the 1989–95 period. The MBE equation also reveals a significant decline in growth opportunities for nonweapons defense firms in the 1989–95 period, although the coefficient is less negative than the corresponding coefficient for weapons firms. The MBA, EPR, and CAPEX equations reveal no significant change in growth opportunities for nonweapons defense firms in the 1989–95 period. The R&D equation reveals an increase in growth opportunities for nonweapons firms over this period. Overall, the results in Table 1 generally reveal that the growth opportunities of weapons firms declined significantly during 1986–88 and 1989–95 vis-a" -vis the benchmark firms. This is true for growth proxies that are based on stock prices as well as for those not based on stock prices. Hence, even though the stock-based measures of growth opportunities increase absolutely for weapons firms, they decline relative to the benchmarks. Because we test whether the debt policy variables of weapons firms change relative to the benchmark sample over time, it is important to

Table 1 Fixed-effect regressions of proxies for growth opportunities on the natural log of asset value, dummy variables for the 1986–88 and 1989–95 periods and variables interacting weapons and nonweapons firms with time. Dependent variable is the natural log of Independent variables

Marketto-book (assets)

Marketto-book (equity)

Earningsto-price ratio

Capital expenditures/ total assets

Research expenditures/ total assets

Intercept

0.307 (2.0)nn 0.056 (3.1)nnn 0.146 (5.3)nnn 0.241 (10.0)nnn 0.157 (3.4)nnn 0.226 (5.9)nnn 0.069 (1.6) 0.147 (4.1)nnn 1920 0.07

1.322 (5.3)nnn 0.119 (4.2)nnn 0.338 (7.7)nnn 0.516 (13.5)nnn 0.199 (2.7)nnn 0.194 (3.2)nnn 0.168 (2.4)nn 0.123 (2.2)nn 1873 0.07

3.604 (8.9)nnn 0.141 (3.0)nnn 0.359 (5.1)nnn 0.522 (8.4)nnn 0.151 (1.3) 0.199 (2.1)nn 0.061 (0.6) 0.027 (0.3) 1702 0.10

1.316 (5.5)nnn 0.130 (4.7)nnn 0.242 (6.0)nnn 0.356 (10.0)nnn 0.014 (0.2) 0.272 (4.9)nnn 0.006 (0.1) 0.007 (0.1) 1918 0.01

1.067 (4.6)nnn 0.304 (11.4)nnn 0.001 (0.0) 0.080 (2.3)nn 0.037 (0.6) 0.132 (2.5)nn 0.162 (2.7)nnn 0.107 (2.1)nn 1679 0.01

Log (Assets) 1986–88 dummy 1989–95 dummy Weapons1986–88 dummy Weapons1989–95 dummy Nonweapons1986–88 dummy Nonweapons1989–95 dummy N R2

The t-statistics are in parentheses. nnn Significant at the 1% level. nn Significant at the 5% level. * Significant at the 10% level.

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

45

establish that the growth opportunities of defense firms declined relative to the benchmark sample, regardless of whether they increased in an absolute sense.

3. Growth opportunities and the level of corporate debt 3.1. Predictions A firm’s debt-to-value ratio is expected to vary inversely with its growth opportunities for at least two reasons. First, the agency costs associated with the debtholder–stockholder conflict are likely to be increasing in a firm’s growth opportunities. One example of this is the underinvestment problem identified by Myers (1977). Myers argues that firms with risky debt have an incentive to underinvest in value-increasing projects. This occurs because shareholders, who control the investment decision, bear the entire cost of the projects but receive only a fraction of the increase in firm value; part of it is shared with the debtholders. Because the cost of the underinvestment problem increases with a firm’s growth opportunities, firms with good growth opportunities have an incentive to finance their operations with equity instead of debt. More generally, debtholders face higher costs of monitoring stockholders in high growth firms than they do in lower growth firms. Because the assets of high growth firms are largely intangible, debtholders have more difficulty observing how stockholders use assets in high growth firms. For example, debtholders and stockholders often conflict over the desirable amount of firm risk, with debtholders generally preferring less risk. It is easier for stockholders to increase firm risk, and more costly for debtholders to detect increases in firm risk, in high growth firms with mostly intangible assets than it is in low growth firms with more fixed assets in place. As a result, the costs of debt financing are higher in firms with more growth opportunities. Hence, a firm’s debt level is expected to vary inversely with its growth opportunities. Second, Jensen (1986) argues that debt can reduce the agency costs of free cash flow, which are most severe for firms with low growth opportunities. According to this argument, the interests of managers and shareholders are likely to diverge in industries that generate abundant free cash flow (i.e., operating cash flow minus cash needed to fund value-increasing investments). Managers supposedly have a stronger preference for retaining free cash flow within the firm, while shareholders have a stronger preference for using free cash flow to fund higher payouts in the form of dividends and share repurchases. Debt, according to Jensen, is one means of resolving this tension. By issuing debt, firms commit to pay out future free cash flows to investors, thereby reducing the likelihood that managers will squander free cash flow on value-reducing investments. By paying the proceeds of the debt issues to shareholders in the form of dividends and share repurchases, stockholders capture the value increase associated with the reduced agency costs of free cash flow. Given that the agency costs of free cash flow vary inversely with a firm’s growth

46

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

opportunities, this theory predicts an inverse relation between growth opportunities and debt ratios. Based on these two theories, we expect the debt-to-value ratios of U.S. defense firms to increase as their growth opportunities decline in the mid-to-late 1980s and early 1990s. 3.2. Results To examine how the capital structure of defense firms changed as their growth opportunities changed, we estimate two sets of fixed effects regression on the panel database. The dependent variable in one set of regressions is the ratio of the book value of a firm’s debt to the market value of its assets (defined as the sum of the book value of debt, the book value of preferred stock, and the market value of equity). The dependent variable in the second set of regressions is the ratio of the book value of a firm’s debt to the book value of its assets. In both sets of regressions, we estimate three equations. In the first equation, the only independent variables are dummy variables for the two subperiods, 1986–88 and 1989–95, and variables that interact these time variables with dummy variables for whether the firm is a weapons or nonweapons defense firm. In the second equation, we use a specification similar to the one used by Rajan and Zingales (1995). Independent variables include proxies for firm size (the natural log of asset value), growth opportunities [market-to-book (asset) ratio], profitability (ratio of operating income to total assets), and tangibility of assets (ratio of net fixed assets to total assets), but not the independent variables included in the first equation. In the third equation, we include all of the independent variables in both the first and second equations. The results are reported in Table 2. In both sets of equations, the key variables of interest, the interaction of weapons firms with the two dummy variables for subperiods, enter with positive and significant coefficients. This result reveals that, relative to the higher growth period of 1980–85, defense firms significantly increased their debt-to-value ratios in the lower growth periods of 1986– 88 and 1989–95. The coefficients on the interaction of weapons firms with the 1989– 95 dummy variable are almost identical to the corresponding coefficients on the interaction of weapons firms with the 1986–88 dummy variable. Hence, the debt-tovalue ratio of defense firms increased in both subperiods relative to the earlier high growth period, but it did not change much from the 1986–88 to the 1989–95 periods. In the debt-to-market value equations, the coefficients on the key interaction variables fall substantially when these variables enter with the Rajan-Zingales variables. For example, when the Rajan-Zingales variables are excluded, the coefficient on the interaction of weapons firms with the 1989–95 dummy variable is 0.106 and is significant at the 0.01 level. When the Rajan-Zingales variables are included, the coefficient on this variable falls to 0.034 and is significant at the 0.05 level. The most likely reason for this is that one of the Rajan-Zingales variables, the market-to-book ratio, is picking up much of the reduction in growth opportunities for weapons firms during the 1989–95 period. Overall, the results are highly consistent with the predictions made above.

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

47

Table 2 Fixed effect regressions of leverage ratios on firm size, market-to-book ratio, tangibility of assets, profitability, dummies for time periods and variables interacting firm dummies and time dummies. Debt-to-market value of assets

Debt-to-book value of assets

Independent variables

(1)

(2)

(3)

(4)

(5)

(6)

Intercept

0.278 (56.0)nnn

0.640 (10.0)nnn 0.129 (19.0)nnn 0.080 (9.7)nnn 0.008 (0.2) 0.749 (12.5)nnn

0.590 (9.2)nnn 0.121 (17.4)nnn 0.084 (9.7)nnn 0.058 (1.5) 0.759 (12.3)nnn 0.002 (0.2) 0.021 (2.1)nn 0.032 (1.8)n 0.034 (2.3)nn 0.048 (2.8)nnn 0.052 (3.7)nnn 1918 0.322

0.214 (53.5)nnn

0.392 (6.8)nnn 0.083 (13.7)nnn 0.040 (5.4)nnn 0.096 (2.9)nnn 0.601 (11.2)nnn

0.347 (6.0)nnn 0.074 (11.9)nnn 0.029 (3.8)nnn 0.036 (1.0) 0.556 (10.1)nnn 0.021 (2.2)nn 0.030 (3.3)nnn 0.031 (1.9)n 0.029 (2.1)n 0.030 (2.0)nn 0.023 (1.8)n 1918 0.203

Log (assets) Market-to-book assets Tangible assets Profitability 1986–88 dummy 1989–95 dummy Weapons1986–88 dummy Weapons1989–95 dummy Nonweapons1986–88 dummy Nonweapons1989–95 dummy N R2

0.003 (0.2) 0.016 (1.4) 0.081 (3.9)nnn 0.106 (6.2)nnn 0.012 (0.6) 0.020 (1.2) 1920 0.048

1918 0.305

The t-statistics are reported in parentheses.

nnn

0.038 (3.8)nnn 0.064 (7.4)nnn 0.044 (2.61)nnn 0.043 (3.1)nnn 0.022 (1.4) 0.032 (2.4)nn 1942 0.087

Significant at the 1% level.

nn

1918 0.184

Significant at the 5% level.

Size enters with a positive and significant coefficient, consistent with the argument that the costs of debt are lower for larger, perhaps less risky, firms. Market-to-book ratio enters negative and significantly in the debt-to-market value equation, consistent with the prediction that higher growth firms rely less on debt financing. However, a significant positive relation exists between market-to-book ratio and the ratio of debt to the book value of assets. This relation most likely is a statistical phenomenon. Because the denominator in both variables is the same (i.e., the book value of assets), it reveals that firms with higher market values have more debt. Tangibility of assets is significant in only one equation. Profitability enters with a negative and significant coefficient in all equations, consistent with previous research and the pecking order theory. We also examine how the net debt and net equity issues of defense firms vary over time as their growth opportunities change. Net debt issues is defined as the value of debt issued in a given year minus the value of debt repayments in the year. Net equity issues is defined as the value of equity issues in a given year minus the value of shares

48

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

Table 3 Fixed effects regressions of new debt issues/MVA (market value of assets) and new equity issue /MVA (market value of assets) on the natural log of assets, dummy variables for the 1986–88 and 1989–95 periods, and variables interacting weapons firms with time. Independent variables

Net debt issues/MVA (1)

Intercept

0.030 (6.7)nnn

Log assets 1986–88 dummy 1989–95 dummy Weapons1986–88 dummy Weapons1989–95 dummy N R2

0.029 (2.6)nnn 0.037 (3.8)nnn 0.053 (3.4)nnn 0.023 (1.8)n 844 0.01

Net equity issues/MVA

(2)

(3)

(4)

0.299 (4.7)nnn 0.040 (5.2)nnn 0.042 (3.7)nnn 0.068 (6.0)nnn 0.046 (3.1)nnn 0.021 (1.7)n 844 0.01

0.001 (0.5)

0.067 (1.6) 0.008 (1.6) 0.009 (1.2) 0.008 (1.1) 0.003 (0.3) 0.003 (0.4) 690 0.00

0.12 (1.7)n 0.001 (0.2) 0.002 (0.2) 0.003 (0.4) 690 0.01

The t-statistics are in parentheses. nnn Significant at the 1% level. nn Significant at the 5% level. n Significant at the 10% level.

repurchased in that year. To adjust for firm size, both variables are divided by the market value of the firm’s assets, as defined above. Because the data on net debt issues and net equity issues were hand collected, we limit this analysis to the 28 weapons firms and their corresponding matches. Table 3 presents results from regressions in which the ratio of net debt issues to the market value of assets and the ratio of net equity issues to the market value of assets serve as dependent variables. The independent variables include the natural log of firm size, the dummy variables for the 1986–88 and 1989–95 subperiods, and variables that interact the subperiod dummy variables with a dummy variable for weapons firms. The results reveal a significant increase in the net debt issues of weapons firms during both subperiods. However, the coefficient on the 1986–88 interaction variable is more than twice the size of the coefficient on the 1989–95 interaction variable, indicating that the large increase in net debt issues occurred when growth opportunities declined more modestly. Furthermore, the coefficient on the 1986– 88 interaction variable is significant at the 0.01 level, whereas the coefficient on the 1989–95 interaction variable is significant at only the 0.10 level. When net equity issues is the dependent variable, the coefficients on the two interaction variables are not statistically significant, indicating that defense firms were not systematically retiring equity at higher rates during the lower growth periods. Fig. 4 plots the average annual value of the two leverage variables used in the regressions for each of the three samples. The graphs show a sharp increase in the

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

49

leverage ratios for weapons firms relative to the two other samples in the mid-1980s and then a reduction in these ratios in the 1990s. Given that leverage ratios could decline because of changes in either the value of leverage (i.e., the numerator) or the firm’s assets (i.e., the denominator), we also plot the average annual value of real

0.50

Debt/capitalization

0.45 0.40 0.35 0.30 0.25 0.20 0.15

95

94

19

19

92

93

19

91

19

90

19

89

19

88

19

87

19

86

(a)

19

85

19

84

19

83

19

19

81

82

19

19

19

80

0.10

Year

Debt/Book value of assets

0.35 0.30 0.25 0.20 0.15 0.10 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 Year

(b) 20000

Real debt

16000 12000 8000 4000 0

(c)

80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 Year Benchmarks

Weapons

Nonweapons

Fig. 4. Average annual value of two leverage variables used in regressions. Panel A shows a plot of market leverage ratio for benchmark, weapons, and nonweapons for the sample period 1980–95; panel B, the book leverage ratio; panel C, a plot of real debt.

50

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

debt for the three samples in panel C of Fig. 4. It shows a striking increase in the value of real debt for weapons firms in the mid-1980s and a continued increase in this value through 1995. This plot suggests that the decline in leverage ratios for weapons firms in the 1990s has resulted from an increase in the value of these firms, not a decrease in the value of their nominal debt. The large increase in the debt-to-value ratios of defense firms began in the mid1980s, before the shock of the end of the cold war. After the shock, they declined from their mid-1980 levels but remained higher than they were during the high growth period of 1980–85. In short, it appears that abrupt reduction in growth opportunities associated with the end of the cold war had less effect on the debt-tovalue ratios of defense firms than did the more moderate reduction in growth opportunities associated with the budgetary reforms of the mid-1980s. One interpretation of these results is that the abrupt end of the cold war increased uncertainty about the future free cash flows of defense firms. If so, the increase in uncertainty offsets, at least in part, the effect that reduced growth opportunities have on the debt-to-value ratios of defense firms. In contrast, the more moderate reduction in growth opportunities for defense firms in the mid-1980s had less effect on uncertainty about the free cash flows of defense firms. Empirically, we find that controlling for other factors, the dispersion of analysts’ earnings forecasts, a commonly used measure of uncertainty, almost doubled for weapons firms during 1989–95 versus 1986–88. This substantial increase in uncertainty might explain why a larger increase in the use of debt occurred during 1986–88 than 1989–95.

4. Growth opportunities and the structure of corporate debt In addition to affecting the level of corporate debt, growth opportunities are expected to affect the structure of corporate debt, including its maturity structure, its priority structure, and the mix of public versus private debt. We presume that these dimensions of corporate debt are jointly chosen and serve to reinforce each other. This perspective would suggest that we estimate a system of equations to test the relation between changes in the growth opportunities of defense contractors and changes in the structure of their debt. However, as a practical matter, we do not have sufficient theory to identify the structural coefficients in such a system of equations. As a result, we treat these dimensions of debt separately in the tests that follow. 4.1. Maturity structure 4.1.1. Predictions Several papers attempt to explain the variation in the maturity structure of corporate debt. Some of the more prominent hypotheses use taxes (Brick and Ravid, 1985; Kane et al., 1985), signaling (Flannery, 1986), and liquidity risk (Diamond, 1991) as explanations for the variation in corporate debt maturity structure. In addition, several papers have examined the relation between a firm’s growth opportunities and the maturity structure of its debt.

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

51

Myers (1977) argues that firms can mitigate the underinvestment problem by issuing short-term debt. According to the argument, using short-term debt that matures before a firm exercises its growth options allows stockholders to capture a larger proportion of the value created by positive net present value projects. Shortterm debt effectively allows stockholders to buy back debt at prices that do not reflect the value of new profitable investment opportunities. Alternatively, if a firm had risky long-term debt outstanding and wanted to exercise a value-increasing growth option, it would have to buy back the debt at a price that reflected the value of the project to the debtholders. Myers’s argument leads to the prediction that an inverse relation exists between a firm’s growth opportunities and the maturity structure of its debt. Barnea et al. (1980) argue that short-term debt also can mitigate the incentives of stockholders to redistribute wealth from bondholders by substituting high risk low value assets for low risk high value assets. This result holds because the value of short-term debt is less sensitive to changes in firm risk than is the value of long-term debt. If the cost of monitoring asset substitution by stockholders is directly related to a firm’s growth opportunities, as might be the case because of the high proportion of intangible assets in high growth firms, then this argument also predicts an inverse relation between growth opportunities and the debt maturity. Several papers examine the cross-sectional relation between proxies for growth opportunities and the maturity structure of corporate debt. Barclay and Smith (1995a) find a significant inverse relation between proxies for growth opportunities and the percentage of debt that matures in more than three years. However, Stohs and Mauer (1996) do not find a significant negative relation between a firm’s marketto-book ratio and the value-weighted average maturity of its debt. Guedes and Opler (1996) find a significant inverse relation between proxies for growth opportunities and the maturity of newly issued debt, which is consistent with Myers’s argument and the evidence in Barclay and Smith.

4.1.2. Results We examine how two measures of debt maturity change over time for defense firms as their growth opportunities change. The first measure, which is similar to the one used by Stohs and Mauer, is the weighted average maturity of a firm’s outstanding liabilities, in which the weights are established by the book values of the liabilities. The second measure, which is similar to the one used by Guedes and Opler, is the value-weighted maturity of newly issued debt. The second measure is more telling for our purposes, because it more directly shows how the maturity structure of debt changes after changes to the firms’ growth opportunities. We estimate two sets of regressions in which the natural log of the two measures of debt maturity serve as the dependent variable. Because the maturity data are hand collected, these regressions are run only for the sample of weapons firms and their corresponding benchmarks. The independent variables are the same as those used in the regressions reported in Table 3. The results from this test are contained in Table 4.

52

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

Table 4 Ordinary least squares regressions of debt maturity on log of firm size, dummy variables for the 1986–88 and 1989–95 periods and variables interacting weapons firms with time. Independent variables

Intercept

Log (maturity of outstanding liabilities) (1)

(2)

(3)

(4)

0.889 (24.5)nnn

2.858 (5.8)nnn 0.462 (7.6)nnn 0.132 (1.5) 0.368 (4.1)nnn 0.175 (1.5) 0.424 (4.3)nnn 886 0.01

1.07 (15.5)nnn

0.465 (0.4) 0.073 (0.6) 0.088 (0.5) 0.139 (0.8) 0.597 (2.6)nnn 0.524 (2.7)nnn 657 0.01

Log assets 1986–88 dummy 1989–95 dummy Weapons1986–88 dummy Weapons1989–95 dummy N R2

Log (value-weighted maturity of debt issues)

0.015 (0.2) 0.009 (0.1) 0.256 (2.0)nn 0.444 (4.3)nnn 886 0.01

The t-statistics are in parentheses.

nnn

Significant at the 1% level.

0.063 (0.4) 0.078 (0.5) 0.604 (2.6)nnn 0.524 (2.7)nnn 657 0.01 nn

Significant at the 5% level.

The results show that the maturity structure of debt lengthened significantly for weapons firms as their growth opportunities declined. When the value-weighted maturity of a firm’s entire debt is used as the dependent variable, the coefficient on the interaction variable for weapons firms in the 1989–95 period reveals that these firms increased the maturity of their debt by more than 40% during this period. Furthermore, this coefficient is significant at the 0.01 level. The coefficient on the simple 1989–95 dummy variable is negative and significant at the 0.01 level, indicating that firms generally were shortening their debt maturity during this period by almost 37%. The proxy for firm size enters with a positive and significant coefficient, indicating that larger firms have longer debt maturities. This result is consistent with Barclay and Smith and Stohs and Mauer. We find similar results for the regressions in which the maturity of new debt issues is the dependent variable. The coefficients on the interaction variables for weapons firms in both the 1986–88 and 1989–95 periods are large and significant at the 0.01 level. The results indicate that the maturity of newly issued debt increased by roughly 60% during the 1986–88 period and 52% during the 1989–95 period vis-a" -vis the earlier high growth period. These results also explain why the interaction variable for weapons firms in the 1989–95 period entered the first set of equations more significantly than did the interaction variable for weapons firms in the 1986–88 period. The cumulative impact of large, long-term debt issues during both the 1986– 88 and 1989–95 periods caused the weighted average maturity of all outstanding liabilities to be substantially higher during the 1989–95 period than it was during the

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

53

1986–88 period. These results are consistent with Guedes and Opler. Our test also examines one of the shortcomings of examining the maturity of new debt issues, as stated by Guedes and Opler (1996): ‘‘The weakness of the incremental approach is that it provides noisy tests of agency theories of maturity choice that depend on slowly changing characteristics such as asset lives and the investment opportunity set’’ (p. 1831). By examining an abrupt change to an industry’s growth opportunities, we are able to detect a strong relation between growth opportunities and debt maturity.

4.2. Private versus public debt 4.2.1. Predictions It is widely believed that firms face the following trade-off in choosing between private placement of debt and public issuance of debt. Private debt, which is usually bank debt, is generally tightly held, while public debt is typically diffusely held. Because private lenders bear a large proportion of the wealth consequences associated with a given debt claim, they have stronger incentives to invest in information and monitor the activities of the borrowing firm. Hence, a major advantage of private debt is that it mitigates adverse selection and moral hazard problems associated with lending activity. A second advantage of private debt is that, compared with public debt, it provides more flexibility to restructure debt contracts if the borrowing firm encounters financial problems. The restructuring of public debt is more difficult to restructure for two reasons. First, the diffuse ownership of public debt results in high coordination costs for debtholders. Second, restructurings of public debt are regulated by the Trust Indenture Act of 1939, which increases the transaction costs of public debt restructurings (Smith and Warner, 1979). Public debt is thought to have two advantages over private debt. First, the transaction costs of issuing public debt may be lower than those associated with private placements of debt, owing to economies of scale in public debt issues (Blackwell and Kidwell, 1988). Second, several papers have argued that a disadvantage of private debt is that tight lender control of a firm can distort a firm’s investment incentives due to possible ‘‘hold up’’ power of lenders (Rajan, 1992; Sharpe, 1990) or inefficient liquidations (Diamond, 1993) prompted by the lender. It is expected that firms with good growth opportunities will rely more on private debt than firms with fewer growth opportunities. Because firms with good growth opportunities are more difficult to monitor, the information production and monitoring provided by private lenders is especially valuable in these firms. Hence, a direct relation is expected between a firm’s growth opportunities and its use of private debt. Houston and James (1996) find that this relation holds for firms with multiple banking relationships, but not for firms with single bank relationships, presumably because of the hold-up power of single bank lenders. Krishnaswami et al.

54

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

(1999) find evidence consistent with their argument that the monitoring provided by private debt is more valuable when firms have more growth options. Because the value of the information production and monitoring function provided by private lenders is inversely related to a firm’s growth opportunities, we expect that defense contractors reduced their use of private debt as their growth opportunities declined. However, a countervailing consideration is that the greater uncertainty created by defense spending shocks increases the costs of monitoring debt agreements, which, in turn, should lead to more private debt. Whether or not this countervailing effect offsets the effect of reduced growth opportunities on private debt is an empirical matter.

4.2.2. Results To examine the relation between the use of private debt by defense firms and their growth opportunities, we collected data on the mix of private and public debt for the 28 weapons firms and their corresponding benchmarks over the sample period. We define private debt as bank debt plus privately placed debt. Public debt is defined as publicly traded notes, bonds, debentures plus commercial paper. Any debt that could not be classified as either private or public based on information in the annual reports is classified as ‘‘other debt’’. To test whether the use of private debt by defense firms changed as their growth opportunities changed, we estimated several regressions in which the dependent variables are different permutations of a firm’s private and public debt. The independent variables are identical to those used in the regressions reported in Table 3. The results from these regressions are reported in Table 5. The results reveal a strong reduction in the use of private debt by defense firms as their growth opportunities contract. When the ratio of private debt to total debt is the dependent variable, the coefficient on the interaction variable for weapons firms in the 1989–95 period is negative and significant at the 1% level. The coefficient of 0:284 reveals an economically significant relation. During this period, weapons firms reduced private debt as a percentage of their total debt by about 28 percentage points. In contrast, the coefficient on the interaction variable for weapons firms in the 1986–88 period is not only less negative ð0:127Þ; but also significant at the 0.01 level. The coefficient on firm size is negative and significant at the 1% level indicating that large firms rely less on private debt than do small firms. As expected, contrary results are found in the regression in which the ratio of public debt to total debt is the dependent variable. In this equation, the coefficient on the interaction variable for weapons firms in the 1989–95 period is positive (0.198) and significant at the 0.01 level, revealing that weapons firms increased public debt as a percentage of their total debt by about 20 percentage points during this period. The corresponding coefficient on the interaction variable for weapons firms in the 1986– 88 period is less positive (0.149) and significant at the 0.01 level, indicating that defense firms also relied more on public debt in the mid-1980s than they did during the higher growth period of the early 1980s.

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

55

Table 5 Fixed effects regressions of the ratio of private debt/total debt and public debt/total debt on the natural log of assets, dummy variables for the 1986–88 and 1989–95 periods, and variables interacting weapons firms with time. Independent variables

Private debt/total debt

Public debt/total debt

Intercept

0.841 (4.5)nnn 0.066 (3.1)nnn 0.058 (1.9)n 0.211 (7.2)nnn 0.127 (3.1)nnn 0.284 (8.0)nnn 821 0.10

0.002 (0.0) 0.067 (3.4)nnn 0.017 (0.6) 0.009 (0.3) 0.149 (3.9)nnn 0.198 (6.0)nnn 821 0.16

Log assets 1986–88 dummy 1989–95 dummy Weapons1986–88 dummy Weapons1989–95 dummy N R2 The t-statistics are in parentheses.

nnn

Significant at the 1% level. n Significant at the 10% level.

4.3. Priority structure of debt 4.3.1. Predictions A relation between growth opportunities and the priority structure of corporate debt also has been documented in the literature. Stulz and Johnson (1985) argue that the underinvestment problem identified by Myers can be mitigated if new investment projects are financed with secured debt. Because secured debt provides debtholders with title to pledged assets, it limits the extent to which debtholders can benefit from positive NPV projects. This, in turn, makes it more likely that shareholders will accept such projects, thereby mitigating the underinvestment problem. Smith and Warner (1979) argue that high priority debt also can mitigate the asset substitution problem. Barclay and Smith (1995b) empirically examine the relation between growth opportunities and the priority structure of corporate debt. Because the underinvestment and asset substitution problems are more severe for firms with high growth opportunities, they hypothesize that firms with high growth opportunities should rely more on high priority debt, such as capitalized leases and secured debt. Barclay and Smith find evidence consistent with this hypothesis. 4.3.2. Results To test whether the priority structure of defense firms’ debt changed systematically as their growth opportunities changed, we estimate several regression equations in which the dependent variables are the proportion of total corporate debt accounted for by debt of different priorities. Following Barclay and Smith, we decompose corporate debt claims into secured debt, capitalized leases, ordinary debt, and

56

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

Table 6 Fixed effect regressions of the debt in each priority class as a fraction of total debt on log of firm size, dummy variables for the 1986–88 and 1989–95 periods and variables interacting weapons firms with time. Independent variables Intercept

Secured debt/ total debt

0.190 (2.3)nn Log assets 0.014 (1.5) 1986–88 dummy 0.005 (0.4) 1989–95 dummy 0.010 (0.8) Weapons1986–88 0.071 dummy (4.1)nnn Weapons1989–95 0.065 dummy (4.2)nnn N 750 R2 0.07

Capital lease/ (Secured debt Ordinary debt/ total debt + capital lease)/ total debt total debt

Subordinated debt/total debt

0.043 (0.3) 0.001 (0.3) 0.026 (3.6)nnn 0.040 (5.8)nnn 0.026 (2.6)nnn 0.007 (0.8) 806 0.15

0.270 (3.6)nnn 0.029 (3.2)nnn 0.015 (1.3) 0.014 (1.2) 0.034 (2.1)nn 0.006 (0.5) 792 0.03

0.213 (1.9)n 0.011 (0.9) 0.019 (1.1) 0.028 (1.7)n 0.088 (3.8)nnn 0.062 (3.0)nnn 746 0.10

0.765 (4.5)nnn 0.031 (1.5) 0.019 (0.7) 0.043 (1.7)n 0.086 (2.3)nn 0.077 (2.4)nn 784 0.02

The t-statistics are in parentheses. The reported R2 is the R2 from the mean-deviated regression. nnn Significant at the 1% level. nn Significant at the 5% level. n Significant at the 10% level.

subordinated debt. Secured debt and capitalized leases are generally regarded as high priority debt, whereas ordinary debt and subordinated debt are regarded as low priority debt. The independent variables in these regressions are the same as those used in previous regressions—the natural log of assets, time dummies for 1986–88 and 1989–95, and the interaction of weapons firms with the two time dummies. The regression results are contained in Table 6. They show that weapons firms significantly reduced their reliance on secured debt in the low growth periods of 1986–88 and 1989–96. These results are significant at the 0.01 level. The results also reveal that weapons firms significantly reduced their use of capitalized leases during the 1986-88 period, but not during the 1989–95 period. The proportion of debt accounted for by the sum of secured debt and capitalized leases fell significantly (at the 0.01 level) for weapons firms during both the 1986-88 and 1989-95 periods. The use of ordinary and subordinated debt by weapon firms generally increased during these two periods. Overall, the results are consistent with the prediction of Barclay and Smith that higher priority debt is less valuable for firms with low growth opportunities.

5. Conclusion The financial economics literature has increasingly focused on the relations between a firm’s growth opportunities and its financial policies. Given that the

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

57

defense industry has experienced an abrupt change in its investment opportunity set in recent years, this industry provides a natural experiment for empirically examining the longitudinal relation between growth opportunities and various corporate policy variables. Evidence in this paper supports the hypothesis that growth opportunities are an important determinant of corporate financial policies. As growth opportunities in the defense industry declined, defense firms increased their use of debt, lengthened the maturity structure of their debt, reduced their use of private debt, increased their use of public debt, and reduced their reliance on high priority debt. This industry and others that have experienced abrupt changes in growth opportunities seem to be fertile ground for studying how growth opportunities affect other variables of interest to corporate finance scholars, such as payout policy, investment policy, and governance structure.

Appendix A A list of the benchmark sample along side its corresponding defence firm is given in Table 7. Table 7 Defense firms

Benchmark firms

ALLIEDSIGNAL AMERADA HESS AMOCO AT&T ATLANTIC RICHFIELD BOEING CERIDIAN CHEVRON COASTAL COMPUTER SCIENCES DIGITAL EQUIPMENT DU PONT (E I) DE NEMOURS EASTMAN KOKAK. EATON EMERSON ELECTRIC E-SYSTEMS EXXON FMC FORD MOTOR GENCORP GENERAL DYNAMICS GENERAL ELECTRIC GENERAL MOTORS CLASS H GOODYEAR TIRE & RUBBER GRUMMAN GTE

DEERE RCA PROCTER & GAMBLE TEXACO GETTY OIL CATERPILLAR ST. REGIS PHILIP MORRIS AMERICAN BRANDS HEILEMAN (G) BREWING BAXTER INTERNATIONAL UNOCAL WEYERHAEUSER. CELANESE BRISTOL MYERS SQUIBB OAK INDUSTRIES SHELL OIL KIMBERLY-CLARK USX-U.S. STEEL GROUP BURLINGTON INDUSTRIES SQUIBB PHILIPS PETROLEUM DOW CHEMICAL INLAND STEEL INDUSTRIES DIAMOND INTERNATIONAL UNION CARBIDE

58

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

Table 7 (continued) Defense firms

Benchmark firms

HARRIS HARSCO HERCULES HEWLETT-PACKARD HONEYWELL INTERNATIONAL BUSINESS MACHINES ITT INDUSTRIES KAMAN LITTON INDUSTRIES LOCKHEED MARTIN LOGICON LORAL LTV MARTIN MARIETTA MCDONNELL DOUGLAS MOBIL MORRISON KNUDSEN MOTOROLA NORTHROP GRUMMAN PHILIPS ELECTRONICS RAYTHEON ROCKWELL ROYAL DUTCH PETROLEUM SUN TELEDYNE TENNECO TEXAS INSTRUMENTS TEXTRON TRW UNISYS UNITED INDUSTRIAL UNITED TECHNOLOGIES VARIAN ASSOCIATES WESTINGHOUSE ELECTRIC XEROX

TEKTRONIX CHAMPION SPARK PLUG DART INDUSTRIES ABBOTT LABORATORIES CROWN ZELLERBACH STANDARD OIL CHRYSLER TAPPAN NORTHWEST INDUSTRIES ANHEUSER-BUSCH BARNES ENGINEERING MANAGEMENT ASSISTANCE KAISERTECH GILLETTE BEATRICE CONOCO LEAR SIEGLER MURPHY OIL PITNEY BOWES MONSANTO PFIZER SPERRY GULF GRACE (W.R.) WARNER COMMUNICATIONS RJR NABISCO HOLDINGS PENNZOIL NORTON SIMON COLGATE-PALMOLIVE PEPSICO MOOG ASHLAND SCOTT & FETZER AVO USX-MARATHON GROUP

Weapons manufacturers are bold-faced.

References Baker, G.P., 1993. Growth, corporate policies, and the investment opportunity set. Journal of Accounting and Economics 16, 161–165. Barclay, M.J., Smith, C.W., 1995a. The maturity structure of corporate debt. Journal of Finance 50, 609–631. Barclay, M.J., Smith, C.W., 1995b. The priority structure of corporate debt. Journal of Finance 50, 899–918. Barnea, A., Haugen, R., Senbet, L., 1980. A rationale for debt maturity structure and call provisions in the agency theoretic framework. Journal of Finance 35, 1223–1243. Blackwell, D.W., Kidwell, D.S., 1988. An investigation of cost differences between public sales and private placements of debt. Journal of Financial Economics 22, 253–278.

V.K. Goyal et al. / Journal of Financial Economics 64 (2002) 35–59

59

Brick, I., Ravid, S.A., 1985. On the relevance of debt maturity structure. Journal of Finance 40, 1423–1437. Chung, K.H., Charoenwong, C., 1991. Investment options, assets in place, and the risks of stocks. Financial Management 20, 21–33. Collins, D., Kothari, S.P., 1989. An analysis of intertemporal and cross sectional determinants of earnings response coefficients. Journal of Accounting and Economics 11, 143–181. Diamond, D., 1991. Monitoring and reputation: the choice between bank loans and directly placed debt. Journal of Political Economy 99, 688–721. Diamond, D., 1993. Seniority and maturity of debt contracts. Journal of Financial Economics 33, 341–368. Flannery, M., 1986. Asymmetric information and risky debt maturity choice. Journal of Finance 41, 19–37. Gaver, J.J., Gaver, K.M., 1993. Additional evidence on the association between the investment opportunity set and corporate financing, dividend, and compensation policies. Journal of Accounting and Economics 16, 125–160. Guedes, J., Opler, T., 1996. The determinants of the maturity of corporate debt issues. Journal of Finance 51, 1809–1834. Houston, J., James, C., 1996. Bank information monopolies and the mix of private and public debt claims. Journal of Finance 51, 1863–1890. Jensen, M.C., 1986. The agency costs of free cash flow, corporate finance, and takeovers. American Economic Review 76, 323–329. Kane, A., Marcus, A., McDonald, R., 1985. Debt policy and the rate of return premium to leverage. Journal of Financial and Quantitative Analysis 20, 479–499. Krishnaswami, S., Spindt, P.A., Subramaniam, V., 1999. Information asymmetry, monitoring, and the placement structure of corporate debt. Journal of Financial Economics 51, 407–434. Myers, S., 1977. Determinants of corporate borrowing. Journal of Financial Economics 5, 147–175. Rajan, R., 1992. Insiders and outsiders: the choice between informed and arm’s-length debt. Journal of Finance 47, 1367–1400. Rajan, R.G., Zingales, L., 1995. What do we know about capital structure?: some evidence from international data. Journal of Finance 50, 1421–1460. Sharpe, S., 1990. Asymmetric information, bank lending and implicit contracts: a stylized model of customer relationships. Journal of Finance 45, 1069–1087. Smith, C.W., Warner, J., 1979. On financial contracting: an analysis of bond covenants. Journal of Financial Economics 7, 117–161. Smith, C.W., Watts, R.L., 1992. The investment opportunity set and corporate financing, dividend and compensation Policies. Journal of Financial Economics 32, 263–292. Stohs, M.H., Mauer, D.C., 1996. The determinants of corporate debt maturity structure. Journal of Business 69, 279–312. Stulz, R.M., Johnson, H., 1985. An analysis of secured debt. Journal of Financial Economics 14, 501–521.