executive compensation and the maturity structure of corporate debt.pdf

8/10/2019 Executive Compensation and the Maturity Structure of Corporate Debt.pdf

1/39

THE JOURNAL OF FINANCE VOL. LXV, NO. 3 JUNE 2010

Executive Compensation and the Maturity

Structure of Corporate Debt

PAUL BROCKMAN, XIUMIN MARTIN, and EMRE UNLU

ABSTRACT

Executive compensation influences managerial risk preferences through executivesportfolio sensitivities to changes in stock prices (delta) and stock return volatility(vega). Large deltas discourage managerial risk-taking, while large vegas encouragerisk-taking. Theory suggests that short-maturity debt mitigates agency costs of debtby constraining managerial risk preferences. We posit and find evidence of a negative

(positive) relation between CEO portfolio deltas (vegas) and short-maturity debt. Wealso find that short-maturity debt mitigates the influence of vega- and delta-relatedincentives on bond yields. Overall, our empirical evidence shows that short-term debtmitigates agency costs of debt arising from compensation risk.

THE USE OF STOCK- and option-based executive compensation has increased dra-matically during the past few decades. The median exposure of CEO wealthto stock prices tripled between 1980 and 1994 (Hall and Liebman (1998)),and then doubled again between 1994 and 2000 (Bergstresser and Philippon

(2006)). Such changes in executive compensation have a direct impact on themanagers exposure to risk, thus altering both incentives and behavior. Car-penter (2000) and Lambert, Larcker, and Verrecchia (1991) discuss two effectsof compensation on managerial incentives. One effect is caused by the sen-sitivity of compensation to stock prices (delta). A second effect is caused bythe sensitivity of compensation to stock return volatility (vega). The higherthe compensation packages sensitivity to stock prices, the weaker will be themanagers appetite for risk (Knopf, Nam, and Thornton (2002)). In contrast,the higher the compensation packages sensitivity to stock return volatility,the stronger will be the managers appetite for risk (Knopf et al. (2002) and

Coles, Daniel, and Naveen (2006)). By altering managerial risk preferences,stock-based compensation also influences third-party (e.g., creditors, suppli-ers, customers) perceptions of those risk preferences. The primary objective ofthis study is to investigate the role of short-term debt in reducing agency costsof debt arising from executive incentive contracts. Specifically, we examine theeffect of the two portfolio sensitivities on the maturity structure of corporate

Lehigh University, Washington University, and University of Nebraska. The authors thank ananonymous referee and associate editor, the editor (Campbell R. Harvey), and co-editor (John R.Graham) for their suggestions and comments during the revision process. The authors also thank

Mary McGarvey, James Schmidt, and seminar participants at Washington University, Universityof Missouri, Texas Christian University, and University of Texas-Dallas for their comments.

1123


2/39

1124 The Journal of Finance R

debt. In addition, we analyze the effect of debt maturity on the relation be-tween portfolio sensitivities and bond yields. The empirical results provide aconsistent picture that short-term debt reduces agency costs of debt associatedwith compensation incentives.

Traditional agency theory posits a conflict of interest between shareholdersand creditors. In their seminal studies, Fama and Miller (1972) and Jensen andMeckling (1976) show that shareholders have an incentive to expropriate bond-holder wealth by substituting into riskier investments, a phenomenon com-monly referred to as asset substitution. Equity-based compensation providesmanagers with a potentially stronger motive for asset substitution. Creditorsunderstand these risk incentives and rationally price them. For example, creditrating agency reports show an awareness of the link between CEO compensa-tion and managerial risk appetites. A 2007 Moodys Investors Service SpecialComment states that Executive pay is incorporated into Moodys credit anal-

ysis of rated issuers because compensation is a determinant of managementbehavior that affects indirectly credit quality (p. 1). The report later explainsthat the primary interest in analyzing pay is to gain insight into the compen-sation committees intent regarding the structure, size and focus of incentives(p. 4). Moodys has also published the results of an internal study conductedin 2005 entitled CEO Compensation and Credit Risk. This study concludesthat pay packages that are highly sensitive to stock price and/or operatingperformance may induce greater risk taking by managers, perhaps consistentwith stockholders objectives, but not necessarily bondholders objectives (p.8). We find similar statements regarding CEO incentives and credit analysis

in Standard & Poors reports.1

Barnea, Haugen, and Senbet (1980) argue that shorter-term debt can reducemanagerial incentives to increase risk. Further, Leland and Toft (1996) claimthat short-term debt can reduce or even eliminate agency costs associated withasset substitution. 2 An important insight from Stulz (2000) is that short-termdebt provides creditors with an extremely powerful tool to monitor manage-ment. Similarly, Rajan and Winton (1995) argue that short-term debt providescreditors with additional flexibility to monitor managers with minimum effort.

Using two measures of risk preference derived from managerial compensa-tion packages in this paper, we test the role of short-term debt in mitigating

agency costs of debt arising from asset substitution. Specifically, we posit thatthe proportion of short-term debt increases in CEO compensation risk. Onemeasure of the managers appetite for risk is the sensitivity of the compen-sation package to underlying stock prices. Creditors recognize that the lowerthis sensitivity, the more likely the manager is to engage in risk-increasingstrategies. We therefore expect that the lower the managers sensitivity tostock prices, the larger the proportion of short-term debt in the firms capital

1 In their 2006 report on Corporate Ratings Criteria, Standard & Poors examines recent exam-ples of poor corporate governance, which contributed to impaired creditworthiness. They identifymanagement incentives that compromised long-term stability for short-term gain as one of four

problem areas.2 Graham and Harveys (2001) survey evidence suggests that short-term debt has little impacton CEOs willingness to accept risky projects.


3/39

Executive Compensation 1125

structure. In contrast, the managers appetite for risk increases with the sen-sitivity of the compensation package to stock return volatility. We thereforeexpect that the higher the managers sensitivity to stock return volatility, thelarger the proportion of short-term debt in the firms capital structure. Prior

studies argue and find some evidence that the cost of debt increases in manage-rial compensation risk (e.g., Daniel, Martin, and Naveen (2004), Billett, King,and Mauer (2007), and Shaw (2007)). If short maturities restrain managerialrisk-seeking, then we expect that short-term debt will attenuate the effect ofcompensation risk on the cost of debt.

We study the causal link between CEO incentive compensation and corporatedebt maturity using a sample of 6,825 firm-year observations during the 14-year period from 1992 to 2005. We employ alternative definitions of short-termdebt, follow Core and Guays (2002) method for estimating option sensitivities,3

and apply several empirical methodologies (e.g., pooled OLS and GMM simul-

taneous equation estimation, fixed-effect regressions, and change-in-variablesregressions) and an alternative new debt issuance sample to analyze the pre-dicted relations. As hypothesized, we find a negative and significant relationbetween CEO portfolio deltas and short-term debt. Also consistent with expec-tations, we find a positive and significant relation between CEO portfolio vegasand short-term debt. Taken together, these findings suggest that short-termdebt is used to reduce agency costs associated with high managerial compensa-tion risk. Our empirical results are robust to controlling for CEO stock owner-ship, leverage, asset maturity, growth opportunities, firm size, term structure,bond rating, and other issuer characteristics.

Next, we use a sample of 268,400 bond-day observations for 114 uniquefirms during the 1994 to 2005 period to examine whether short-maturity debtmitigates the impact of managerial incentives on the cost of debt. Consistentwith prior studies (e.g., Daniel et al. (2004), Shaw (2007), and Billett, Mauer,and Zhang (2009)), we find that higher deltas (vegas) lead to lower (higher)borrowing costs. More importantly, we show that short-term debt attenuatesthe negative (positive) relation between bond yields and deltas (vegas).

Our study makes several contributions to the literature. First, we provideempirical support for theory that argues that short-maturity debt mitigatesagency costs of debt related to asset substitution (Barnea et al. (1980), Leland

and Toft (1996)). Related research by Johnson (2003) finds that short-termdebt mitigates the debt overhang problem for high growth firms. Our papercomplements this finding by showing that short-term debt can also mitigateasset substitution problems for firms with high CEO compensation risk.

Our empirical results also add to the literature on the maturity structure ofcorporate debt. Barclay and Smith (1995), Guedes and Opler (1996), and Stohsand Mauer (1996), among others, find that debt maturity is determined by firmcharacteristics such as asset maturity, growth opportunities, and firm size.

3 Other studies that use the Core and Guay (2002) approximation include Knopf et al. (2002),

Rajgopal and Shevlin (2002), Graham and Rogers (2002), Billett et al. (2007), Cheng and Warfield(2005), Burns and Kedia (2006), Erickson, Hanlon, and Maydew (2006), and Hanlon et al. (2004).


4/39


Datta, Iskandar-Datta, and Raman (2005) provide evidence that managerialownership is an additional determinant of corporate debt maturity. Our studyextends this literature by showing that CEO compensation incentives alsoaffect debt maturity structure.

Next, we expand our understanding of how managerial compensation charac-teristics impact corporate capital structure. Berger, Ofek, and Yermack (1997)document that firms with weak managerial incentives avoid high levels ofleverage. Novaes and Zingales (1995) show that the optimal level of leveragefor shareholders differs from that chosen by entrenched managers. Datta et al.(2005) document that managerial stock ownership inversely affects debt matu-rity. Our study provides new evidence that the two sensitivities of managerialcompensation affect debt maturity in different ways: the sensitivity of CEOwealth to stock price increases debt maturity, whereas the sensitivity of CEOwealth to stock return volatility reduces debt maturity.

Finally, our study sheds light on creditors assessment of managerial in-centives and risk-seeking behavior. Option-based compensation is designed inpart to encourage risk-averse, underdiversified managers to invest in positivebut risky NPV projects. However, it is an open question as to whether option-based compensation might also encourage excessive risk-taking, thus aggra-vating stockholder-debtholder conflicts. John and John (1993) and Parrinoand Weisbach (1999) find evidence that suggests option-based compensationmay increase risk-taking, whereas Carpenter (2000), Ross (2004), and Hanlon,Rajgopal, and Shevlin (2004) find no such relation. Our empirical results con-firm that creditors take into consideration the impact of option-based compen-

sation on managerial risk-seeking behavior and adjust debt maturity and yieldspreads accordingly.

This paper proceeds as follows. In Section I, we review related research anddevelop our testable hypotheses. Section II discusses the sample and data. Wepresent our empirical findings in Section III, and we summarize and concludein Section IV.

I. Related Research and Hypotheses

A. Related Research

Lambert et al. (1991) suggest that measuring the partial derivative (or, sen-sitivity) of the change in managerial compensation with respect to a changein a performance variable is the preferred way to assess managerial incen-tives. Recent empirical research examines the relation between CEO portfoliosensitivities and the firms investment decisions. Guay (1999) finds that thesensitivity of CEO portfolios to stock return volatility is positively related tothe firms growth opportunities. Similarly, Rajgopal and Shevlin (2002) findthat greater sensitivity of CEO options to stock return volatility is relatedto greater exploration risk and less risk hedging for a sample of oil and gas

industry firms. Further, Hanlon et al. (2004) show that firms with CEO op-tion portfolios that are more sensitive to stock return volatility have greater


5/39


1-year-ahead stock return volatility, though such effects are economically small.Coles et al. (2006) endogenize CEO compensation and a firms financial andinvestment policies and find that greater sensitivity of CEO wealth to stockreturn volatility is associated with higher leverage, larger research and de-

velopment expenditures, and less capital expenditures. Taken together, thesefindings support the claim that option-based sensitivities exert significant in-fluence on managerial decision-making.

Knopf et al. (2002) disentangle the opposing effects of portfolio sensitivitieson the managers appetite for risk. They argue that the sensitivity to stockreturn volatility should,ceteris paribus, give the manager an incentive to takemore risk (p. 801), while the sensitivity to stock price should give a risk-averse manager an incentive to avoid risk (p. 802). They find that managerswith high portfolio sensitivities to stock return volatility tend to hedge lesswith derivative securities than managers with relatively low sensitivities. This

finding supports the vega effect of executive compensation. They also show thatmanagers with high portfolio sensitivities to stock prices (i.e., deltas) tend tohedge more with derivative securities than managers with low sensitivities.

There is considerable evidence that creditors understand the effect of priceand volatility sensitivities on the CEOs risk-seeking behavior. Billett et al.(2009) find that when firms announce new CEO stock option grants, stock-holders experience positive abnormal returns while bondholders experiencenegative abnormal returns. Bondholder returns are more negative with highvega stock option grants and less negative with high delta option grants. Danielet al. (2004) examine the impact of the CEOs delta and vega on the firms cost

of capital. They conclude that these CEO incentives affect the firms cost of debtbecause the bond markets understand and account for the effect of incentiveson risk-taking (p. 5).

Barnea et al. (1980) argue that with rational expectations, debt-holdersrecognize the entrepreneurs incentive problem, and discount debt valueaccordingly. . . . In the absence of mechanisms which resolve the incentive prob-lem, the entrepreneur incurs the agency costs (p. 1229). The authors note thatthe risk incentive problem can be solved by shortening debt maturity. Lelandand Toft (1996) similarly argue that short-maturity debt reduces or eliminatesasset substitution agency costs. Short-term debt has also been shown to reduce

agency costs of managerial discretion by subjecting managers to more frequentmonitoring by underwriters, investors, and rating agencies (Rajan and Winton(1995) and Stulz (2000)).

In this study, we follow the Knopf et al. (2002) framework and extend theabove literature by examining the impact of both dimensions of compensationrisk (i.e., vega and delta effects) on the firms debt maturity. Similar to Johnson(2003), we argue that short-term debt can be used to mitigate bondholder-shareholder conflicts. However, whereas Johnson examines the underinvest-ment agency problem caused by growth opportunities, we examine the as-set substitution agency problem caused by the incentive structure of execu-

tive compensation. Without the use of short-term debt, managers with highvega (risk-seeking) incentives would bear higher agency costs and find it more


6/39


difficult to access the credit markets. Berlin (2006) summarizes this relationbetween risk and maturity as follows (p. 4): For some very risky firms, lendersare simply unwilling to lend long term because lenders will lose money toooften . . . As a result, lenders will provide only very short-term financing for

such firms to keep them on a short leash. Like a short leash, short maturitiescan mitigate principal-agent problems.

B. Hypotheses

Although linking managerial compensation to firm performance by manage-rial stock and option ownership reduces ownermanager conflicts of interest,it may exacerbate ownercreditor conflicts. Previous empirical studies suggestthat compensation contracts have two opposing effects on managerial incen-tives. The first effect is the sensitivity of the compensation package to move-

ments in the underlying stock price. Tying a managers wealth to the firmsstock price decreases a risk-averse managers appetite for risk. The second ef-fect is the sensitivity of the compensation package to stock return volatility.Due to the convex payoff structure of options, the value of a managers stockoption portfolio increases with the volatility of the firms stock returns. Thus,tying managerial wealth to stock returns increases a risk-averse managersappetite for risk.

Previous theoretical research also argues that asset substitution problemscan be mitigated through the use of short-maturity debt since the value ofshorter-term debt is less sensitive to shifts from low to high variance projects.

Building on these arguments and previous empirical evidence, we posit thatthe maturity structure of debt can be used to mitigate compensation-relatedagency costs of debt for firms with high managerial compensation risk.4 Ourfirst hypotheses can be stated as follows:

H1a: The proportion of short-term debt is negatively related to the sensitivityof the CEOs portfolio to stock prices (delta).

H1b: The proportion of short-term debt is positively related to the sensitivityof the CEOs portfolio to stock return volatility (vega).

Prior studies posit and find evidence of a positive relation between manage-rial risk seeking behavior and the cost of debt (Barnea et al. (1980), Danielet al. (2004), Billett et al. (2007), and Shaw (2007)). If short-maturity debt dis-courages risk-tolerant managers from shifting toward risky projects and allowscreditors to monitor managers with less effort, then we expect short-term debt

4 Short-maturity debt also imposes costs on risk-averse managers for at least two reasons. First,self-interested managers have an inherent preference for less monitoring and, as discussed above,short-term debt is an efficient tool for the creditor to monitor the manager. Second, the cost ofrolling over short-term debt is greater than the cost of issuing long-term debt. The higher costs ofshort-term debt potentially include higher out-of-pocket flotation costs, greater opportunity costs

of management time in dealing with more frequent debt issues, reinvestment risk, and potentialcosts of illiquidity.


7/39


8/39


B. Variable Descriptions

B.1. Debt Maturity Regression

B.1.1. Dependent Variables: Debt Maturity Structure. We use two measures for

debt maturity structure. Our first variable, ST3, measures the proportion oftotal debt maturing in 3 years or less. Alternatively, we measure the maturitystructure of debt as the proportion of debt maturing in 5 years or less (denotedbyST5).9 These two measures are consistent with prior literature.10

B.1.2. Treatment Variables: CEO Portfolio Sensitivities. We define the CEOsportfolio price sensitivity (PRCSEN) as the change in the value of the CEOsstock and option portfolio in response to a 1% increase in the price of thefirms common stock. Volatility sensitivity (VOLSEN) is similarly defined asthe change in the value of the CEOs option portfolio due to a 1% increase in theannualized standard deviation of the firms stock return.11 Partial derivativesof the option price with respect to stock price (delta) and stock return volatility(vega) are based on the Black and Scholes (1973) option pricing model adjustedfor dividends by Merton (1973). We follow Guay (1999) and Core and Guay(2002) in calculating vega and delta, consistent with recent papers including

Yermack (1995), Hall and Liebman (1998), Aggarwal and Samwick (2006),Core and Guay (2002), Guay (1999), Cohen, Hall, and Viceira (2000), Datta etal. (2005), and Rajgopal and Shevlin (2002). We discuss the derivation of deltaand vega in more detail in Appendix A.

B.1.3. Control Variables. We choose the control variables based on previous

debt maturity literature. Earlier studies analyze the relation between debtmaturity and firm size (LSIZEin logs), the square of firm size (LSIZE2),leverage (LEVERAGE), asset maturity (ASSET MAT), managerial ownership(OWN), market-to-book (M/B), term structure of interest rates (TERM), ab-normal earnings (ABNEARN), asset return standard deviations (STD DEV),and dummy variables for firms with S&P credit ratings (RATE DUM), firmswith a high Altman (1977) Z-score (ZSCORE DUM), and firms from regu-lated industries (REG DUM). We provide more detailed definitions and datasources for all variables in the maturity regression in Appendix B. We alsoprovide motivations for each of the right-hand-side variables in a supple-

mental Internet Appendix, available online at the Journal of Financewebsite(http://www.afajof.org/supplements.asp).

9 COMPUSTAT contains 63 (250) firm-year observations in which ST3(ST5) exceeds one. Wedelete these observations from our data set. As a robustness check, we also set each of theseobservations to one and rerun all empirical analyses. The results are similar to those reportedherein.

10 Our definitions are consistent with Johnson (2003). Other studies (e.g., Barclay and Smith(1995) and Datta et al. (2005)) use the complements of the maturity measures (i.e., 1-ST3 or1-ST5).

11 Guays (1999) findings suggest that the CEOs combined stock and option portfolio vega canbe precisely estimated using the option portfolio vega since the stock portfolio vega is relativelysmall. More specifically, Guay (1999) reports that the median CEO option portfolio vega ($20,915)

is 10,457 times larger than the median CEO stock portfolio vega ($2). Other papers that use thesame approximation include Coles et al. (2006), Hanlon et al. (2004), Knopf et al. (2002), andRajgopal and Shevlin (2002).


9/39


B.1.4. Sample Distribution and Summary Statistics. We estimate the matu-rity regression based on the debt maturity sample. We present the debt matu-rity sample distribution in Table I. The full sample includes 6,825 observationsover the 1992 to 2005 period. Panel A presents the time series distribution of

ST3(proportion of total debt maturing in 3 years or less), ST5 (proportion oftotal debt maturing in 5 years or less),LEVERAGE(long-term debt divided bytotal assets),PRCSEN(change in value of CEOs portfolio due to a 1% increasein stock price), and VOLSEN(change in value of CEOs portfolio due to a 1%increase in stock return volatility); Panel B presents the cross-sectional distri-butions of the same variables by industry. The industry breakdown is based ontwo-digit SIC codes.

Our debt maturity measures (includingST3 and ST5) andLEVERAGE haveremained relatively stable over the course of our sample period (Panel A). Thereis an upward trend in the use of short-term debt from 1992 until 2000, followed

by a general decline. Both ST3 and ST5 reach their highest median valuesof 43.2% and 66.6%, respectively, at the top of the bull market in 2000. ST3obtains its lowest median value of 28.5% in 2005, while ST5 obtains its lowestmedian value of 46.6% in 1992.

In contrast to the gradual rise and fall of short-term debt usage over oursample period, CEO sensitivities display a strong secular trend. The mediansensitivity of CEO portfolios to a 1% increase in stock price (PRCSEN) increasesfrom 1.111 in 1992 to 2.960 in 2005. This result suggests that compensationcommittees have strengthened the connection between CEO wealth and share-holder value. As described in our hypothesis development section, an increase

in PRCSENis expected to reduce the CEOs appetite for risk (all else equal).The median sensitivity of CEO portfolios to a 1% increase in stock returnvolatility (VOLSEN) rises from 0.109 in 1992 to 0.748 in 2005. An increase inVOLSEN is expected to increase the CEOs appetite for risk (all else equal),exacerbating conflicts of interest between owners and creditors.

Panel B exhibits the cross-sectional variation (by industry) in short-termdebt usage and portfolio sensitivities. The number of firms in each industryranges from a low of five (tobacco products industry) to a high of 883 (electric,gas, and sanitary services industry). There is considerable variation in the useof short-term debt across these industrial categories.

In Table II, we present summary statistics for our dependent and right-hand-side variables in the maturity regression. The dependent variable ST3 has amean value of 39.8%, which is similar to the reported mean of 40% in Table Iof Datta et al. (2005). The second dependent variable, ST5, has a mean valueof 58.3%. Turning to the treatment variables, PRCSENhas a mean of 6.914andVOLSENhas a mean of 1.108. Because the distributions ofPRCSENandVOLSENare right-skewed, we use natural logarithm transformations in ourempirical tests.12 The statistics for both treatment variables are similar tothose reported in Table I of Coles et al. (2006). The summary statistics for our

12

More specifically, we defineLPRCSENas ln(1+PRCSEN) andLVOLSENas ln(1+VOLSEN).We add one to each sensitivity measure sinceLPRCSENand LVOLSENcan be zero.


10/39


TableI

Sample

Distribution

Thistableshowsthetimeseries(PanelA)andindustrial(PanelB)di

stributionoftheproportionofshort-termdebtandthepricesensitivitiesofthe

CEOsoptionportfolio.Thesamplecontains6,825observationsandcoversthe1992to2005period.AllvariablesaredefinedinAppendixB.

PanelA:MediansbyFiscalYear

Yea

r

N

ST3

ST5

PRCSEN

VOLSEN

LEVERAGE

199

2

135

0.309

0.466

1.111

0.109

0.113

199

3

398

0.320

0.489

1.044

0.141

0.131

199

4

516

0.347

0.549

0.991

0.127

0.141

199

5

533

0.303

0.547

1.183

0.190

0.136

199

6

557

0.287

0.557

1.491

0.217

0.130

199

7

559

0.292

0.568

1.870

0.285

0.120

199

8

489

0.306

0.574

1.690

0.417

0.135

199

9

463

0.358

0.593

1.906

0.447

0.136

200

0

469

0.432

0.666

2.076

0.547

0.119

200

1

481

0.354

0.604

2.610

0.709

0.137

200

2

542

0.315

0.548

2.043

0.802

0.156

200

3

566

0.307

0.541

2.758

0.871

0.143

200

4

598

0.317

0.547

3.061

0.837

0.124

200

5

519

0.285

0.522

2.960

0.748

0.113

PanelB:M

ediansbyIndustry

Ind

ustry

2-digitSIC

N

ST3

S

T5

PRCSEN

VOLSEN

Foo

dandkindredproducts

20

258

0.335

0.553

4.239

0.432

Tob

accoproducts

21

5

0.339

0.541

2.496

1.253

Tex

tilemillproducts

22

69

0.316

0.588

1.031

0.107

Apparelandproductsmadefromfab

rics

23

93

0.512

0.701

1.603

0.208

Lumberandwoodproducts

24

52

0.249

0.429

3.271

0.589

Furnitureandfixtures

25

74

0.276

0.568

2.418

0.525

Pap

erandalliedproducts

26

192

0.285

0.464

1.374

0.436

Printing,publishing,andalliedindu

stries

27

214

0.322

0.648

2.618

0.564

(continued)


11/39


TableIContinued

Sample

Distribution

PanelB:M

ediansbyIndustry

Ind

ustry

2-digitSIC

N

ST3

ST5

PRCSEN

VOLSEN

Chemicalsandalliedproducts

28

731

0.381

0.659

2.

805

0.628

Pet

roleumrefiningandrelatedindustries

29

100

0.228

0.393

4.

151

1.121

Rubberandmiscellaneousplasticsp

roducts

30

100

0.354

0.656

1.

932

0.280

Lea

therandleatherproducts

31

38

0.526

0.887

0.

407

0.117

Stone,clay,glass,andconcreteproducts

32

53

0.248

0.659

1.

266

0.223

Primarymetalindustries

33

197

0.245

0.524

0.

984

0.238

Fab

ricatedmetalproducts

34

172

0.352

0.616

1.

259

0.306

Machineryandcomputerequipment

35

562

0.440

0.706

1.

809

0.478

Electronicandotherelectricalequipment

36

565

0.396

0.742

2.

260

0.421

Tra

nsportationequipment

37

227

0.340

0.529

1.

946

0.408

Measuring,analyzing,andcontrollin

ginstruments

38

478

0.449

0.850

2.

996

0.515

Mis

cellaneousmanufacturingindustries

39

73

0.418

0.634

3.

647

0.564

Railroadtransportation

40

40

0.170

0.335

4.

972

1.953

Hig

hwaypassengertransportation

41

6

0.179

0.977

0.

614

0.172

Motorfreighttransportationandwa

rehousing

42

74

0.411

0.756

2.

845

0.172

Watertransportation

44

36

0.169

0.415

1.

229

0.557

Tra

nsportationbyair

45

96

0.264

0.426

1.

542

0.409

Tra

nsportationservices

47

21

0.590

1.000

14

.

130

0.439

Com

munications

48

190

0.272

0.415

3.

604

1.134

Electric,gas,andsanitaryservices

49

883

0.256

0.372

0.

436

0.151

Wh

olesaletrade:durablegoods

50

237

0.342

0.672

1.

511

0.325

Wh

olesaletrade:non-durablegoods

51

115

0.349

0.530

1.

522

0.363

Buildingmaterialsandhardware

52

41

0.158

0.785

7.

522

0.155

Generalmerchandisestores

53

147

0.262

0.453

5.

351

0.844

Foo

dstores

54

78

0.254

0.526

1.

624

0.328

Automotivedealersandgasolineservicestations

55

60

0.595

0.836

2.

993

0.224

Apparelandaccessorystores

56

149

0.309

0.538

2.

199

0.412

Homefurnitureandfurnishings

57

80

0.396

0.619

3.

166

0.775

Eatinganddrinkingplaces

58

161

0.366

0.714

1.

939

0.342

Mis

cellaneousretail

59

158

0.389

0.629

3.

144

0.556


12/39


Table II

Summary StatisticsThis table shows the summary statistics for our dependent and right-hand-side variables in thematurity regression. The sample contains 6,825 observations and covers the 1992 to 2005 period.

All variables are defined in Appendix B.

Variable Mean Std. Dev. Min. 25% 50% 75% Max.

ST3 0.398 0.308 0.000 0.157 0.321 0.589 1.000ST5 0.583 0.306 0.000 0.347 0.560 0.888 1.000PRCSEN 6.914 17.160 0.017 0.689 1.894 5.487 132.340VOLSEN 1.108 1.913 0.000 0.112 0.408 1.158 11.446SIZE 10,182.7 26,408.9 25.095 916.6 2,414.1 7,632.2 449,110.2LEVERAGE 0.154 0.124 0.000 0.052 0.132 0.236 0.522ASSET MAT 11.619 10.280 0.769 4.395 7.846 15.394 45.425OWN 0.024 0.057 0.000 0.001 0.003 0.014 0.328M/B 1.888 1.146 0.803 1.193 1.508 2.133 7.593

TERM 1.535 1.137 0.440 0.570 1.360 2.550 3.550REG DUM 0.114 0.317 0.000 0.000 0.000 0.000 1.000ABNEARN 0.006 0.100 0.432 0.012 0.006 0.023 0.512STD RET 0.064 0.046 0.011 0.034 0.051 0.080 0.260RATE DUM 0.604 0.489 0.000 0.000 1.000 1.000 1.000ZSCORE DUM 0.861 0.346 0.000 1.000 1.000 1.000 1.000

control variables are consistent with those reported in Johnson (2003), Dattaet al. (2005), and Billett et al. (2007), among others.

B.2. Cost of Debt Regression

B.2.1. Dependent Variable: Cost of Debt. In our cost of debt regression, weuse the yield spread (SPREAD), measured as the daily difference between thecorporate bonds daily yield-to-maturity and the linearly interpolated bench-mark Treasury bond yield (BENCHMARK TREAS), as our dependent variable.Benchmark Treasury yields are based on 1-, 2-, 3-, 5-, 7-, 10-, 20-, and 30-yearconstant maturity series.13

B.2.2. Treatment and Control Variables. Our main variables of interest are theinteraction terms between managerial incentives (LPRCSENandLVOLSEN)and maturity (LMAT). We defineLMATas the natural logarithm of the bondsmaturity, in years. We match the previous fiscal years information for alltreatment and control variables with current bond spreads to ensure that theCEOs portfolio sensitivities (as well as the accounting information) are knownby the bond market at the time that borrowing costs are determined.

We also control for several variables that have been shown by previous liter-ature to influence yield spreads at both the firm level and the aggregate level.

13Yields for 30-year Treasury bonds are unavailable for the 02/19/2004 to 02/08/2006 period.For this period, we use the Treasurys extrapolation factors added to the 20-year yields to derive

the 30-year yields. The data can be found at http://www.treas.gov/offices/domestic-finance/debt-management/interest-rate/yield historical huge.shtm.


13/39


Specifically, we control for stock volatility (STD RET), stock return (AVG RET),bond rating (RATING), profitability (ROS, or return on sales), leverage(LEVERAGE), interest coverage (INTCOVERAGE), coupon rate (COUPON),bond illiquidity (ILLIQUIDITY), bond issue size (ISSUE SIZE), Treasury rate

(BENCHMARK TREAS), yield curve slope (YLDCRV SLOPE), and the Euro-Treasury spread (EURO TREAS SPREAD). We provide more detailed defini-tions and data sources for all variables in the cost of debt regression in AppendixB, and we discuss the motivation for each of the right-hand-side variables inthe supplemental Internet Appendix.

B.2.3. Sample Distribution and Summary Statistics. Table III reports sum-mary statistics for the variables used in the cost of debt regression based onthe bond yield sample. Panel A shows generally increasing yield spreads acrosseach risk category, moving from AAA-rated bonds in the far left column to CCC-rated bonds in the far right column.14 We find monotonically increasing yieldspreads within each risk category, moving from short-term to long-term maturi-ties. In Panel B, we present means and standard deviations, as well as variablevalues at the minimum, maximum, 5th, 25th, 50th, 75th, and 95th percentiles.We find considerable variation for the dependent variable, SPREAD, as well asfor the independent variables. Bond maturity (MAT), for example, ranges from7.718 years at the 25th percentile to 21.182 years at the 75th percentile.

III. Estimation Methods and Empirical Results

A. Debt Maturity Regression

A.1. Pooled Cross-sectional, Time-Series Analysis

We estimate the following pooled cross-sectional, times-series regression us-ing the debt maturity sample as follows:

ST3i,t(ST5i,t) = 0 + 1LPRCSENi,t + 2LVOLSENi,t + 3LSIZEi,t+4LSIZE2i,t + 5LEVERAGEi,t + 6ASSET MATi,t+7OWNi,t + 8M/Bi,t + 9TERMi,t + 10REG DUMi,t

+11ABNEARNi,t

+12STD RETi,t

+13RATE DUMi,t

+14ZSCORE DUMi,t + i,t, (1)

where all variables are defined in Appendix B.In Table IV, we report the empirical results from pooled regression model (1).

We use Rogers (1993) industry-year clustered standard errors in assessing sta-tistical significance. According to hypothesis H1a, we expect a negative relationbetween the use of short-term debt (ST3) and the CEOs portfolio sensitivityto stock prices (LPRCSEN). Our regression results support this hypothesis

14

The only exception, the average spread for short-term BB-rated bonds (2.923%), is slightlyhigher than the average spread for short-term B-rated bonds (2.826%).


14/39


TableIII

AverageCorporateBondYie

ldSpreadsandSummaryStatistics

Thistableshowsthesummarystatis

ticsforthebondpricingsample.

Thesamplecontains268,400day

-bondobservationsbasedonthe

1994to2005

period.AllvariablesaredefinedinA

ppendixB.

P

anelA:AverageCorporateBondYieldSpread(%)BrokenDownbyMaturity

Maturity

AAA

AA

A

BBB

BB

B

CCC

Sho

rt(17years)

0.359

0.

533

0.

644

1.

016

2.

923

2.

826

5.

671

Medium(715years)

0.534

0.

730

1.

006

1.

546

3.

111

4.

483

7.

212

Lon

g(1530years)

0.612

0.

829

1.

125

1.

725

3.

258

4.

810

8.

659

PanelB:S

ummaryStatistics

Var

iable

Mean

Std.Dev.

Min.

5%

25%

5

0%

75%

95%

Max

SPREAD

1.64

0

1.

750

0.

312

0.

465

0.

783

1

.

086

1.

688

4.

720

11

.

651

MA

T

14.68

2

7.

385

2.

386

3.

814

7.

718

15

.

211

21

.

182

25

.

959

28

.

116

LM

AT

2.52

0

0.

627

0.

870

1.

339

2.

044

2

.

722

3.

053

3.

257

3.

336

LPRCSEN

1.99

8

0.

933

0.

251

0.

608

1.

315

1

.

843

2.

672

3.

499

4.

966

LVOLSEN

1.34

3

0.

707

0.

000

0.

253

0.

862

1

.

298

1.

824

2.

521

3.

141

STDRET

1.82

2

0.

815

0.

765

0.

918

1.

255

1

.

595

2.

200

3.

302

5.

572

AVGRET

0.06

0

0.

150

0

.

396

0

.

184

0

.

027

0

.

058

0.

145

0.

309

0.

512

RATING

12.43

1

3.

328

2.

000

5.

500

11

.

000

14

.

000

14

.

000

17

.

000

19

.

000

ROS

0.15

5

0.

087

0

.

024

0.

046

0.

096

0

.

136

0.

202

0.

317

0.

422

LEVERAGE

0.18

9

0.

111

0.

012

0.

045

0.

105

0

.

166

0.

264

0.

406

0.

477

INT

COVERAGE

1.89

5

0.

706

0.

000

0.

745

1.

463

1

.

843

2.

294

3.

055

3.

850

COUPON

7.66

6

1.

321

3.

800

5.

400

6.

875

7

.

625

8.

750

9.

700

10

.

200

ILL

IQUIDITY

0.16

1

0.

258

0.

023

0.

031

0.

039

0

.

047

0.

070

0.

844

1.

000

ISS

UESIZE

5.33

5

0.

628

3.

219

4.

605

5.

011

5

.

298

5.

704

6.

215

6.

908

BENCHMARKTREAS

4.72

7

0.

816

2.

873

3.

325

4.

250

4

.

732

5.

149

6.

260

7.

010

YLDCRVSLOPE

1.20

8

0.

965

0

.

260

0

.

090

0.

190

1

.

310

2.

160

2.

480

2.

640

EUROTREASSPREAD

0.21

9

0.

138

0.

030

0.

060

0.

110

0

.

180

0.

310

0.

490

0.

700


15/39


Table IV

Relation between Debt Maturity and CEO PortfolioPrice/Volatility Sensitivities

This table shows the pooled regression results for two specifications. In the first (second) specifi-

cation, the dependent variable is ST3 (ST5). The sample contains 6,825 observations and coversthe 1992 to 2005 period. All variables are defined in Appendix B. Statistical significance is basedon Rogers (1993) industry-year clustered standard errors. ,, and denote significance at the1%, 5%, and 10% levels, respectively.

Dependent Variables

ST3 ST5

Independent Predicted Coefficient CoefficientVariables Signs Estimate p-value Estimate p-value

Intercept 1.4281 0.000 1.2059 0.000

LPRCSEN 0.0385 0.000 0.0179 0.012LVOLSEN + 0.0309 0.001 0.0277 0.003LSIZE 0.1558 0.000 0.0777 0.000LSIZE2 + 0.0090 0.000 0.0041 0.001LEVERAGE 1.1068 0.000 0.6251 0.000ASSET MAT 0.0026 0.000 0.0036 0.000OWN + 0.4562 0.000 0.1567 0.093M/B + 0.0021 0.714 0.0044 0.382TERM 0.0014 0.699 0.0049 0.209REG DUM 0.0108 0.429 0.0501 0.001ABNEARN + 0.0171 0.648 0.0253 0.500STD RET + 0.0163 0.889 0.2232 0.036RATE DUM

0.0945 0.000

0.1413 0.000

ZSCORE DUM 0.1194 0.000 0.047 0.000R2adj 0.251 0.226N 6,825 6,825

by showing that LPRCSENs estimated coefficient (0.0385) is negative andhighly significant. According to H1b, we expect a positive relation between theuse of short-term debt (ST3) and the CEOs portfolio sensitivity to stock returnvolatility (LVOLSEN). Again, the regression results support this hypothesis byshowing thatLVOLSENs estimated coefficient (0.0309) is positive and highly

significant.UsingST5 as the dependent variable, we find that the estimated coefficient

onLPRCSEN(0.0179) is negative and significant. This finding suggests thatlonger-term maturities are more likely to be chosen when managers incen-tives are aligned with creditors. We also find that the estimated coefficient on

LVOLSEN(0.0277) is positive and significant.Following Datta et al. (2005), we also estimate a firm fixed-effects model.

The results, which are reported in the supplemental Internet Appendix,15 aresimilar to those reported in Table IV. Specifically, the LPRCSENcoefficients

15 We also present other untabulated results in the supplemental Internet Appendix.


16/39


are negative and statistically significant, and the LVOLSENcoefficients arepositive and statistically significant. Taken together, the evidence shows thatwhen the ex ante incentive of managers to substitute risky assets for safeassets is high (i.e., high portfolio vegas), short-maturity debt is more likely to

be chosen to mitigate bondholder-shareholder agency conflicts.16Besides the main variables of interest, LPRCSEN and LVOLSEN, our

Table IV pooled regressions also yield consistent results for the control vari-ables. Most of the control variables are statistically significant and display theexpected sign based on previous studies (e.g., Datta et al. (2005)). Specifically,the estimated coefficients on LSIZE, LEVERAGE, ASSET MAT, REG DUM(ST5regression only),RATE DUM, andZSCORE DUMare negative and sta-tistically significant, consistent with expectations and previous findings. Theestimated coefficients on LSIZE2,OWN, andSTD RET(ST5regression only)are positive and significant, also consistent with expectations and previous

findings. We obtain insignificant results forM/B,TERM,REG DUM(ST3re-gression only),ABNEARN, andSTD RET(ST3regression only). Overall, ourpooled regression results explain between 22.6% and 25.1% of the variation inshort-term debt.

We also evaluate the economic significance of our findings. Our Table IV slopeestimates onLPRCSENand LVOLSENare 0.0385 and 0.0309, respectively.When price sensitivity changes from the 50th percentile to the 95th percentile,the firms use of short-term debt (ST3) decreases by 8.7%. Thus, for a firm withroughly 40% of its total debt in short-term maturities (the sample mean inTable II), a change in LPRCSENfrom the 50th to the 95th percentile would re-

duce this firms short-term component from 40% to 31.3% of total firm debt. Thesame computation forLVOLSENshows an increase of 4.4% in the firms use ofshort-term debt (ST3), implying an increase in the short-term component from40% to 44.4% of total firm debt.17 We conclude that there is an economicallysignificant relation between portfolio sensitivities and the maturity structureof corporate debt.

A.2. Generalized Method of Moments (GMM) Estimation

If debt maturity and leverage are jointly determined, then ordinary least

squares estimation can lead to a biased leverage coefficient. To address thisconcern, in this section we estimate a system that models leverage and debt

16 We also investigate the effect of the interaction LPRCSEN LVOLSENon debt maturity.Our results show that the estimated coefficient on this interaction term is negative (0.0163)and significant when ST3 is the dependent variable. We find similar results using ST5 as thedependent variable. The partial derivative ofST3with respect toLVOLSENis 0.0681 0.0163 LPRCSEN, and the partial derivative ofST3 with respect to LPRCSEN is0.0296 0.0163LVOLSEN. The economic interpretation is that vega effects are mainly driven by CEOs withlow delta compensation packages, and delta effects are mainly driven by CEOs with high vegacompensation packages.

17

Similar computations for ST5 show a decrease (increase) of 4.0% (3.9%) when LPRCSEN(LVOLSEN) increases from the 50th percentile to the 95th percentile of the sample distribution.


17/39


maturity as jointly endogenous. Our two-equation system is specified as follows:

LEVERAGEi,t= 0 + 1ST3i,t(ST5i,t) + 2LPRCSENi,t + 3LVOLSENi,t

+4LSIZEi,t

+5OWNi,t

+6M/Bi,t

+7REG DUMi,t

+ 8ABNEARNi,t + 9STD RETi,t + 10FIX ASSETi,t+ 11ROAi,t + 12NOL DUMi,t + 13ITC DUMi,t + i,t (2)

ST3i,t(ST5i,t) = 0 + 1LPRCSENi,t + 2LVOLSENi,t + 3LSIZEi,t+4LSIZE2i,t + 5LEVERAGEi,t + 6ASSET MATi,t+7OWNi,t + 8M/Bi,t + 9TERMi,t + 10REG DUMi,t+ a11ABNEARNi,t + 12STD RETi,t + 13RATE DUMi,t+14ZSCORE DUMi,t + i,t, (3)

where dependent and right-hand-side variables are as defined in Appendix B.We rely on earlier theoretical studies to guide our selection of right-hand-side

variables in our simultaneous equations. Theoretical capital structure studiesshow that the fixed asset ratio (FIX ASSET), profitability measure (ROA), andexpected marginal tax rate (NOL DUMand ITC DUM) are important deter-minants of leverage (see Johnson (2003), Barclay, Marx, and Smith (2003)). Incontrast, theoretical debt maturity studies do not find that fixed asset ratios(when asset maturity is controlled for), profitability measures, or the expected

marginal tax rate are important determinants of maturity. We therefore as-sume that these variables are orthogonal to the error term and restrict theircoefficients to be zero in the maturity equation.18 Further, theoretical capitalstructure studies make no claims about the square of firm size (LSIZE2), as-set maturity (ASSET MAT), term structure (TERM), the rated-firm dummy(RATE DUM), or financial distress (ZSCORE DUM). 19 We therefore treatthese variables as orthogonal to the error term and restrict their coefficients tobe zero in the leverage equation. 20

Table V reports the empirical results for the maturity equation (3). Theresults support our H1a, with a negative (0.0829) and significant coefficienton LPRCSEN. The empirical results also support our H1b, with a positive(0.0984) and significant coefficient on LVOLSEN. We obtain similar resultswhen we extend our definition of short-term debt from 3 years to 5 years. Using

18 It is important to note that our simultaneous equation estimates can be biased if this andrelated assumptions fail to hold. To address this concern, we test for and confirm the reliability ofour reported results in subsequent sections of this study.

19 Faulkender and Petersen (2006) argue that debt ratings allow firms to access the public debtmarket, thus alleviating financial constraints. They find that, indeed, firms with debt ratings havehigher leverage. For this reason, we also include a rating dummy (RATE DUM) in the leverageequation and reestimate the two-equation system using GMM. Our results remain unaltered.

20

We also estimate this system of equations using two-stage least squares (2SLS) and three-stage least squares (3SLS). Our findings remain unchanged.


18/39


Table V

Relation between Debt Maturity and CEO Portfolio Price/VolatilitySensitivities, Joint Determination of Maturity, and Leverage

This table shows results for the two-equation system allowing joint determination of maturity and

leverage based on GMM. The sample contains 6,825 observations and covers the 1992 to 2005period. For brevity, only the debt maturity equation estimations are reported. Debt maturity ismeasured as ST3 and ST5. All variables are defined in Appendix B. , ,and denote significanceat the 1%, 5%, and 10% levels, respectively.

Dependent Variables

ST3 ST5


Intercept 0.8225 0.596 2.5835 0.156

LPRCSEN 0.0829 0.000 0.0698 0.004LVOLSEN + 0.0984 0.014 0.1269 0.009LSIZE 0.4047 0.267 0.8355 0.051LSIZE2 + 0.0246 0.258 0.0505 0.049LEVERAGE 1.2759 0.000 0.3228 0.421ASSET MAT 0.0016 0.001 0.0026 0.000OWN + 1.0197 0.000 0.7855 0.008M/B + 0.0037 0.796 0.0113 0.486TERM 0.0002 0.966 0.0058 0.094REG DUM 0.0035 0.855 0.0982 0.000ABNEARN + 0.0146 0.707 0.0526 0.193STD RET + 0.0853 0.765 0.5645 0.085RATE DUM

0.1165 0.008

0.2245 0.000

ZSCORE DUM 0.0807 0.077 0.0185 0.736R2adj 0.182 0.204N 6,825 6,825

ST5 as the dependent variable, we find a negative (0.0698) and significantcoefficient onLPRCSEN, and a positive (0.1269) and significant coefficient on

LVOLSEN. Overall, these GMM results confirm the earlier pooled regressionresults and show that debt maturity structure can reduce the agency costs ofdebt.21

A.3. Change-in-Variables Analysis

We investigate the robustness of the results above with respect to variablechanges, as opposed to variable levels.22 We estimate the following pooled

21 We also use simulated marginal tax rates to control for the tax incentives of debt financing.We thank John Graham for providing simulated marginal tax rates. To maximize the sample size,we also employ the procedure outlined in Graham and Mills (2008) to fill in any missing values.Our Table IV results are robust to the inclusion of marginal tax rates.

22 Weber (2006), for example, suggests the use of change-in-variables regressions to address

endogeneity concerns. Moreover, change-in-variables regressions are useful in mitigating problemswith correlated omitted variables if such variables are time-invariant in the levels regressions.


19/39


Table VI

Relation between Debt Maturity and CEO Portfolio Price/VolatilitySensitivities: Change Regressions

This table shows the results of the change regressions. is used as a prefix to denote the change.

In the first (second) specification, the dependent variable is the change in ST3 (ST5). The samplecontains 5,513 observations and covers the 1993 to 2005 period. All variables are defined in Ap-pendix B. Statistical significance is based on Rogers (1993) year clustered standard errors. ,,and denote significance at the 1%, 5%, and 10% levels, respectively.

Dependent Variables

ST3 ST5


Intercept 0.0058 0.093 0.0014 0.530

LPRCSEN 0.0440 0.003 0.0259 0.025LVOLSEN + 0.0483 0.012 0.0247 0.090LSIZE 0.0831 0.176 0.0647 0.291LSIZE2 + 0.0018 0.650 0.0029 0.475LEVERAGE 1.1325 0.000 0.6638 0.000ASSET MAT 0.0007 0.449 0.0005 0.614OWN + 0.7247 0.007 0.5065 0.015M/B + 0.0009 0.940 0.0289 0.017TERM 0.0040 0.339 0.0073 0.020ABNEARN + 0.0226 0.377 0.0553 0.026STD RET + 0.0376 0.798 0.1341 0.302RATE DUM 0.1201 0.000 0.1929 0.000ZSCORE DUM

0.1242 0.000

0.0571 0.000

R2adj 0.095 0.063N 5,513 5,513

regression using 5,513 firm-year observations:

ST3i,t(ST5i,t) = 0 + 1LPRCSENi,t + 2LVOLSENi,t + 3LSIZEi,t+4LSIZE2i,t + 5LEVERAGEi,t + 6ASSET MATi,t

+7OWNi,t

+8M/Bi,t

+9TERMi,t

+10ABNEARNi,t + 11STD RETi,t+12RATE DUMi,t + 13ZSCORE DUMi,t + i,t, (4)

where all variables are as defined previously. Taking first differences () re-duces our sample size from 6,825 to 5,513 observations. We use the variablesvalue att2 if the value att1 is missing. We also eliminateREG DUMfromthe change-in-variables specification since it does not change for any firm in oursample. The remaining right-hand-side variables (after taking first differences)are described in Appendix B.

In Table VI, we report the empirical results from pooled regression model(4). The results support our H1a with a negative (0.0440) and significant


20/39


coefficient on LPRCSEN; the results also support our H1b by finding a pos-itive (0.0483) and significant coefficient for LVOLSEN. We obtain similarresults for changes in short-term debt with maturities up to 5 years. UsingST5as the dependent variable, we find a negative (positive) and significant

coefficient on LPRCSEN (LVOLSEN). Overall, these change-in-variablesresults confirm the earlier findings based on variable levels.

A.4. Managerial Incentives, Investment Policy, and Capital Structure

Coles et al. (2006) argue that it is important to disentangle the effects of man-agerial compensation incentives on the firms investment and financing policiesfrom the effects of these policies (and the corresponding risk profile of the firmsassets) on managerial compensation incentives. Accordingly, we follow Coleset al.s (2006) framework and examine the relations among a firms manage-

rial incentives (LPRCSENand LVOLSEN), financial policies (ST3, ST5, andLEVERAGE), and investment policies (i.e., research and development expen-ditures (RD) and net capital expenditures (CAPEX)).

In Panel A of Table VII, we report our results for the dependent variablesST3, LPRCSEN, LVOLSEN, LEVERAGE, RD, and CAPEXbased on GMMestimation. In Panel B, we use the alternative definition for short-term debt,ST5, in place ofST3. Our right-hand-side variables include the variables de-scribed above and in Appendix B, augmented by additional variables motivatedby prior literature (e.g., Coles et al. (2006)).23 The additional right-hand-sidevariables include the following: LTENURE, the logarithmic transformation of

the CEOs tenure measured in years; SURCASH, the cash from assets-in-place;EQUITY RISK, the logarithmic transformation of monthly stock return vari-ance during the fiscal year; CASHCOMP, the sum of the CEOs salary andbonus (in 100 thousands); SGR, the sales growth rate; and STOCKRET, thebuy-and-hold return during the fiscal year. We provide more detailed definitionsand data sources in Appendix B.

As Barclay et al. (2003) note, it becomes increasingly difficult to find iden-tifying independent variables for increasing numbers of structural equations.In spite of this difficulty, we continue to follow previous theoretical studies toguide our choice of structural restrictions and help identify our six-equation

system. Previous theory shows that firm size and the investment opportunityset are important determinants of a firms financial, investment, and compen-sation policies. We therefore view firm size (LSIZE) and market-to-book (M/B)as exogenous or predetermined and include them in all six equations. To thebest of our knowledge, previous theory does not posit significant relations be-tween our other instruments (in the debt maturity and leverage regressions)and compensation incentives (LPRCSENand LVOLSEN) or investment poli-cies (RDand CAPEX). We therefore treat these variables as orthogonal to theerror terms in the compensation and investment equations. We note, however,

23 See the Internet Appendix for more detailed motivations of the instrumental variables.


21/39


TableVII

R

elationbetweenDebtM

aturityandCEOPortf

olioPrice/VolatilitySen

sitivities,JointDetermination

ofMaturity,Leverage,C

ompensation,andInvestment

Thistableexaminestherobustness

oftherelationbetweendebtmaturityandmanagerialincentivesbyallowingforthejointdete

rminationof

maturity,compensation,capitalstru

cture,andinvestmentpoliciesba

sedonGMM.Thesamplecontains6,180firm-yearobservations.

Allvariables

are

definedinAppendixB.

,

,an

d

denotesignificanceatthe1%

,5%,and10%levels,respectively.

PanelA:Mat

urityMeasuredasST3

DependentVariables

(1)

(2)

(3)

(4)

(5)

(6

)

ST3

LPRCSEN

LVOLSEN

LEVERAGE

RD

CAP

EX

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

Inte

rcept

0.

986

0.000

0.

6491

0.001

0

.

5806

0.000

0.

334

0.000

0.

0679

0.000

0.

026

0.016

ST3

0

.

8341

0.000

0.

09

0.317

0

.

3495

0.000

0.

0128

0.084

0

.

0561

0.000

LPR

CSEN

0

.

156

0.000

0.

0411

0.003

0

.

0952

0.000

0

.

0064

0.000

0

.

0032

0.105

LVO

LSEN

0.

1904

0.000

0.

7465

0.000

0.

072

0.000

0.

0393

0.000

0

.

0462

0.000

LEVERAGE

0

.

8036

0.000

2

.

8914

0.000

0

.

571

0.000

0

.

0655

0.000

0

.

0623

0.000

RD

0

.

3743

0.094

6

.

5266

0.000

2.

0733

0.000

0

.

5518

0.000

CAPEX

0

.

5284

0.004

3

.

1809

0.000

1

.

3842

0.000

0.

719

0.000

LSI

ZE

0

.

0859

0.000

0.

1149

0.000

0.

1603

0.000

0.

0084

0.04

0

.

0051

0.000

0.

0067

0.000

LSI

ZE2

0.

0053

0.000

ASS

ETMAT

0

.

0022

0.000

OW

N

1.

8137

0.000

1.

1553

0.000

M/B

0.

0419

0.000

0.

2582

0.000

0

.

0614

0.000

0.

0092

0.014

0.

0112

0.000

0.

0041

0.002

TERM

0

.

0005

0.795

REGDUM

0.

0075

0.619

0.

0559

0.000

ABNEARN

0

.

0597

0.031

0

.

0017

0.919

STD

RET

0.

3006

0.046

0

.

2491

0.000

RAT

EDUM

0

.

105

0.000

ZSC

OREDUM

0

.

0385

0.003

LTE

NURE

0.

2975

0.000

0.

0014

0.011

0.

0037

0.000

SURCASH

0.

1641

0.432

0.

2953

0.000

0.

0029

0.838

(continued)


22/39


TableV

II(Continued)

PanelA:Mat

urityMeasuredasST3

DependentVariables

(1)

(2)

(3)

(4)

(5)

(6

)

ST3

LPRCSEN

LVOLSEN

LEVERAGE

RD

CAP

EX

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

EQUITYRISK

0.

1247

0.000

0.

0538

0.000

CASHCOMP

0.

019

0.000

0

.

0009

0.000

0.

0005

0.004

FIX

ASSET

0

.

1033

0.001

ROA

0

.

3731

0.000

0

.

3088

0.000

0.

1229

0.000

NOLDUM

0.

011

0.000

ITC

DUM

0

.

0068

0.036

SGR

0.

0149

0.000

0.

0028

0.343

STO

CKRET

0

.

0057

0.000

0

.

0119

0.000

PanelB:Mat

urityMeasuredasST5

DependentVariables

(1)

(2)

(3)

(4)

(5)

(6

)

ST5

LPRCSEN

LVOLSEN

LEVERAGE

RD

CAP

EX

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

Inte

rcept

0.

8778

0.000

0.

1503

0.541

0

.

7485

0.000

0.

3643

0.000

0.

0571

0.000

0.

0512

0.000

ST5

0

.

2248

0.186

0.

1799

0.034

0

.

2841

0.000

0.

0184

0.008

0

.

058

0.000

LPR

CSEN

0

.

136

0.000

0.

0412

0.003

0

.

0746

0.000

0

.

0072

0.000

0

.

0012

0.551

LVO

LSEN

0.

1514

0.000

0.

7412

0.000

0.

025

0.072

0.

0369

0.000

0

.

0389

0.000

LEVERAGE

0

.

4533

0.000

2

.

4342

0.000

0

.

4848

0.000

0

.

0637

0.000

0

.

0582

0.000

RD

0

.

571

0.014

6

.

3876

0.000

2.

2043

0.000

0

.

7144

0.000

CAPEX

0

.

7156

0.000

2

.

4907

0.000

1

.

0933

0.000

0.

9192

0.000

(continued)


23/39


TableV

II(Continued)

PanelB:Mat

urityMeasuredasST5

DependentVariables

(1)

(2)

(3)

(4)

(5)

(6

)

ST5

LPRCSEN

LVOLSEN

LEVERAGE

RD

CAP

EX

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

LSI

ZE

0

.

0265

0.

277

0.

1333

0.

000

0.

1658

0.000

0.

0127

0.002

0

.

0043

0.000

0.

0045

0.001

LSI

ZE2

0.

0019

0.

16

ASS

ETMAT

0

.

0034

0.

000

OW

N

1.

541

0.

000

0.

8728

0.000

M/B

0.

0427

0.

000

0.

25

0.

000

0

.

0678

0.000

0.

004

0.246

0.

0114

0.000

0.

0035

0.006

TERM

0

.

0044

0.

059

REGDUM

0

.

065

0.

000

0.

0339

0.000

ABNEARN

0

.

0189

0.

498

0.

0077

0.673

STD

RET

0.

5021

0.

001

0

.

3083

0.000

RAT

EDUM

0

.

1423

0.

000

ZSC

OREDUM

0.

0101

0.

503

LTE

NURE

0.

3017

0.

000

0.

0015

0.007

0.

0032

0.000

SURCASH

0

.

0213

0.

92

0.

2936

0.000

0

.

0092

0.514

EQUITYRISK

0.

1118

0.

000

0.

0464

0.000

CASHCOMP

0.

019

0.000

0

.

0008

0.000

0.

0003

0.083

FIX

ASSET

0

.

1334

0.000

ROA

0

.

4741

0.000

0

.

3142

0.000

0.

1358

0.000

NOLDUM

0.

0148

0.000

ITC

DUM

0

.

0096

0.013

SGR

0.

0156

0.000

0.

0036

0.228

STO

CKRET

0

.

0059

0.000

0

.

0109

0.000


24/39


that if these assumptions are violated, then our estimation procedure can leadto biased estimates.

The results in the first column of Panel A support our main hypotheses.24

We find that the estimated coefficient on LPRCSENis negative (

0.1560) and

significant, and the estimated coefficient on LVOLSEN is positive (0.1904)and significant. In column 4, we examine the relation between managerialincentives and financing policies. We find that the managers use of leverageis negatively (0.3495) related to delta and positively (0.0720) related to vega.Both coefficients are significant. These findings are consistent with Coles et al.(2006), who show that managers with high vega seek risky financing policies.The results in column 4 also show that the relation between short-term debtand leverage is negative (0.3495) and significant.

In columns 5 and 6, we examine the relation between managerial incen-tives and investing policies. We focus on the vega dimension of managerial

incentives since Coles et al. (2006) hypothesize a positive (negative) relationbetweenLVOLSENandRD (CAPEX); in contrast, their predictions onLPRC-SENandRD (CAPEX) are ambiguous. All else equal, higherLVOLSENshouldencourage high risk investments in R&D expenditures and discourage low-riskinvestments in fixed-asset capital expenditures. Our results support these hy-potheses. Estimated coefficients onLVOLSENare positive (0.0393) and signif-icant in theRD equation, and negative (0.0462) and significant in the CAPEXequation. More importantly, the results in Panel A suggest that even after con-trolling for the simultaneity between managerial compensation incentives anda firms investment and financing policies, our main results continue to hold:

short-maturity debt increases in vega but decreases in delta.25,26

24 We estimate our simultaneous equations using GMM. As a robustness check, we also estimate2SLS and 3SLS regressions. Our results in Table VII are robust to these alternative estimationprocedures.

25 We run additional robustness tests on our Table VII results by reducing the number of equa-tions from six to five, and then from five to three. First, we reduce the number of equations inour system from six to five by separately endogenizing the two compensation variables and thenestimate two runs of five-equation systems. Second, we further reduce the number of equations tothree by eliminating the two investment equations (RDand CAPEX) and separately endogenizethe two compensation variables, which gives two runs of three-equation systems. The resulting co-

efficients onLPRCSENand LVOLSENare consistent with both the six-equation and five-equationresults. Both coefficients maintain their hypothesized signs and are statistically significant.

26 There are two related issues that might lead to biased estimators in a six-equation system: (1)weak instrumentation, and (2) overidentification. To test for a weak identification bias, we computethe Cragg and Donald (1993) test statistic and compare it to the corresponding critical value inStock and Yogo (2004). We find that the Cragg and Donald statistic is 71.62, which is significantlyhigher than the corresponding critical value of 10.56. This result shows that our estimates areunlikely to be biased from weak instrumentation. Next, we examine potential overidentificationproblems by implementing Hansens J-test. We reject the null hypothesis that the overidentifi-cation restrictions are satisfied and conclude that our six-equation system is overidentified. Weimplement two alternative methods to deal with the potential overidentification problems. First,we reestimate our system using limited information maximum likelihood (LIML) estimation. This

method provides unbiased estimates for systems subject to overidentification problems (David-son and MacKinnon (1993), Angrist and Krueger (2001), Mariano (2001), and Greene (2003)).


25/39


A.5. New Debt Issues

Our previous results focus on the relation between CEO portfolio sensitivi-ties and current maturity structures of debt. In this section, we examine the

relation between portfolio sensitivities and the maturity of new debt issuesan incremental approach.27 This setting allows us to take the perspective of aprospective creditor who analyzes the firm characteristics that will determinethe maturity structure of new lending. Consistent with our central hypothe-sis, we expect that short-maturity debt is more (less) likely to be used when

LVOLSEN (LPRCSEN) is high.We obtain the maturity structure of new debt issues from the Securities Data

Company (SDC). Our sample includes 7,388 issues representing 873 uniquefirms over the period 1992 to 2005.28 We also construct two alternative samplesthat consolidate the original sample into a firm-year format by treating multipleissues throughout the year as a single issue. In the first consolidated sample,maturity is computed using an issue sizeweighted average maturity for firmswith multiple issues; in the second consolidated sample, maturity is computedusing an equal-weighted average maturity for firms with multiple issues. Theconsolidated samples include 3,122 firm-year observations.

We present pooled regression results for each of our three samples inTable VIII. For the unconsolidated sample, the dependent variable is the nat-ural logarithm of the maturity of the new debt issues (LMAT). In the con-solidated samples, the dependent variable is the natural logarithm of the is-sue sizeweighted maturity, LWEIGHT AVG MAT, or the natural logarithmof the equal-weighted maturity, LAVG MAT. As hypothesized, the estimatedcoefficients on LPRCSENare positive and significant across all three regres-sions, and the estimated coefficients on LVOLSEN are negative and signif-icant across all three regressions. Similar to previous results, we find thatthe CEOs portfolio sensitivities influence the maturity choice of new debtissues. The coefficients on the control variables are consistent with priorliterature.

The results from LIML estimation are similar to the results reported in Table VII. Second, weestimate a just-identified model with a single instrument for each equation. Angrist and Pischke(2009) recommend using just-identified models as an alternative to LIML estimation. Following

their approach, we identify LTENURE, CASHCOMP, STD RET, SURCASH, STOCKRET, andRATEDUM as the best instruments for LPRCSEN, LVOLSEN, LEVERAGE, RD, CAPEX, andST3, respectively. The results confirm our prior findings that the proportion of short-term debt ispositively (negatively) related to LVOLSEN (LPRCSEN) at the 1% level. Overall, these findingssuggest that our Table VI results are robust to weak instrumentation and overidentification issues.

27 Guedes and Opler (1996, p. 1810) argue that some questions about the determinants of debtmaturity are better answered by the incremental approach relative to the balance sheet approach.The use of incremental debt issues is well suited to situations in which the determinants of debtmaturity fluctuate substantially over time. In addition, the incremental approach is better ableto identify the determinants of financing choices at all points of the maturity spectrum. Otherstudies using the incremental approach include Denis and Mihov (2003), Hovakimian, Opler, andTitman (2001), and Jung, Kim, and Stulz (1995).

28

More specifically, our sample contains 355 public issues, 2,368 private (nonRule 144A) issues,642 Rule 144A issues, and 4,343 syndicated issues.


26/39


Table VIII

Relation between Debt Maturity and CEO Portfolio Price/VolatilitySensitivities: Evidence from New Debt Issues

This table shows the pooled regression results for three specifications. The first specification

is based on an unconsolidated sample of new debt issues. The second and third specificationsare based on a consolidated sample at the firm-year level. The dependent variables are LMAT,LWEIGHT AVG MAT, and LAVG MATin the first, second, and third specifications. All variablesare defined in Appendix B. Issue type fixed effects are implemented but not reported for brevity.Fixed effect dummies are constructed for public, Rule 144A, and syndicated issues relative to pri-vate (nonRule 144A) issues. Control variables are based on the previous fiscal year. Statisticalsignificance is based on Rogers (1993) industry-year clustered standard errors. , , and denotesignificance at the 1%, 5%, and 10% levels, respectively.

UnconsolidatedSample (Transaction

Level) Consolidated Sample (Firm-Year Level)

Dependent Variable:LMAT

Dependent Variable:LWEIGHT AVG MAT

Dependent Variable:LAVG MAT

Independent Coefficient Coefficient CoefficientVariables Estimate p-value Estimate p-value Estimate p-value

Intercept 2.9590 0.000 1.8058 0.000 1.9432 0.000LPRCSEN 0.1114 0.000 0.1132 0.000 0.1069 0.000LVOLSEN 0.0770 0.006 0.1300 0.000 0.1369 0.000LSIZE 0.0644 0.482 0.1407 0.082 0.0682 0.414LSIZE2 0.0003 0.947 0.0082 0.068 0.0021 0.661LEVERAGE 0.0154 0.919 0.2871 0.029 0.3062 0.019ASSET MAT 0.0045 0.002 0.0058 0.000 0.0066 0.000OWN 1.3134 0.005 1.5970 0.000 1.4757 0.000M/B 0.0113 0.485 0.0520 0.001 0.0566 0.000TERM 0.0201 0.376 0.0002 0.989 0.0023 0.875REG DUM 0.1943 0.040 0.0332 0.702 0.0374 0.688ABNEARN 0.0052 0.961 0.0300 0.799 0.0443 0.71STD RET 2.4487 0.000 2.6250 0.000 2.7106 0.000RATE DUM 0.0913 0.003 0.0803 0.011 0.0678 0.032ZSCORE DUM 0.0139 0.768 0.0627 0.175 0.0803 0.097

R2adj 0.376 0.343 0.317N 7,388 3,122 3,122

B. Cost of Debt Regression

We test H2 by estimating the following system of simultaneous equations,which allows for the joint determination of yield spreads and debt maturity:29

SPREADi,j,t,d= 0 + 1LMATi,j,t,d + 2LPRCSENi,t1+3LVOLSENi,t1 + (4LPRCSENi,t1xLMATi,j,t,d+5LVOLSENi,t1xLMATi,j,t,d) + 6STD RETi,t,d1+7AVG RETi,t,d1 + 8RATINGi,j,t,d1 + 9ROSi,t1

29 We also estimate a three-equation system allowing for joint determination of spread, maturity,and leverage. Our findings remain similar.


27/39


+10LEVERAGEi,t1 + 11INTCOVERAGEi,t1+12COUPONi,j+ 13ILLIQUIDITYi,j,t,d1

+14ISSUE SIZEi,j

+15BENCHMARK TREASi,j,t,d

+16YLDCRV SLOPEt,d + 17EURO TREAS SPREADt,d+ issuer fixed effects + i,j,t,d (5)

LMATi,j,t,d= 0 + 1SPREADi,j,t,d + 2LPRCSENi,t1 + 3LVOLSENi,t1+4STD RETi,t,d1 + 5LEVERAGEi,t1 + 6YLDCRV SLOPEt,d+7LSIZEi,t1 + 8LSIZE2i,t1 + 9ASSET MATi,t1+10OWNi,t1 + 11M/Bi,t1 + 12ABNEARNi,t1+12ZSCORE DUMi,t1 + issuer fixed effects + i,j,t,d, (6)

wherei,j,t, andd denote the ith firm andjth bond for yeart and dayd. 30 Ourmain variables of interest are the interaction terms LPRCSEN LMATand

LVOLSEN LMAT. All other variables are as defined in Appendix B.Since an endogenous variable (LMAT) is interacted with portfolio sensitivi-

ties we use nonlinear GMM.31 Under H2, we expect 4to be negative and 5tobe positive. Prior literature documents a positive relation between managerialcompensation risk and the cost of debt. We therefore expect 2 to be negativeand 3 to be positive. The coefficient 1 captures the maturity premium (i.e.,

the premium required by creditors for a unit increase in the bonds maturity)and is expected to be positive.32

We report our simultaneous equation results in Table IX based on the bondyield sample.33 In the spread equation (system 1), the coefficient on LPRCSEN

30 We also repeat our analyses on monthly and annually consolidated samples and find resultssimilar to those reported herein.

31 We obtain similar results using nonlinear 2SLS (consistent with Chen et al. (2007)) andnonlinear 3SLS.

32 Easterwood and Kadapakkam (1994) argue that agency costs of borrowing increase with debtmaturity for two reasons. First, the positive relation between equity value and asset volatility is an

increasing function of debt maturity (Barnea et al. (1980)). Second, longer-term debt increases thelikelihood that the manager will exercise the option to invest in new projects before the outstandingdebt matures (Myers (1977)).

33 In addition to credit spread models, we also estimate credit rating models using ordered probitregressions based on two samples. The first regression uses creditor ratings for existing long-termdebt (COMPUSTAT sample) and the second regression model uses credit ratings for new debtissuances (FISD sample). While the COMPUSTAT sample is larger than the FISD sample, it isalso more likely to include stale credit ratings. For each sample, we estimate the credit rating modelusing two categorical dependent variables and independent variables based on previous literature.The first dependent variable (RATINGCODE1) ranks credit ratings into 18 categories (i.e., AAA=18, . . . ,CCC = 1); the second dependent variable ranks the ratings into seven categories (AAA= 7,

AA+, AA, AA = 6, . . . ,CCC+, CCC, CCC = 1). The empirical results are robust to both samplesand both dependent variables. Consistent with expectations, we find a negative and significantrelation between LVOLSEN and credit ratings for both samples. Rating agencies assign lower


28/39


(0.1790) is negative and significant and the coefficient on LVOLSEN(0.2209)is positive and significant. Both results are consistent with prior studies sug-gesting that creditors understand the risk incentives imbedded in manage-rial compensation and rationally price these risks (Daniel et al. (2004), Billett

et al. (2007), and Shaw (2007)).34 In system 2, we include two interaction termson LVOLSEN and LPRCSENwith LMAT. Consistent with H2, we find thatthe coefficient on LVOLSEN LMAT(0.7847) is positive and significant, andthe coefficient on LPRCSEN LMAT (1.0830) is negative and significant,suggesting that short-term debt reduces agency cost of debt arising from man-agerial compensation incentives. In the maturity regression of both systems,we find a positive and significant coefficient onLPRCSENand a negative andsignificant coefficient on LVOLSEN. Both of these results are consistent withthe evidence reported in Tables IV to VIII.

The combined results in Table IX lead to the conclusion that short-term debt

mitigates agency costs of debt arising from managerial compensation incen-tives. At the same time, the results also indicate that some firms with highcompensation risk still use longer-term debt, and consequently they incur ahigher cost of debt. Although creditors attempt to steer risk-tolerant managerstoward short-term debt, creditors are also willing to issue longer-term debt torisk-tolerant managers for a commensurate risk premium. Some risk-tolerantmanagers are willing to incur the risk premium associated with longer-termdebt for reasons such as liquidity risk. For example, Diamond (1991, 1993)and Sharpe (1991) argue that too much short-maturity debt creates additionalrisks of suboptimal liquidation. This liquidity (or rollover) risk increases ex-

pected bankruptcy costs and reduces equity value. In addition, Leland and Toft(1996) argue that short-term debt can reduce managers ability to pursue riskyinvestments, and the reduction in such behavior can decrease the option valueof managerial compensation.

Our Table IX results also allow us to show how changes in debt maturityaffect the partial derivative of credit spreads with respect to LVOLSENand

LPRCSEN. Using our system 2 results, these two partial derivatives are asfollows:

SPREAD

LVOLSEN= 1.7985 + 0.7847 LMAT (7)

SPREAD

LPRCSEN= 2.5596 1.0830 LMAT. (8)

(higher) ratings when LVOLSEN is high (low). In contrast, we do not find a significant relationbetweenLPRCSENand credit ratings. In addition, we show that the estimated coefficient on theinteraction betweenST3and LVOLSENis positive and significant, while the estimated coefficienton the interaction between ST3 and LPRCSEN is insignificant.

34 In addition, coefficient estimates on LPRCSEN and LVOLSEN in the maturity equation(equation (7)) confirm our previous results in Tables IV through VIII. We find a positive and

significant relation betweenLPRCSENand maturity. Similarly, we find a negative and significantrelation betweenLVOLSENand maturity.


29/39


TableIX

RelationbetweenCorporateBondYieldSpre

adsandCEOPortfolioPrice/VolatilitySensitiv

ities

Thistableexaminesthepricingofmanagerialincentivesonbondspreads.Ourdependentvariablesare

LMATandSPREAD.Allvariable

saredefined

inA

ppendixB.

,

,and

denotesignificanceatthe1%,5%,and10

%levels,respectively.

System1

System2

DependentVariables

DependentVariables

SPREAD

LMAT

SPREAD

LM

AT

IndependentVariables

Estimate

p-value

Estimate

p-value

Estimate

p-value

Estimate

p-value

Inte

rcept

3.

7261

0.

000

4.

5772

0.

000

1.

4008

0.

002

4.

0067

0.

000

LMAT

1.

7282

0.

000

3.

2548

0.

000

SPR

EAD

0.

0724

0.

000

0.

0634

0.

000

LPR

CSEN

0

.

179

0.

000

0.

1392

0.

000

2.

5595

0.

000

0.

1384

0.

000

LVO

LSEN

0.

2209

0.

000

0

.

2000

0.

000

1

.

7985

0.

000

executive compensation and the maturity structure of corporate debt.pdf

Documents