0

Do Banks Price Independent Directors’ Attention?

Henry He Huang

Sy Syms School of Business

Yeshiva University

New York, New York 10033

Tel: (212) 960-0845, E-mail: henry.huang@yu.edu

Gerald J. Lobo

C. T. Bauer College of Business

University of Houston

Houston, Texas 77204-6021

Tel: (713) 743-4838, E-mail: gjlobo@uh.edu

Chong Wang

Von Allmen School of Accountancy

Gatton College of Business and Economics

University of Kentucky

Lexington, Kentucky 40506

Tel: (832) 651-2995, E-mail: chongwang85@uky.edu

Jian Zhou

School of Accountancy

Shidler College of Business

University of Hawaii at Manoa

Honolulu, HI 96822

Tel: (808) 956-7608, E-mail: jianzhou@hawaii.edu

1

Do Banks Price Independent Directors’ Attention?

Abstract

Masulis and Mobbs (2014a, 2014b) find that directors with multiple directorships allocate

their effort unequally based on a directorship’s relative prestige. We investigate whether

bank loan contract terms reflect such unequal allocation of monitoring effort by directors.

We find that bank loans of firms with a greater proportion of independent directors for

whom the board is among their most prestigious have lower spreads, longer maturities,

fewer covenants, lower syndicate concentration, and lower annual fees. Furthermore,

these firms also have higher bond ratings. Our evidence is consistent with directors’

attention reducing the cost of debt capital.

Keywords: multiple directorships, directors’ attention, cost of debt, bank loan contracting

JEL: G3 G12

2

Do Banks Price Independent Directors’ Attention?

I. Introduction

Bank loans comprise a significant source of corporate financing. According to

Bradley and Roberts (2004), new issuances of private debt, including bank loans, range

from two to three times the amount of new issuances of public debt. Consequently,

understanding the determinants of the cost of bank loans is economically important. In

this study, we examine whether the relative importance of directorships to independent

directors who serve on multiple boards is associated with firms’ bank loan contracting.

Prior literature shows that corporate governance is an important determinant of

the cost of debt. Specifically, studies demonstrate that higher board quality, greater

disclosure quality, and higher institutional ownership can reduce the cost of debt

(Sengupta (1998), Anderson, Mansi, and Reeb (2003), Bhojraj and Sengupta (2003)).

With regard to board quality, the focus has been on traditional board quality measures,

such as board independence and size (Bhojraj and Sengupta (2003), Anderson, Mansi,

and Reeb (2004), Fields, Fraser and Subrahmanyam (2012)). Recently, a few studies have

begun to explore the effect of multiple directorships on governance quality (Masulis and

Mobbs (2014a, 2014b), Huang, Lobo, Wang, and Zhou (2014)). Following these studies,

we focus on the differential economic effects resulting from directors’ unequal

prioritization of their time and effort across multiple directorships.

Multiple directorships are very common for independent directors in U.S. public

firms. More than half of the independent directors at S&P 1,500 companies serve on

more than one directorship (Masulis and Mobbs (2014a)). Although multiple

directorships can signal the talent and quality of a director (Shivdasani and Yermack

3

(1999)), strong board monitoring demands time and energy (Yermack (1996)), and the

limited availability of these resources may constrain a director with multiple directorships

from effectively fulfilling her/his directorial responsibilities (Core, Holthausen, and

Larcker (1999), Fich and Shivdasani (2006)). To minimize the potential reputation

damage that may result from directorship overload and the ensuing ineffectiveness,

directors will rationally distribute their time and energy to different directorships based

on each directorship’s relative importance (and associated prestige) (Masulis and Mobbs

(2014a)). Using firm size to proxy for the relative importance of a directorship, Masulis

and Mobbs (2014a) find that directors have a better attendance record at relatively more

important directorships and are less likely to relinquish their more important director

seats when these firms perform poorly. Further, they show that firms viewed as more

important by their directors enjoy better performance, as measured by return on assets

(ROA) and Tobin’s q. These findings are consistent with Masulis and Mobbs’ conjecture

that firm-size-based reputation incentives motivate directors to prioritize their time and

energy to important directorships.

Directors’ relative attention among multiple directorships can affect the firm’s

cost of bank loans through several channels. First, creditors, such as banks, use

accounting-based numbers to assess firm health and viability. Given that bank loan

contracts are closely tied to accounting numbers (Drucker and Puri (2009)), the integrity

of accounting numbers and the financial reporting process are important for bank loan

contracting. Because board directors can directly monitor firms’ accounting practices,

board effectiveness is critically important in constraining managers’ opportunistic

accounting behavior (Hermalin and Weisbach (1998), Klein (2002), Garcia Lara, Garcia

4

Osma, and Penalva (2009)). Effective boards and audit committees are more diligent and

spend more time and energy in fulfilling their directorial responsibilities (Menon and

Williams (1994), Vafeas (1999), Xie, Davidson, and DaDalt (2003)).1

Given that

directors will unequally distribute their time and effort among their multiple directorships

in accordance with their relative importance (Masulis and Mobbs (2014a, 2014b)), firms

viewed as more important are likely to receive more attention from and monitoring of

managerial accounting practice by directors. Consistent with this argument, Huang et al.

(2014) find that directors are more (less) likely to constrain earnings management when

they serve on more (less) prestigious boards. We expect that such unequal monitoring

among multiple directorships is also reflected in bank loan contracting.

Second, strong board monitoring improves the borrowing firm’s performance and

thereby reduces the cost of borrowing that banks charge. Specifically, the prior literature

documents a negative relation between firm performance and cost of debt (Graham, Li,

and Qiu (2008)). Because firms with a greater proportion of devoted independent

directors have better firm performance (Masulis and Mobbs (2014b)), we conjecture that

these firms obtain bank loans with more favorable terms.

Third, debt holders price the quality of a borrowing firm’s governance and will

rationally require higher loan rates from firms with more opportunistic managerial

behavior (Boubakri and Ghouma (2010)). For example, poor governance increases the

risks that rent-seeking managers would engage in irregular activities that endanger the

wealth of the debt holders, resulting in higher cost of debt. Bhojraj and Sengupta (2003),

Anderson et al. (2004), and Fields et al. (2012) document that firms with higher quality

1 For example, McMullen and Raghunandan (1996) and Garcia Lara et al. (2009) show that more frequent

meetings by boards and audit committees are associated with more conservative accounting and a lower

likelihood of earnings restatement.

5

boards have lower interest rates.2 Given that boards that are viewed as more prestigious

by their directors can more effectively constrain managerial opportunism, we conjecture

that the strong corporate governance associated with such boards will be priced favorably

by banks and, therefore, lead to lower cost of loans.

We base our empirical tests on a sample of 8,512 firm-year observations of S&P

1,500 firms from 1998 to 2011. We find that firms with a greater proportion of

independent directors for whom the board is one of their most prestigious have lower

loan spreads, longer loan maturities, and fewer loan covenants. Further analysis indicates

that the effect of directorship importance on loan spreads is more pronounced when a

majority of board members are independent. This result indicates that independent

directors’ attention is more important when the board relies on the independent directors

to carry out its mission.

To ensure that the differences in loan contract terms are attributable to directors’

attention, we also conduct a first difference analysis. We find that differences in the

proportion of directors who rank the directorship high are associated with differences in

bank loan contracting terms. To address the concern that the results may be driven by the

determinants of the relative importance of a board (e.g., firm size), we employ

propensity-score-matching. We find consistent results with this matched sample analysis.

Additionally, we obtain similar results when we match firm-years by firm size instead of

propensity score. The results of these tests indicate that the relationship between

independent director attention and cost of borrowing does not merely reflect the

differences in firm size and other determinants of directorship importance. Instead, they

2 The board quality measures in Fields et al. (2012) include size, independence, advisory presence, tenure,

busyness, gender, share ownership, and base pay.

6

capture the effect of unequal allocation of attention of board directors on the cost of

borrowing; firms that receive more attention from their independent directors have lower

cost of bank loans.

In additional tests, we extend the effect of directorship importance to independent

directors to audit committee directors and find similar results. Furthermore, we find that

boards viewed as more important by directors exhibit lower loan syndicate concentration

(that is, more diverse lender pools) and pay lower annual loan fees. Finally, we show that

firms with more director attention have higher bond ratings. Overall, we provide robust

evidence that a strong board, as measured by the attention paid by its independent

directors, can reduce the cost of borrowing.

We make several important contributions to the literature. First, despite the

prevalence of directors serving on multiple boards, few studies have explored the impact

of the unequal allocation of effort by independent directors on firm behavior, Masulis

and Mobbs (2014a), (2014b) and Huang et al. (2014) being notable exceptions. We thus

contribute to the emerging literature on the unequal allocation of effort by directors with

multiple directorships, by linking the unequal monitoring to bank loan contracting. Our

results suggest that when retaining a director, a firm needs to take into consideration the

relative importance (and time and effort) it will receive from the director because, among

other effects, this has implications for the pricing of its loans.

Second, we add to the literature linking board governance and cost of debt.

Bhojraj and Sengupta (2003), Anderson et al. (2004), and Fields et al. (2012) find that the

cost of debt is inversely related to traditional measures of board quality such as board

independence, size, and meeting frequency. We contribute to this line of research by

7

focusing on the unequal attention of directors with multiple directorships, which to date is

unexplored in the literature on board governance and cost of debt.

The remainder of the paper is organized as follows. We discuss prior research and

develop our hypotheses in Section 2. Section 3 describes the sample selection process and

presents descriptive statistics. Section 4 discusses the results of our main analyses, and

Section 5 presents the results of additional analyses. Section 6 concludes the study.

II. Prior Literature and Hypotheses Development

A. Multiple Directorship and Board Effectiveness

Research on multiple directorships has primarily focused on how busy directors

are and generally finds that having multiple directorships can make directors too busy to

effectively monitor all the firms under their supervision. For example, Core et al. (1999)

suggest that busy directors are associated with excessive CEO compensation, which leads

to lower firm performance. Shivdasani and Yermack (1999) find that CEOs are likely to

appoint busy directors who serve on multiple boards, and interpret this phenomenon as

CEOs’ attempts to reduce monitoring pressure. Fich and Shivdasani (2006) propose that

firms with busy boards (e.g., firms with a majority of members holding three or more

directorships) have weaker governance and are associated with lower operating

profitability (i.e., ROA) and market valuation (i.e., market-to-book ratio). They also show

that the market responds positively to news of departures of busy directors. However,

Ferris, Jagannathan, and Pritchard (2003) find no evidence that busy directors shirk their

responsibilities to serve on subcommittees; nor do they find an association between

multiple directorships and the likelihood of securities litigation.

Given busy directors’ time constraints, Masulis and Mobbs (2014a) expect that

8

directors differentially value each directorship based on its relative importance to their

reputation, and allocate their time and energy among these boards accordingly. Since firm

size is associated with greater visibility, prestige, compensation, and opportunity to attract

additional external director appointments (Ferris et al. (2003), Ryan and Wiggins (2004),

Adams and Ferreira (2008)), Masulis and Mobbs (2014a) use relative size of each firm

where they serve as independent directors to proxy for the relative importance the

director places on this directorship and the relative amount of time and effort allocated to

this directorship. At the director level, they find that independent directors are more likely

to attend the board meetings of their more important directorships and less likely to

relinquish these board seats, even when these firms perform poorly. At the firm level,

they find that the fraction of independent directors who view the board as relatively more

important is positively associated with the sensitivity of CEO departure to poor

performance as well as to overall firm performance (i.e., ROA) and valuation (i.e.,

Tobin’s q). Further, in a related study, Masulis and Mobbs (2014b) find that greater

reputation incentives of a board lead to less negative outcomes, such as stock delistings,

violations of debt covenants, financial report restatements, options backdating, securities

class actions, and reductions of cash dividends, and more positive firm outcomes.

B. Multiple Directorships, Accounting Quality, and the Cost of Debt

Lenders explicitly rely on financial statement numbers in setting debt covenants

and performance pricing provisions (Dichev and Skinner (2002), Li (2010)). Thus, bank

loan contracting must reflect the quality of a firm’s accounting. Anderson et al. (2004)

propose that creditor reliance on accounting-based debt covenants suggests that debtors

are potentially concerned with board of director characteristics that influence the integrity

9

of financial accounting reports. In a sample of S&P 500 firms, they find that the cost of

debt is negatively related to board independence and board size. They also find that fully

independent audit committees are associated with a significantly lower cost of debt.

Similarly, yield spreads are also negatively related to the size and meeting frequency of

the audit committee. Overall, these results provide market-based evidence that boards and

audit committees are important elements affecting the reliability of financial reports. The

importance of accounting numbers in bank loan contracting and the incentives of

management to manipulate these numbers make banks rationally sensitive to major

changes in governance dimensions that may affect the integrity and reliability of the

financial reporting process (Smith (1993)).

Board monitoring, and especially monitoring by audit committees, is critically

important in constraining managers’ opportunistic accounting behavior (Beasley (1996),

Klein (2002), Xie et al. (2003), Beekes, Pope, and Young (2004), DeFond, Hann and Hu

(2005), Ahmed and Duellman (2007), Larcker, Richardson, and Tuna (2007), Garcia Lara

et al. (2009),). Effective boards and audit committees demand more attention, effort,

time, and energy (Vafeas (1999), Xie et al. (2003)). If directors distribute their attention

unevenly among multiple directorships according to their relative importance, then the

monitoring effectiveness of a firm’s accounting decisions will vary with the amount of

attention given to a directorship. Consistent with this reasoning, Huang et al. (2014) find

that directors are more (less) likely to constrain earnings management when they serve on

more (less) prestigious boards.

Given that the cost of borrowing reflects the accounting quality and that the

distribution of directors’ efforts among multiple directorships affects accounting quality,

10

we therefore expect firms that are viewed as more prestigious by their independent

directors to have lower cost of borrowing.

C. Multiple Directorships, Firm Performance, and the Cost of Debt

Strong board monitoring is associated with higher profitability and lower return

volatility, which reduce the cost of borrowing charged by banks. Specifically, firms with

higher profitability have lower default risk and can obtain loans at lower rates and better

terms (Graham et al. (2008)). Prior literature also documents a positive relation between a

firm’s earnings volatility and cost of debt, consistent with banks benefiting from a stable

stream of cash flow that can support bond repayments (e.g., Graham et al. (2008)). Firms

with a greater proportion of devoted independent directors have better firm performance

(e.g., higher ROA). These firms also have a lower likelihood of violating loan covenants

(Masulis and Mobbs (2014b)). We therefore conjecture that these firms will enjoy more

favorable bank loans terms.

D. Multiple Directorships, Firm Governance, and the Cost of Debt

Boubakri and Ghouma (2010) argue that debt holders price the quality of a

borrowing firm’s governance and will rationally require higher interest rates from firms

with more opportunistic managerial behavior. For example, to extract private benefits,

opportunistic managers may divert cash away from paying debt holders, and this

likelihood of managerial opportunism will negatively affect the loan terms. Strong board

monitoring of managers also alleviates the risks that rent-seeking managers would engage

in fraudulent activities that endanger the wealth of debt holders, or that they would divert

assets away from supporting debt cash flow. For example, poor governance can lead to

financial fraud (e.g., Enron, WorldCom, Xerox) that can inflict catastrophic damages on

11

debt holders and also damage a firm’s disclosure credibility about forecasted future cash

flows, resulting in a higher cost of debt. Consistent with this reasoning, Graham et al.

(2008) document that financial irregularities are followed by higher interest rates, shorter

maturities, and more covenant restrictions by banks.

Specific evidence indicates that board quality impacts the cost of bank debt.

Fields et al. (2012) analyze the relation between comprehensive measures of board

quality and the cost as well as the non-price terms of bank loans. They show that firms

that have higher quality boards with a greater advisory presence borrow at lower interest

rates. This relation holds even after controlling for other firm characteristics, such as

ownership structure, CEO compensation policy, and shareholder protection, as well as the

size and financial condition of the borrower. They also document that board quality and

other governance characteristics influence the likelihood that loans have covenant

requirements, but the relations differ by covenant type. When they combine the direct and

indirect costs of bank loans, they find that firms with large, independent, experienced,

and diverse boards, and with lower institutional ownership, borrow more cheaply. Other

studies in the literature also show empirically that measures of corporate governance

quality, such as disclosure quality, institutional ownership, and board independence, are

negatively related to the cost of debt (Sengupta (1998), Anderson et al. (2003), Bhojraj

and Sengupta (2003)).

Given the costly renegotiation following default (references) and the costly

enforcement of loan contracts (Bolton and Scharfstein (1996)), lenders are willing to

extend credit on more favorable terms if they know ex ante that the borrower has strong

governance. The resulting lower governance risk that debt holders perceive will be priced

12

favorably in the firm’s loan terms (Graham et al. (2008), Boubakri and Ghouma (2010),

Qi, Roth, and Wald (2010)). Firms with more attentive directors have better governance

quality (Masulis and Mobbs (2014a)). When directors view the board as more

prestigious, they can better constrain managerial opportunism. As a consequence, we

conjecture that the strong corporate governance associated with these boards will lead

banks to offer more attractive loan features, including lower cost loans.

Based on the aforementioned arguments, we propose the following hypothesis:

H1: Firms associated with more attentive directors receive more favorable bank

loan contract terms.

III. Main Variables, Data Sources, Sample, and Descriptive Statistics

A. Measurement of Board Importance, Sample Selection, and Data Sources

Following Masulis and Mobbs (2014a, 2014b), we rank the relative importance of

directorships for a director sitting on multiple boards using the size of each firm’s equity

market capitalization. Among all the directorships a director serves, the directorship with

the largest (smallest) firm size is deemed the most (least) important and prestigious.

Percent_Ranked_High (Percent_Ranked_Low) captures the percentage of independent

directors for whom this directorship’s rank is 10% larger (smaller) than their smallest

(largest) directorship measured by a firm’s market capitalization. We focus on

independent directors because they are more effective monitors (Beasley (1996), Klein

(2002), Xie et al. (2003), Ahmed and Duellman (2007), Larcker et al. (2007), Garcia Lara

et al. (2009)).

Panel A of Table 1 presents the sample selection procedure. We start by collecting

independent director board data from RiskMetrics and merge this data with bank loan

information from DealScan for the years 1998 to 2011. To ensure that directorship

13

importance drives loan contract terms, we match the loan contract terms with the

preceding year’s director board data. Our initial sample consists of 10,996 firm-year

observations with director data available in RiskMetrics, which covers board information

for the S&P 1,500 firms. We then exclude all 1,348 observations in the financial services

and utility industries. Next, we drop 325 observations with bridge loans and non-fund-

based facilities, such as leases and standby letters of credit. Last, we exclude 811

observations that lack the necessary data to construct the control variables used in our

empirical tests. Our final sample comprises 8,512 firm-year observations. We extract

financial data from Compustat, and stock return data from CRSP.

Pane B of Table 1 presents the sample distribution by year. The sample size varies

widely over time, ranging from a high of 1,187 firms in 2001 to a low of 152 firms in

2011.3

B. Descriptive Statistics

Panel C of Table 1 shows the descriptive statistics for our sample. The first part

shows the descriptive statistics for the bank loan characteristics. The mean and median of

“all-in-drawn” spread are 175 and 150 basis points, respectively. The average maturity is

45.63 months. The mean and median facility amounts are 506.6 and 250 million,

respectively. The percentage of secured bank loans is 39.2%. The averages of total,

general and financial covenants are 3.827, 2.344 and 1.483, respectively. Of the bank

loans in the sample, 48.7% have performance pricing and the average number of lenders

per loan is 10. 3 The main reason for the large drop in our sample after 2007 is that new loans fell by almost 50%

during the financial crisis (2007-2009). This is consistent with Ivashina and Scharfstein (2010),

who also document a large drop-off. Specifically, the full sample in DealScan decreases

significantly, from 6,927 in 2006 to 4,943 in 2007 and 3,843 in 2008. There are only 152 firms in

2011 because most data for 2011 were not available at the time of our data cutoff.

14

The second part of Panel C presents the descriptive statistics for board

characteristics. Across all boards, 19.2 percent of independent directors classify the

directorship as of high importance (Percent_Ranked_High) and 24.2 percent of

independent directors classify the directorship as of low importance

(Percent_Ranked_Low). While 9.3 percent of firms are ranked by the majority of their

independent board members as of high importance (Majority_Ranked_High), 12.3

percent of firms are ranked as of low importance (Majority_Ranked_Low). On average,

CEO ownership is 2.5% of outstanding shares and independent director ownership is

1.1%.

The last part of Panel C presents the descriptive statistics for firm characteristics.

We do not discuss these statistics in detail for the sake of brevity.

IV. Methodology and Empirical Results

A. Methodology

We estimate the relationships between directorship importance and bank loan

contract terms using the following specification:

Bank Loan Contracting =

β0 + β1Percent_Ranked_High + β2Percent_Ranked_Low + β3Majority independent

+ β4CEO ownership + β5CEO ownership squared + β6Independent ownership

+ β7Ln Facility amount + β8Ln Maturity + β9Number of covenants

+ β10Number of lenders + β11Performance pricing + β12Total assets + β13Leverage

+ β14Return on assets + β15Operating cash flow volatility + β16Tangibility

+ β17Z-score + β18MB + β19Credit spread + β20Term spread + Loan Purpose

+ Loan Type + Industry + Year + ε (1)

where, Bank Loan Contracting separately refers to one of the following bank loan

characteristics: cost of borrowing, loan maturity, and number of covenants.4

4 Model (1) does not include Ln (Maturity) as a control variable when the dependent variable, Bank Loan

Contracting, represents loan maturity, and does not include Number of covenants as a control variable when

15

Following Masulis and Mobbs (2014b), we control for several governance

characteristics in our regression analysis. Specifically, we include an independent board,

termed Majority independent, CEO ownership, CEO ownership squared, and

Independent ownership in Equation (1). Drawing on prior studies (Qian and Strahan

(2007), Boubakri and Ghouma (2010), and Qi et al. (2010)), we also control for loan

characteristics and firm characteristics. We use the natural log of the amount of the

facility committed by the facility’s lender pool (Ln (Facility amount)), the natural log of

the number of months to maturity (Ln (Maturity)), the total number of covenants (Number

of covenants), the number of lenders (Number of lenders), and whether the facility has a

performance pricing provision (Performance pricing), to capture other loan

characteristics besides spread.5 We control for firm characteristics, including the natural

log of book value of assets (Ln (Total assets)), total liabilities divided by total assets

(Leverage), income before extraordinary items divided by total assets (Return on assets),

standard deviation of quarterly cash flows from operations divided by total assets

(Operating cash flow volatility), property, plant and equipment scaled by total assets

(Tangibility), Altman’s Z-score (Z-score), and market-to-book ratio (MB), all measured at

the beginning of the fiscal year. We also use two variables to control for macroeconomic

factors, the difference in the yield between BAA- and AAA-rated corporate bonds

measured one month before the loan becomes active (Credit spread), and the difference

in the yield between ten-year and two-year U.S. Treasury bonds measured one month

before the loan becomes active (Term spread). We include Loan Purpose and Loan Type

to control for differences in loan purpose and type. Specifically, Loan Purpose captures

the dependent variable represents the number of covenants. 5 We do not include annual loan fees as a control variable because only 33.6% of our sample facilities have

annual loan fee information.

16

the primary purpose for the loan, including acquisition line, commercial paper backup,

corporate purposes, debt repayment, takeover, working capital, and other purposes. Loan

type includes 364-day facilities, delay draw term loans, revolvers and term loans. Finally,

we include industry and year fixed effects. Appendix A provides detailed descriptions of

all the variables used in the model.

B. Relation between Cost of Borrowing and Importance of Directorship

Table 2, Panel A presents the estimation results relating relative importance of

multiple directorships to the cost of debt, measured using the log of all-in-drawn spread

as the dependent variable. Columns (1)-(3) show the results using Percent_Ranked_High

and Percent_Ranked_Low. Column (1) controls for only board characteristics, Column

(2) controls for all the control variables except loan characteristics, and Column (3)

controls for all the control variables. We find similar results in all three columns.

Specifically, in Column (3), we find that Percent_Ranked_High is negatively related to

the cost of debt, as indicated by the coefficient β1, which equals -0.161 with a t-value of -

5.04. Moving from the first quartile (0.000) to the third quartile (0.333) of

Percent_Ranked_High reduces the cost of debt, measured by the all-in-drawn spread, by

1.131 basis points.6 We also find that Percent_Ranked_Low is significantly positively

related to the cost of debt, as indicated by β2, which has a coefficient estimate of 0.065

with a t-value of 2.78. Moving from the first quartile (0.000) to the third quartile (0.375)

for Percent_Ranked_Low increases the cost of debt by 1.057 basis points for all-in-drawn

spread.7 The difference between β1 and β2 is highly significant (F = 30.21, p = 0.00).

These results show that firms ranked high by directors with multiple directorships are

6 It is calculated as follows: exp((0.333 – 0.000)*0.161) = 1.131.

7 It is calculated as follows: exp((0.375-0.000)*0.065)=1.057.

17

associated with lower cost of debt, while firms ranked low by directors with multiple

directorships are associated with higher cost of debt.8

Columns (4), (5) and (6) present regressions analogous to those in Columns (1) -

(3), except that we use the variables Majority_Ranked_High and Majority_Ranked_Low

in place of Percent_Ranked_High and Percent_Ranked_Low. We find consistent evidence

with this alternative specification; firms with boards valued highly by multiple

directorship directors have lower cost of debt.

Table 2, Panel B examines the interaction effect of directorship importance and

overall board independence (Majority Independence) on cost of debt. Column (1) shows

the results for Percent_Ranked_High and Percent_Ranked_Low, and Column (2) shows

the corresponding results for Majority_Ranked_High and Majority_Ranked_Low. All

columns include all the control variables. The results show significantly negative

coefficients on the interaction term Percent_Ranked_High*Majority_Independent in

Column (1) (β4 = -0.244, t = -4.12, p = 0.00), and

Majority_Ranked_High*Majority_Independent in Column (2) (β4 = -0.160, t = -2.94, p =

0.00). These estimates indicate that the effect of strong director attention on reducing the

cost of debt is more pronounced when the majority of the board is independent.

We also find significantly positive coefficients on the interaction term

Percent_Ranked_Low*Majority_Independent in Column (1) (β5 = -0.146, t = -3.02, p =

0.00), indicating that the effect of weak director attention on increasing the cost of debt is

also more pronounced when the majority of the board is independent. These results

suggest that when the majority of a board is composed of independent directors, the

8 Ln (Maturity), as a control variable, is not significant in this and most of the subsequent tables. This is

because we control for the loan type fixed effect, which is highly correlated with loan maturity. Once we

drop the loan type fixed effect, the coefficient on Ln (Maturity) becomes significant in most regressions.

18

effectiveness of the board depends largely on these independent directors’ efforts and

attention; as a result, these directors’ attention to the board becomes more important in

affecting the cost of debt.9

C. Relation between Loan Maturity and Importance of Directorship

Table 3, Panel A presents the results of the relation between the relative

importance of multiple directorships and loan maturity. Columns (1) - (3) show the

regression results using Percent_Ranked_High and Percent_Ranked_Low. All three

columns yield similar results. Specifically, in column (3), we find that

Percent_Ranked_High is positively related to loan maturity (β1 = 0.052, t = 2.45, p =

0.01). Moving from the first quartile (0.000) to the third quartile (0.333) for

Percent_Ranked_High increases loan maturity by 1.040 months, which represents a

2.279% increase in loan maturity.10

Percent_Ranked_Low is insignificantly positively

related to loan maturity (β2 = -0.003, t = -0.17, p =0.87). The difference between β1 and β2

is significant at the 0.1 level (F = 1.88, p = 0.06). These results show that firms ranked

high by directors with multiple directorships are associated with longer loan maturity

than firms ranked low by directors with multiple directorships.

Columns (4) - (6) present the results of regressions using the variables

Majority_Ranked_High and Majority_Ranked_Low. We find consistent evidence with

this alternative specification; firms valued highly by directors with multiple directorships

9 The results indicate that the coefficients on Percent_Ranked_High, Percent_Ranked_Low,

Majority_Ranked_High, and Percent_Ranked_Low, in Columns (1) and (2) of Table2, Panel B are no

longer significant. One explanation for these results is that these coefficients reflect the effect of

directorship importance for firms that do not have a majority of independent directors on their boards (i.e.,

for 16.5% of sample observations). In other words, directorship importance is relevant for bank loan

contracting only when the board comprises a majority of independent directors (i.e., for 83.5% of sample

observations). 10

The loan maturity increase is calculated as follows: exp((0.333 – 0.000)*0.052) = 1.040, and the

percentage is calculated as follows: 1.040/45.630=2.279%.

19

enjoy longer loan maturity.

Table 3, Panel B presents the results of regressions relating multiple directorships

to loan maturity measured using loan maturity instead of log loan maturity. We find

results qualitatively similar to those reported in Table 3, Panel A.

D. Relation between Number of Covenants and Importance of Directorship

Table 4, Panel A presents the results of regressions relating multiple directorships

to number of total covenants. Columns (1) - (3) show the regression results with

Percent_Ranked_High and Percent_Ranked_Low. The results in all three columns are

similar. For example, in Column (3), Percent_Ranked_High is significantly negatively

related to the number of total covenants (β1 = -0.079; t = -1.85, p = 0.06)11

and

Percent_Ranked_Low is insignificantly negatively related to the number of total

covenants (β2 = -0.017, t = -0.52, p =0.60). These results show that firms ranked high by

directors with multiple directorships are associated with a smaller number of total

covenants. Columns (4) - (6) present results with Percent_Ranked_High and

Percent_Ranked_Low. These results are consistent with the evidence in Columns (1) -

(3); we continue to find that firms valued highly (lowly) by multiple directorship

directors have fewer (more) total covenants.

Table 4, Panel B presents the results of regressions relating multiple directorships

to number of general covenants. We find that firms valued highly (lowly) by multiple

directorship directors have fewer (more) general covenants in Columns (1) and (4), but

not for the models in the other columns.

Table 4, Panel C presents the results of regressions relating multiple directorships

11

The marginal effect of Percent_Ranked_High is -0.25. That is, for a one unit increase in

Percent_Ranked_High, the number of covenants will decrease by 0.25.

20

to number of financial covenants. Except for the model in Column (2), we find that firms

valued highly by multiple directorship directors have fewer financial covenants.

V. Additional Analysis

A. First Differences Analysis

The results reported earlier are based on a levels analysis, which is inherently

vulnerable to omitted variables that may bias the coefficient estimates. To alleviate this

concern, we conduct a first differences analysis by relating the two-year change (from

year t-1 to year t+1) in the loan contract terms to the corresponding two-year change

(from year t-2 to year t) in the percentage of directors that view the directorship as more

or less important (i.e., two-year change in Percentage_Ranked_High and

Percentage_Ranked_Low).12

We also limit the sample to observations with non-zero

changes in directorship importance from year t-2 to year t. Table 5 presents the results of

this first differences analysis. We continue to find results consistent with the results

reported earlier. Specifically, we find that increases (decreases) in

Percentage_Ranked_High are associated with decreases (increases) in loan spread and

increases (decreases) in loan maturity. Furthermore, increases (decreases) in

Percentage_Ranked_Low are associated with increases (decreases) in the number of total

covenants and general covenants. These results provide further evidence that the

differences in the cost of bank loans can be attributed to the changes in importance of

directorships.

B. Using a Propensity-score-matched Sample

Given that we use firm size to proxy for the relative prestige of a firm, and given

12

The results are similar if we use annual changes.

21

that prior research indicates that larger firms are likely to have better credit history and,

therefore, lower cost of debt, our results relating loan characteristics to the relative

importance of multiple directorships could be driven by differences in firm size. Although

we include total assets (Ln(Total Assets)) in our regressions to control for this size

effect,13

we conduct two additional tests to alleviate this potential concern.

First, we select a matched sample using the propensity score matching technique

and compare the results for this sample to those for our test sample. Table 6, Panel A

presents the results for the first-stage regression. The dependent variable, High_Low, is

an indicator variable that equals one if Majority_Ranked_High equals one, and zero if

Majority_Ranked_Low equals one. Firms with larger market value and higher operating

earnings volatility are more likely to have Majority_Ranked_High boards. Firms with

lower percentage of female directors and with majority independent directors are less

likely to have Majority_Ranked_High boards. We use the estimated coefficients from this

first-stage regression to compute the propensity score for each observation in our sample.

We then match each firm-year that is ranked as relatively important by the majority of its

independent directors (i.e., a firm-year with Majority_Ranked_High = 1) with a firm-year

that has the opposite ranking (i.e., Majoroty_Ranked_Low = 1), that is in the same-

industry (based on the Fama-French 48 industry classification), and has the closest

13 Ln(Total Assets), as a measure of size, usually has the same association with the loan contract terms as

Percent_Ranked_High and Majority_Ranked_High. However, in Table 3, loan maturity is negatively

associated with Ln(Total Assets) and positively associated with Percent_Ranked_High and

Majority_Ranked_High. This is not inconsistent with the previous literature, which presents mixed

evidence between loan maturity and firm size. For example, when using loan maturity as the dependent

variable, Graham et al. (2008) and Chan, Chen, and Chen (2013) find that the coefficients on size are

negative but insignificant. Furthermore, Costello and Wittenberg-Moerman (2011) report a positive and

significant coefficient on size in Table 5, but a negative and significant coefficient on size in Table 8.

22

propensity score.14

We use these matched firms as the control sample.

Table 6, Panel B presents the regression results using the propensity-score-

matched sample. Column (1) shows that High_Low is associated with significantly lower

loan spreads. Column (2) shows that High_Low is not associated with loan maturity. The

regression results in Columns (3), (4) and (5) show that High_Low is associated with

significantly lower number of total covenants, general covenants and financial covenants.

These results are qualitatively similar to our primary results, except for loan maturity.

C. Relation between Loan Contract Terms and Importance of Audit Committee

Directorship

Because monitoring by audit committees is critically important in ensuring the

reliability of a firm’s accounting information, we also examine the effect of audit

committee directorship importance on loan contract terms. We define Audit

Percent_Ranked_High (Audit Percent_Ranked_Low) as the percentage of independent

audit committee members to whom this ranked directorship is 10% larger (smaller) than

their smallest (largest) directorship measured by the firm’s market capitalization. We also

define Audit Majority_Ranked_High (Audit Majority_Ranked_Low) as an indicator

variable that equals one if the majority of independent outside directors in the audit

committee classify this directorship as high (low) ranked (i.e., 10% larger (smaller) than

their smallest (largest) directorship by market capitalization of the firm), and zero

otherwise.

Table 7 presents the regression estimates. The results are very similar to our main

findings for the full board. Specifically, the results indicate that greater audit committee

directorship importance is associated with lower loan spread, longer maturity, and fewer

14

The results are similar if we match by firm size (market value of equity) instead of propensity score.

23

covenants. This suggests that the audit committee plays an important role in reducing the

cost of debt, probably because of its importance in monitoring a firm’s accounting

decisions.

D. Relation between Syndicate Concentration, Annual Loan Fees, and Importance of

Directorship

We also examine the effect of relative board importance on two additional

measures of bank loan contracting: syndicate concentration and annual loan fees.

Following Sufi (2007), high information asymmetry forces the lead arranger to take a

larger stake in the loan and form a more concentrated syndicate. We measure syndicate

concentration using the Herfindahl index calculated using each syndicate member’s share

in the loan. A higher index score indicates more concentration by lenders. Table 8

presents the estimation results using Equation (1) with syndicate concentration as the

dependent variable. We find that the coefficients on Percent_Ranked_High and

Majority_Independent_High are significantly negative in Columns (1) and (3), where

only board characteristics are included as control variables. In addition, we find that the

coefficient on Majority_Independent_High is significantly negative in Column (4), where

all the control variables are included in the regression. Overall, the results indicate that

boards with more motivated directors have lower levels of syndicate concentration. In

other words, more lenders are willing to lend to a firm if it has a greater fraction of

directors that view its board as more important.15

Table 9 shows the results using Equation (1) with annual loan fee as the 15

We also examine the effect of directorship importance on the percentage stake of the lead arranger. We

find some evidence that firms with more devoted independent directors have lead arrangers with a lower

percentage stake, consistent with directorship importance reducing loan syndicate concentration.

Furthermore, we examine the effect of directorship importance on the collateral requirement, and find some

evidence consistent with higher (lower) directorship importance decreasing (increasing) the likelihood of

collateral requirement.

24

dependent variable. Annual fee is the natural log of the number of basis points of a

facility commitment amount that a borrower is required to pay on an annual basis

regardless of any loan outstanding. In Columns (1) and (2), Percent_Ranked_High

(Percent_Ranked_Low) is significantly negatively (positively) related to the annual loan

fee. Similarly, in Column (3), Majority_Ranked_High (Majority_Ranked_Low) is

significantly negatively (positively) related to the annual loan fee. These results indicate

higher annual loan fees for firms that are ranked as less important by their directors.

In sum, the results in Tables 8 and 9 suggest that boards that are viewed as more

prestigious by their members have more diffuse lender groups and lower annual loan

fees.

E. Relation between Bond Ratings and Importance of Directorship

We also examine the effect of directorship importance on bond ratings. Since

bond rating is an important determinant of bond cost, this test also indirectly reflects the

relation between director attention and bond cost. We obtain bond data from Mergent

Fixed Income Securities Database (FISD), and convert the bond ratings to numerical

values using the conversion in Becker and Milbourn (2011).16

We compute Ln (Bond

Maturity) as the natural logarithm of the number of days to maturity, and include it as a

control variable. Following Mansi et al. (2011), we include two major determinants of

bond ratings, Idiosyncratic Risk and Firm Risk, as additional control variables.

Furthermore, we include Ln (Total Assets) to control for firm size and information

environment, Leverage to control for capital structure, Return on assets to control for

firm performance, Z-score to control for default risk, and MB to control for growth. We

16

Following Becker and Milbourn (2011), we code AAA as 28, AA+ as 26, AA as 25, AA- as 24, A+ as 23,

A as 22, A- as 21, BBB+ as 20, BBB as 19, BBB- as 18, BB+ as 17, BB as 16, BB- as 15, B+ as 14, B as

13, B- as 12, CCC as 11, CC as 7, and C as 4.

25

also include all the board characteristics in Equation (1).

Table 10 presents the results of estimating the effect of directorship importance on

bond ratings. We find that Percent_Ranked_High is significantly and positively

associated with bond ratings in Columns (1) and (2), and Majority_Ranked_High has a

similar association with bond ratings in Columns (3) and (4). Further,

Percent_Ranked_Low (Majority_Ranked_Low) has a significantly negative association

with bond ratings in Column (1) (Column (3)). These results indicate that firms whose

directorships are viewed as more (less) important by their directors are associated with

higher (lower) bond ratings.

E. Robustness Tests

Loan terms such as loan spread, maturity, and number of covenants are often

determined simultaneously. We conduct robustness tests to address this potential

endogeneity concern but, for brevity, do not tabulate the results. First, we follow Graham

et al. (2008) and employ two stage least squares estimation. In the first stage, we regress

bank loan maturity on the borrower’s total assets maturity.17

In the second stage, we re-

estimate our main model using the fitted value of loan maturity from the first stage as the

instrument for loan maturity. We find results that are consistent with our main results.

Second, to address the endogeneity concern for cost of debt and number of

covenants, we follow Costello and Wittenberg-Moerman (2011) and estimate a

simultaneous equations model, with Prior lending rate as the instrument for cost of debt

17

In the first stage, following Johnson (2003), we measure the total assets maturity as two different

components. The first component, current assets maturity equals current assets (ACT) scaled by cost of

goods sold (COGS). The second component, long-term assets maturity equals gross PPE (PPEGT) scaled

by depreciation expense (OIBDP-OIADP). Total assets maturity equals the weighted (ACT/AT and

PPEGT/AT) sum of these two components.

26

and Lead Reputation as the instrument for number of covenants.18

Once again, we find

consistent results.

In additional robustness tests, we use other measures of importance (reputation

incentives) of a directorship. First, similar to Masulis and Mobbs (2014), we rank

directorships using the market value of total assets (i.e., sum of book value of liabilities

and market value of equity). Second, we change the threshold of the relative importance

of directorships from 10% of a firm’s market capitalization to 20% and 50%. We find

results under these three alternative measures to be very similar to the primary results

using the 10% threshold. Third, we control for potential nonlinearity in the relation with

firm size (Ln(Total Assets)) by including the square of this variable as an additional

control variable and find consistent results.

We control for the direction of performance pricing provisions in our fourth

robustness test. Our sample is considerably reduced because only 48.7% of the

observations include a performance pricing provision. Following Asquith, Au, Covert,

18

This simultaneous equations model is specified as follows. Prior Lending Rate equals the cost (Ln(All-in-

drawn)) of the previous deal for the same borrower. Lead Reputation is an indicator variable that equals

one if the loan is syndicated by a reputable lead arranger in the syndicated loan market. Following

Bushman and Wittenberg-Moerman (2012), we define a lead arranger bank as reputable if its market share

is larger than 2% in our sample period. The market share is the ratio of total amount of loans syndicated as

a lead arranger to the total amount of syndicated loans in our sample period. All other variables are defined

as in Equation (1).

Ln(All-in-drawn)= β0 + β1Percent_Ranked_High (Majority_Ranked_High) + β2Percent_Ranked_Low

(Majority_Ranked_Low )+ β3Majority independent + β4CEO ownership + β5CEO

ownership squared + β6Independent ownership + β7Ln Facility amount + β8Ln Maturity

+ β9Number of covenants + β10Number of lenders + β11Performance pricing + β12Total

assets + β13Leverage + β14Return on assets + β15Operating cash flow volatility +

β16Tangibility + β17Z-score + β18MB + β19Credit spread + β20Term spread +β21Prior

lending rate + Industry + Year + Loan Purpose + Loan Type + ε (2)

Number of covenants = β0 + β1Percent_Ranked_High (Majority_Ranked_High) + β2Percent_Ranked_Low

(Majority_Ranked_Low )+ β3Majority independent + β4CEO ownership + β5CEO

ownership squared + β6Independent ownership + β7Ln Facility amount + β8Ln(All-in-

drawn) + β9Ln Maturity + β10 Prior Leader + β11Number of lenders + β12 Secured +

β13Total assets + β14Leverage + β15Return on assets + β16Operating cash flow volatility

+ β17Tangibility + β18Z-score + β19MB + β20Credit spread + β21Term spread + + β22 Lead

Reputation + Industry + Year + Loan Purpose + Loan Type + ε (3)

27

and Pathak (2013), we define Interest-Increasing PP as an indicator variable that equals

one if the loan contract contains an interest increasing performance pricing provision, and

zero otherwise. We include Interest-Increasing PP in Equation (1) and find results similar

to our main results, except for the number of covenants which are no longer significant,

possibly due to the reduced sample size. Furthermore, we find some evidence that greater

director attention is negatively associated with the likelihood of having an interest-

increasing performance pricing provision.

Fifth, because dual-class firms have unique voting features that may affect the

reputation incentives of independent directors, we repeat our analyses after excluding

these firms and continue to find similar results.

Sixth, since firms with majority block holders may limit the reputation incentives

of independent directors, we re-estimate the previous test after deleting these firms and

find that our results are robust.

VI. Conclusions

In this study, we examine the relation between reputation incentives in the

director labor market and bank loan contracting. Previous literature, such as Masulis and

Mobbs (2014a, 2014b) and Huang et al. (2014), finds that directors with multiple

directorships distribute their effort unequally based on the directorships’ relative prestige.

Directors with multiple directorships spend more time and effort on their more

prestigious boards. We investigate whether bank loan contracting reflects such unequal

distribution of monitoring effort. We find that firms with a greater proportion of

independent directors for whom the board is one of their most prestigious have lower cost

of debt, longer loan maturity, and fewer covenants. These results are robust to various

28

sensitivity tests, including first differences, propensity score matching, firm-size

matching, and using alternative measures of board importance. In addition, we also find

similar results for the audit committee level. Furthermore, we find that more important

boards are associated with lower syndicate concentration and lower annual fees. Finally,

we find that directorship importance is also significantly related to bond ratings.

To the best of our knowledge, this is the first study to directly link unequal

attention of directors to the cost of debt. We contribute to the understanding of the

economic effects of director attention and how director attention affects bank loan

contracting. By linking the unequal monitoring effort to bank loan contracting, we thus

contribute to the emerging literature on directors’ unequal distribution of their efforts

among multiple directorships (e.g., Masulis and Mobbs (2014a), (2014b), Huang et al.

(2014)). Our results indicate that a firm needs to take into consideration director

reputation incentives because of the importance of this characteristic for loan pricing.

We also add to the literature linking strong board governance to lower cost of debt

(Bhojraj and Sengupta (2003), Anderson et al. (2004), and Fields et al. (2012)). We

contribute to this line of research by focusing on the unequal distribution of directors’

efforts across multiple boards, an area that has not been previously studied. Our study

highlights the importance of building a board that will receive the most attention from its

board members.

29

References

Abbott, L.; S. Parker; and G. Peters. “Audit Committee Characteristics and

Restatements.” Auditing: A Journal of Practice and Theory, 23 (2004), 69–87.

Adams, R., and D. Ferreira. “Do Directors Perform for Pay?” Journal of Accounting and

Economics, 46 (2008), 154–171.

Aggarwal, R., and J. Goodell. “Markets and Institutions in Financial Intermediation:

National Characteristics as Determinants.” Journal of Banking and Finance, 33 (2009),

1770–1780.

Ahmed, A., and S. Duellman. “Evidence on the Role of Accounting Conservatism in

Corporate Governance.” Journal of Accounting and Economics, 43 (2007), 411–437.

Anderson, C.; A. Mansi; and D. Reeb. “Founding Family Ownership and the Agency

Cost of Debt.” Journal of Financial Economics, 68 (2003), 263–283.

Asquith, P.; A. Au; T. Covert; and P. Pathak. “The Market for Borrowing Corporate

Bonds.” Journal of Financial Economics, 107 (2013), 155–182.

Beasley, M. “An Empirical Analysis of the Relation between the Board of Director

Composition and Financial Statement Fraud.” The Accounting Review, 71 (1996), 443–

465.

Beasley, M., and K. Petroni. “Board Independence and Audit-firm Type.” Auditing: A

Journal of Practice and Theory, 20 (2001), 97–114.

Beck, T.; A. Demirguc-Kunt; and R. Levine. “A New Database on Financial

Development and Structure.” World Bank mimeo (1999).

Becker, B., and T. Milbourn. “How did Increased Competition affect Credit Ratings?”

Journal of Financial Economics, 101 (2011), 493-730.

Beekes, W.; P. Pope; and S. Young. “The Link between Earnings Timeliness, Earnings

Conservatism and Board Composition: Evidence from the UK.” Corporate Governance:

An International Review, 12 (2004), 47–59.

Bhojraj, S., and P. Sengupta. “Effect of Corporate Governance on Bond Ratings and

Yields: The Role of Institutional Investors and the Outside Directors.” Journal of

Business, 76 (2003), 455–475.

Bolton, P., and D. Scharfstein. “Optimal Debt Contracts and the Number of Creditors.”

Journal of Political Economy, 104 (1996), 1–25.

Boubakri, N., and H. Ghouma. “Control/Ownership Structure, Creditor Rights Protection,

30

and the Cost of Debt Financing: International Evidence.” Journal of Banking and

Finance, 34 (2010), 2481–2499.

Bradley, M., and M. Roberts. “The Structure and Pricing of Corporate Debt Covenants.”

Working Paper, University of Pennsylvania (2004).

Bushman, R.., and R. Wittenberg-Moerman. “The Role of Bank Reputation in

“Certifying” Future Performance Implications of Borrowers’ Accounting Numbers.”

Journal of Accounting Research, 50 (2012), 883-930.

Cai, Y.; D. Dhaliwal; Y. Kim; and C. Pan. “Board Interlocks and the Diffusion of

Disclosure Policy.” Review of Accounting Studies, 19 (2014), 1086–1119.

Chava, S.; D. Livdan; and A. Purnanandam. “Do Shareholder Rights Affect the Cost of

Bank Loans?” Review of Financial Studies, 22 (2009), 2973–3004.

Chen, K., and J. Zhou. “Audit Committee, Board Characteristics and Auditor Switch

Decision by Andersen’s Clients.” Contemporary Accounting Research, 24 (2007), 1085–

1117.

Chiu, P.; S. H. Teoh; and F. Tian. “Board Interlocks and Earnings Management

Contagion.” The Accounting Review, 88 (2013), 915–944.

Coffee Jr., J. “The Rise of Dispersed Ownership: The Role of Law and the State in the

Separation of Ownership and Control.” Yale Law Journal, 111 (2001), 1–82.

Conroy, S. J., and T. L. N. Emerson. “Business Ethics and Religion: Religiosity as a

Predictor of Ethical Awareness among Students.” Journal of Business Ethics, 50 (2004),

383–396.

Core, J. E.; R. W. Holthausen; and D. F. Larcker. “Corporate Governance, Chief

Executive Officer Compensation, and Firm Performance.” Journal of Financial

Economics, 51 (1999), 371–406.

Cornwall, M.; S. L. Albrecht; P. H. Cunningham; and B. L. Pitcher. “The dimensions of

religiosity: A Conceptual Model with an Empirical Test.” Review of Religious Research,

27 (1986), 226–244.

Costello, A., and R. Wittenberg-Moerman. “The Impact of Financial Reporting Quality

on Debt Contracting: Evidence from Internal Control Weakness Reports.” Journal of

Accounting Research, 49 (2011), 97–136.

DeFond, M., R. Hann; and X. Hu. “Does the Market Value Financial Expertise on the

Audit Committees of Boards of Directors?” Journal of Accounting Research, 43 (2005),

154-194.

31

Dehejia, R.; T. Deleire; and E. Luttmer. “Insuring Consumption and Happiness through

Religious Organizations.” Working Paper, NBER (2005).

Dichev, I., and D. Skinner. “Large-Sample Evidence on the Debt Covenant Hypothesis.”

Journal of Accounting and Economics, 40 (2002), 1091–1123.

Djankov, S.; C. McLiesh; and A. Shleifer. “Private Credit in 129 Countries.” Journal of

Financial Economics, 84 (2007), 299–329.

Drucker, S., and M. Puri. “On Loan Sales, Loan Contracting and Lending Relationships.”

Review of Financial Studies, 22 (2009), 2835–2872.

Edwards, S. “Trade Orientation, Distortions, and Growth in Developing Countries.”

Journal of Development Economics, 39 (1992), 31–58.

Eum, W. “Religion and Economic Development –A Study on Religious Variables

Influencing GDP Growth over Countries.” Working Paper. University of California,

Berkeley (2011).

Fama, E. F., and J. D. MacBeth. “Risk, Return, and Equilibrium: Empirical Tests.” The

Journal of Political Economy, 81(1973), 607–636.

Ferris, S.; M. Jagannathan; and A. Pritchard. “Too Busy to Mind the Business?

Monitoring by Directors with Multiple Board Appointments, Journal of Finance, 58

(2003), 1087–1111.

Fich, E. M., and A. Shivdasani. “Are Busy Boards Effective Monitors?” Journal of

Finance, 61 (2006), 689–724.

Fields, P.; D. R. Fraser; and A. Subrahmanyam. “Board Quality and the Cost of Debt

Capital: The Case of Bank Loans.” Journal of Banking and Finance, 36 (2012), 1536–

1547.

Garcia Lara, J. M.; B. Garcia Osma; and F. Penalva. “Accounting Conservatism and

Corporate Governance.” Review of Accounting Studies, 14 (2009), 161–201.

Ghoul, S.; O. Guedhami; Y. Ni; J. Pittman; and S. Saadi. “Does Religion Matter to Equity

Pricing?” Journal of Business Ethics, 111 (2012), 491–518.

Giannett, M., and Y. Yafeh. “Do Cultural Differences between Contracting Parties

Matter? Evidence from Syndicated Bank Loans.” Management Science, 58 (2012), 365–

383.

Graham, J.; S. Li; and J. Qiu. “Corporate Misreporting and Bank Loan Contracting.”

Journal of Financial Economics, 89 (2008), 44–61.

32

Grullon, G.; G. Kanatas; and J. Weston. “Religion and Corporate (Mis)Behavior.”

Working Paper, Rice University (2010).

Hermalin, B., and M. Weisbach. “Boards of Directors as an Endogenously Determined

Institution: A Survey of the Economic Literature.” Economic Policy Review, 9 (2003), 7–

26.

Hofstede, G. Culture’s Consequences: Comparing Values, Behaviors, Institutions, and

Organizations across Nations. Thousand Oaks, CA: Sage Publications (2001).

Huang, H.; G. J. Lobo; C. Wang; and J. Zhou. “Attention Please: How Does Directors’

Uneven Attention among Multiple Directorships Affect Firms’ Earnings Management.”

Working Paper, Yeshiva University, University of Houston, University of Kentucky, and

University of Hawaii at Manoa (2014).

Iannaccone, L. “Introduction to the Economics of Religion.” Journal of Economic

Literature, 36 (1998), 1465–1496.

Ivashina, V., and D. Scharfstein. “Bank Lending during the Financial Crisis of 2008.”

Journal of Financial Economics, 97 (2010), 319–338

Jensen, M. C., and W. H. Meckling. “Theory of the Firm: Managerial Behavior, Agency

Costs and Ownership Structure.” Journal of Financial Economics, 3 (1976), 305–360.

Kanagaretnam, K.; C. Y. Lim; and G. J. Lobo. “Effects of National Culture on Earnings

Quality of Banks.” Journal of International Business Studies, 42 (2011), 853–874.

Klein, A. “Audit Committee, Board of Director Characteristics, and Earnings

Management.” Journal of Accounting and Economics, 33 (2002), 375–400.

Krishnan, J.; Y. Wen; and W. Zhao. “Legal Expertise on Corporate Audit Committees and

Financial Reporting Quality.” The Accounting Review, 86 (2011), 2099–2031.

Lagace, M. “Can Religion and Business Learn from Each Other?” Working Paper,

Harvard University (2001).

La Porta, R.; F. Lopez-de-Silanes; A. Shleifer; and R. W. Vishny. “Legal Determinants of

External Finance.” Journal of Finance, 52 (1997), 1131–1150.

La Porta, R.; F. Lopez-de-Silanes; A. Shleifer; and R. W. Vishny. “Law and Finance.”

Journal of Political Economy, 106 (1998), 1113 –1155.

La Porta, R.; F. Lopez-de-Silanes; A. Shleifer; and R. W. Vishny. “The Quality of

Governance.” Journal of Law, Economics, and Organization, 15 (1999), 222–279.

Larcker, D.; S. Richardson; and A. Tuna. “Corporate Governance, Accounting Outcomes,

33

and Organizational Performance.” The Accounting Review, 82 (2007), 963–1008.

Li, N. “Negotiated Measurement Rules in Debt Contracts.” Journal of Accounting

Research, 48 (2010): 1103–1144.

Longenecker, J. G.; J. A. McKinney; and C. W. Moore. “Religious Intensity, Evangelical

Christianity, and Business Ethics: An Empirical Study.” Journal of Business Ethics, 55

(2004), 373–386.

Mansi, S.; W. Maxwell; and D. Miller. “Analyst Forecast Characteristics and the Cost of

Debt.” Review of Accounting Studies, 16 (2011), 116–142.

Masulis, R. W., and S. Mobbs. “Independent Director Incentives: Where Do Talented

Directors Spend Their Limited Time and Energy?” Journal of Financial Economics, 111

(2014a), 406–429.

Masulis, R. W., and S. Mobbs. “Independent Director Reputation Concerns.” Working

Paper, University of New South Wales and University of Alabama (2014b).

McGuire, S.; T. Omer; and N. Sharp. “The Impact of Religion on Financial Reporting

Irregularities.” The Accounting Review, 87 (2012), 645–673.

McCleary, R. M., and R. J. Barro. “Religion and Economy.” Journal of Economic

Perspectives, 20 (2006), 49–72.

McMullen, D., and K. Raghunandan. “Enhancing Audit Committee Effectiveness.”

Journal of Accountancy, 182 (1996), 79–81.

Menon, K., and J. D. Williams. “The Use of Audit Committee for Monitoring.” Journal

of Accounting and Public Policy, 13 (1994), 121–139.

Merriman, S. Religion and the Law in America: An Encyclopedia of Personal Belief and

Public Policy. Santa Barbara, CA: ABC-CLIO (2007).

Merton, R. C. “Theory of Rational Option Pricing.” Bell Journal of Economics and

Management Science, 4 (1973), 141–183.

Milbourn, T., and B. Becker. “How Did Increased Competition Affect Credit Ratings?”

Journal of Financial Economics, 101 (2011), 493–514.

Miller, A., and J. Hoffmann. “Risk and Religion: An Explanation of Gender Differences

in Religiosity.” Journal for the Scientific Study of Religion, 34 (1995), 63–75.

Mintz, S., and R. Morris. Ethical Obligations and Decision Making in Accounting. New

York, NY: McGraw-Hill/Irwin (2011).

34

Montalvo, J., and M. Reynal-Querol. “Ethnic Diversity and Economic Development.”

Journal of Development Economics, 76 (2005), 293–323.

Osoba, B. “Risk Preferences and the Practice of Religion: Evidence from Panel Data.”

Working Paper, West Virginia University (2003).

Parboteeah, K. P.; M. Hoegl; and J. B. Cullen. “Ethics and Religion: An Empirical Test of

a Multidimensional Model.” Journal of Business Ethics, 80 (2008), 387–398.

Qi, Y.; L. Roth; and J. Wald. “Political Rights and the Cost of Debt.” Journal of Financial

Economics, 95 (2010), 202–226.

Qian, J., and P. Strahan. “How Laws and Institutions Shape Financial Contracts: The

Case of Bank Loans.” Journal of Finance, 62 (2007), 2803–2834.

Rajan, R., and L. Zingales. “The Great Reversals: The Politics of Financial Development

in the Twentieth Century. Journal of Financial Economics, 69 (2003), 5–50.

Roe, M., and J. Siegel. “Political Instability: Its Effects on Financial Development, Its

Roots in the Severity of Economic Inequality.” Working Paper, Harvard University

(2008).

Roth, L. M., and J. C. Kroll. “Risky Business: Assessing Risk Preference Explanations

for Gender Differences in Religiosity.” American Sociological Review, 72 (2007), 205–

220.

Ryan, H., and R. Wiggins. “Whose Is in Whose Pocket? Director Compensation, Board

Independence, and Barriers to Effective Monitoring.” Journal of Financial Economics,

73 (2004), 497–524.

Sengupta, P. “Corporate Disclosure Quality and the Cost of Debt.” The Accounting

Review, 73 (1998), 459–474.

Shi, L.; R. Dharwadkar; and D. Harris. “Board Interlocks and Earnings Quality.”

Working Paper, SUNY Binghamton and Syracuse University (2014).

Shivdasani, A., and D. Yermack. “CEO Involvement in the Selection of New Board

Members: An Empirical Analysis.” Journal of Finance, 54 (1990), 1829–1853.

Smith, C. “A Perspective on Accounting-Based Debt Covenant Violations.” The

Accounting Review, 68 (1993), 289–303.

Stulz, R., and R. Williamson. “Culture, Openness, and Finance.” Journal of Financial

Economics, 70 (2003), 313–349.

Sufi, A. “Information Asymmetry and Financing Arrangements: Evidence from

35

Syndicated Loans.” Journal of Finance, 62 (2007), 629–668.

Vafeas, N. “Board Meeting Frequency and Firm Performance.” Journal of Financial

Economics, 53 (1999), 113–142.

Weaver, G. R., and B. R. Agle. “Religiosity and Ethical Behavior in Organizations: A

Symbolic Interactionist Perspective.” Academy of Management Review, 27 (2002), 77–

97.

Weber, M. The Protestant Ethic and the Spirit of Capitalism. New York, NY: Harper

Collins (1930).

Williamson, O. E. “The Institutions of Governance.” American Economic Review, 88

(1998), 75–79.

Williamson, O. E. “The New Institutional Economics: Taking Stock, Looking Ahead.”

Journal of Economic Literature, 38 (2000), 359–613.

Xie, B.; W. N. Davidson; and P. DaDalt. “Earnings Management and Corporate

Governance: The Roles of the Board and the Audit Committee.” Journal of Corporate

Finance, 3 (2003): 295–316.

Yermack, D. “Higher Market Valuation of Companies with a Small Board of Directors.”

Journal of Financial Economics, 40 (1996), 185–211.

Zahn, G.C. “The Commitment Dimension.” Sociology of Religion, 31 (1970), 203–208.

Zheng, X.; S. Ghoul; O. Guedhami; and C. Kwok. “National Culture and Corporate Debt

Maturity.” Journal of Banking and Finance, 36 (2012), 468–488.

36

Appendix A

Variable name Variable definitions and constructions

Board Characteristics

Percent_Ranked_High

Percentage of independent directors for whom this ranked

directorship is 10% larger than their smallest directorship, measured

by firm market capitalization. Source: Risk Metrics.

Percent_Ranked_low

Percentage of independent directors for whom this ranked

directorship is 10% smaller than their largest directorship, measured

by firm market capitalization. Source: Risk Metrics.

Majority_Ranked_High

Indicator variable that equals 1 if the majority of independent

directors classify this as a high ranked directorship (i.e., 10% larger

than their smallest directorship by market capitalization of the firm),

and is 0 otherwise. Source: Risk Metrics.

Majority_Ranked_low

Indicator variable that equals 1 if the majority of independent

directors classify this as a low ranked directorship (i.e., 10% smaller

than their largest directorship by firm market capitalization), and 0

otherwise. Source: Risk Metrics.

High-Low Indicator variable: 1 if Majority_Ranked_High equals1 and 0

if Majority_Ranked_Low equals 1.

Majority independent Indicator variable: 1 if the majority of directors are independent

directors, and 0 otherwise. Source: Risk Metrics.

CEO ownership Percentage of common shares outstanding held by the CEO at year-

end, including stock options. Source: Risk Metrics.

CEO ownership squared Square of percentage of common shares outstanding held by the CEO

at year-end, including stock options. Source: Risk Metrics.

Independent ownership Percentage of common shares outstanding held by the independent

directors at year-end, including stock options. Source: Risk Metrics

Audit Committee

Characteristics

Audit_Percent_Ranked_High

Percentage of independent audit committee members for whom this

ranked directorship is 10% larger than their smallest directorship,

measured by firm market capitalization. Source: Risk Metrics.

Audi_ Percent_Ranked_Low

Percentage of independent audit committee members for whom this

ranked directorship is 10% smaller than their largest directorship,

measured by firm market capitalization. Source: Risk Metrics.

Audit_Majority_Ranked_High

Indicator variable that equals 1 if the majority of independent outside

directors in the audit committee classify this as a high ranked

directorship (i.e., 10% larger than their smallest directorship by firm

market capitalization), and 0 otherwise. Source: Risk Metrics.

Audit_Majority_Ranked_Low

Indicator variable that equals 1 if the majority of independent outside

directors in the audit committee classify this as a low ranked

directorship (i.e., 10% smaller than their largest directorship by firm

market capitalization), and 0 otherwise. Source: Risk Metrics.

Bank Loan Characteristics

Ln(All-in-drawn) Natural logarithm of the all-in drawn spread of each facility. Source:

DealScan

Ln (Facility amount) Natural logarithm of the actual amount (in million) of the facility

committed by the facility's lender pool. Source: DealScan

Ln(Maturity) Natural logarithm of the number of the months to maturity. Source:

DealScan

Secured Indicator variable that equals 1 if the facility has secured asset, and 0

otherwise. Source: DealScan

Number of total covenants Number of total covenants, measured as the sum of number of

general covenants and number of financial covenants. Source:

37

DealScan

Number of general covenants Number of general covenants. Source: DealScan

Number of financial covenants Number of financial covenants. Source: DealScan

Performance pricing Indicator variable that equals 1 if the loan contract includes a

performance pricing provision, and 0 otherwise. Source: DealScan

Number of lenders Natural log of the number of lenders in a loan deal. Sources:

DealScan

Loan purpose Indicator variable that equals 1 for primary loan purposes, including

acquisition line, commercial paper backup, corporate purposes, debt

repayment, takeover, working capital, and other purposes, and 0

otherwise. Source: DealScan

Loan type Indicator variable that equals 1 for loan types, including 364-day

facility, delay draw term loan, revolver, and term loan, and 0

otherwise. Source: DealScan

Syndicate concentration Herfindahl index calculated using each syndicate member’s share in

the loan. Source: DealScan

Annual fee Natural log of the number of basis points of a facility commitment

amount that a borrower is required to pay on an annual basis,

regardless of any loan outstanding. Source: DealScan

Firm characteristics

Ln (Total assets) Natural log of the book value of assets at the beginning of the fiscal

year. Source: Compustat

Leverage Total liabilities divided by total assets, measured at the beginning of

the fiscal year. Source: Compustat

Operating cash flow volatility Cash flow volatility, measured as the standard deviation of quarterly

cash flows from operations (change in quarterly Compustat data item

108) divided by total assets (Compustat data item 6) over the past

five fiscal years. Source: Compustat

Return on assets Income before extraordinary items divided by total assets, measured

at the beginning of the fiscal year. Source: Compustat

Tangibility Property, plant, and equipment (PP&E) divided by total assets,

measured at the beginning of the fiscal year. Source: Compustat

Sales growth Change in sales from prior fiscal year divided by prior fiscal year

sales. Source: Compustat

Z-score Modified Altman (1968) Z-score=(1.2 working capital + 1.4 etained

earnings + 3.3EBIT + 0.999sales)/total assets. We use a modified Z-

score, which does not include the ratio of market value of equity to

book value of total debt, measured at the beginning of the fiscal year,

because a similar term, market-to-book, is included in the regressions

as a separate variable. Source: Compustat

MB Ratio of market value of equity to book value of equity, measured at

the beginning of the fiscal year. Source: Compustat

Macroeconomic Factors

Credit spread Difference in the yield between BAA-rated and AAA-rated corporate

bonds measured one month before the loan becomes active. Source:

Federal Reserve Board of Governors

Term spread Difference in the yield between ten-year and two-year U.S. Treasury

bonds measured one month before the loan becomes active.

Source:Federal Reserve Board of Governors

Bonds characteristics

Bond Rating Coded as follows (Becker and Milbourn, 2011; (Table 2, page 500):

AAA as 28, AA+ as 26, AA as 25, AA- as 24, A+ as 23, A as 22, A-

as 21, BBB+ as 20, BBB as 19, BBB- as 18, BB+ as 17, BB as 16,

38

BB- as 15, B+ as 14, B as 13, B- as 12, CCC as 11, CC as 7, and C as

4. Source: Mergent Fixed Income Securities Database.

Ln (Bond Maturity) Natural log of number of the days to maturity (maturity date minus

offering date). Source: Mergent Fixed Income Securities Database.

Additional Firm

Characteristics for Bonds

Regressions

Idiosyncratic Risk Standard deviation of the firm’s abnormal return over the calendar

year, which equals daily stock return minus value-weighted

market return. To calculate this variable, we require at least 50

trading days each year for a given stock. Source: CRSP.

Firm Risk ROA volatility, measured as the standard deviation of quarterly

Income Before Extraordinary Items divided by total assets over the

past five fiscal years. Source: Compustat.

39

Table 1: Sample Distribution and Descriptive Statistics

Panel A: Sample Development

The sample consists of 8,512 firm-year observations from 1998 to 2011. Variable definitions are in the

Appendix

Number of firm

years

Total firm-year independent director observations available in Risk Metrics and

merged with DealScan 10,996

Less:

Observations from financial services and utilities industries (1,348)

Observations with bridge loans and non-fund-based facilities, such as leases and

standby letters of credit (325)

Observations with insufficient data to calculate control variables (811)

Final sample 8,512

Panel B: Distribution by year

Year Frequency Percentage Cumulative

1998 619 7.270 7.270

1999 676 7.940 15.21

2000 800 9.400 24.61

2001 1,187 13.95 38.56

2002 641 7.530 46.09

2003 702 8.250 54.34

2004 724 8.510 62.84

2005 600 7.050 69.89

2006 710 8.340 78.23

2007 359 4.220 82.45

2008 248 2.910 85.36

2009 398 4.680 90.04

2010 696 8.180 98.21

2011 152 1.790 100

Total 8,512 100

40

Panel C: Descriptive statistics

Variables N Mean Std. P25 Median P75

Bank Loan Characteristics

All-in-drawn 8125 175.50 149.10 62.50 150.00 250.00

Ln(All-indrawn) 8125 4.798 0.921 4.135 5.011 5.521

Maturity (months) 8293 45.63 22.49 25.00 59.00 60.00

Ln(Maturity) 8293 3.631 0.698 3.219 4.077 4.094

Facility amount(millions) 8512 506.60 745.60 100.00 250.00 550.00

Ln (Facility amount) 8512 5.494 1.262 4.605 5.521 6.310

Secured 8512 0.392 0.488 0.000 0.000 1.000

Number of total covenants 8512 3.827 3.823 0.000 3.000 6.000

Number of general covenants 8512 2.344 2.704 0.000 1.000 3.000

Number of financial covenants 8512 1.483 1.432 0.000 2.000 2.000

Performance Pricing 8512 0.487 0.500 0.000 0.000 1.000

Number of lenders 8504 10.02 9.00 4.00 8.00 13.00

Board Characteristics

Percent_Ranked_High 8512 0.192 0.245 0.000 0.111 0.333

Percent_Ranked_Low 8512 0.242 0.280 0.000 0.167 0.375

Majority_Ranked_High 8512 0.093 0.291 0.000 0.000 0.000

Majority_Ranked_Low 8512 0.123 0.328 0.000 0.000 0.000

Majority Independent 8452 0.835 0.371 1.000 1.000 1.000

CEO ownership 8512 0.025 0.063 0.001 0.007 0.020

CEO ownership squared 8512 0.005 0.030 0.000 0.000 0.000

Independent ownership 8512 0.011 0.043 0.000 0.002 0.005

Firm characteristics

Ln (Total assets) 8512 7.802 1.454 6.731 7.622 8.803

Leverage 8512 0.587 0.202 0.461 0.580 0.702

Operating cash flow volatility 8512 0.041 0.036 0.015 0.027 0.058

Return on assets 8512 0.035 0.091 0.013 0.046 0.081

Tangibility 8512 0.574 0.357 0.288 0.508 0.804

Z-score 8512 3.353 2.580 1.804 2.746 4.151

MB 8512 3.046 3.831 1.339 2.129 3.431

Credit spread 8512 1.011 0.353 0.810 0.910 1.130

Term spread 8512 1.256 1.025 0.190 1.520 2.190

41

Table 2: Relation between Directorship Importance and Cost of Debt

This table presents the OLS estimation results of the impact of directorship importance on the cost of bank

loans. The following model is used in Panel A: Ln(All-in-drawn) = β0 + β1Percent_Ranked_High

(Majority_Ranked_High) + β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The following

model is used in Panel B: Ln(All-in-drawn) = β0 + β1Percent_Ranked_High (Majority_Ranked_High) +

β2Percent_Ranked_Low (Majority_Ranked_Low) + β3 Majority Independent + β4Percent_Ranked_High

(Majority_Ranked_High)* Majority Independent + β5Percent_Ranked_Low (Majority_Ranked_Low)*

Majority Independent + Controls + ε. The dependent variable is Ln (All-in-drawn). Panel B presents the

results for testing the interaction between directorship importance and board independence with the

dependent variable Ln (All-in-drawn). All models include year, industry, loan purpose, and loan type fixed

effects. The t-statistics reported in parentheses are based on standard errors that are heteroskedasticity

robust. To conserve space, we do not report the coefficient estimates for the year, industry, loan purpose,

and loan type dummies. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% level (two-

tailed), respectively. All variables are defined in Appendix A.

42

Panel A: Directorship Importance and Cost of Bank Loans (Ln (All-in-drawn))

(1) (2) (3) (4) (5) (6)

Ln(All-in-

drawn)

Ln(All-in-

drawn)

Ln(All-in-

drawn)

Ln(All-in-

drawn)

Ln(All-in-

drawn)

Ln(All-in-

drawn)

Percent_Ranked_High -0.731*** -0.155*** -0.161***

(-19.99) (-4.62) (-5.04)

Percent_Ranked_Low 0.341*** 0.073*** 0.065***

(12.60) (2.98) (2.78)

Majority_Ranked_High -0.333*** -0.040 -0.069***

(-11.35) (-1.54) (-2.77)

Majority_Ranked_Low 0.239*** 0.045** 0.043**

(10.26) (2.20) (2.18)

Majority Independent -0.093*** -0.061*** -0.049*** -0.095*** -0.060*** -0.048***

(-4.34) (-3.36) (-2.80) (-4.41) (-3.30) (-2.75)

CEO Ownership 1.018*** 0.337* 0.192 1.207*** 0.334* 0.185

(4.09) (1.71) (0.99) (4.76) (1.69) (0.96)

CEO Ownership Square -1.703*** -0.535 -0.312 -1.950*** -0.507 -0.282

(-3.37) (-1.45) (-0.84) (-3.88) (-1.37) (-0.76)

Independent Ownership -0.203 -0.351** -0.309** -0.122 -0.338** -0.299**

(-1.20) (-2.56) (-2.21) (-0.69) (-2.44) (-2.12)

Ln (Facility amount) -0.101*** -0.100***

(-12.36) (-12.33)

Ln(Maturity) 0.000 -0.001

(0.01) (-0.04)

Number of total covenants 0.047*** 0.047***

(24.16) (24.25)

Number of lenders -0.002** -0.002**

(-2.37) (-2.40)

Performance Pricing -0.185*** -0.186***

(-13.56) (-13.57)

Ln (Total Assets) -0.211*** -0.121*** -0.222*** -0.131***

(-31.50) (-14.67) (-35.61) (-16.39)

Leverage 0.502*** 0.453*** 0.508*** 0.458***

(10.88) (10.20) (10.91) (10.25)

Return on assets -1.660*** -1.382*** -1.676*** -1.399***

(-18.09) (-15.50) (-18.20) (-15.62)

Operating cash flow volatility 0.0450 0.271 0.0470 0.271

(0.24) (1.52) (0.25) (1.52)

Tangibility -0.121*** -0.099*** -0.120*** -0.098***

(-5.29) (-4.48) (-5.24) (-4.44)

Z-score -0.043*** -0.043*** -0.045*** -0.044***

(-10.91) (-11.05) (-11.24) (-11.30)

MB -0.017*** -0.016*** -0.018*** -0.017***

(-8.59) (-8.57) (-9.10) (-9.05)

Credit_ Spread 0.247*** 0.212*** 0.249*** 0.214***

(10.59) (9.35) (10.64) (9.39)

Term_ Spread 0.164*** 0.150*** 0.165*** 0.151***

(14.65) (13.26) (14.75) (13.33)

Intercept 4.330*** 5.636*** 5.485*** 4.271*** 5.705*** 5.546***

(57.13) (55.74) (50.38) (55.27) (56.91) (51.41)

No. of observations 8066 8066 7852 8066 8066 7852

Adjusted R2 0.539 0.669 0.701 0.514 0.668 0.700

Test of β1=β2

F

524.73***

27.93***

30.21***

224.70***

6.40**

11.99***

43

Panel B: Interaction between Directorship Importance and Board Independence

(1) (2)

Ln(All-in-drawn) Ln(All-in-drawn)

Percent_Ranked_High 0.002

(0.05)

Percent_Ranked_ Low -0.040

(-0.98)

Majority_Ranked_High 0.055

(1.14)

Majority_Ranked_Low -0.015

(-0.39)

Majority Independent -0.036 -0.040**

(-1.55) (-2.05)

Percent_Ranked_High* Majority Independent -0.244***

(-4.12)

Percent_Ranked_ Low* Majority Independent 0.146***

(3.02)

Majority_Ranked_High* Majority Independent -0.160***

(-2.94)

Majority_Ranked_Low* Majority Independent 0.071

(1.60)

CEO Ownership 0.183 0.196

(0.95) (1.01)

CEO Ownership Square -0.246 -0.275

(-0.67) (-0.75)

Independent Ownership -0.319** -0.310**

(-2.31) (-2.21)

Ln (Facility amount) -0.100*** -0.100***

(-12.31) (-12.28)

Ln(Maturity) -0.001 -0.002

(-0.06) (-0.10)

Number of total covenants 0.047*** 0.047***

(24.19) (24.25)

Number of lenders -0.002** -0.002**

(-2.36) (-2.40)

Performance Pricing -0.187*** -0.187***

(-13.69) (-13.58)

Ln (Total Assets) -0.117*** -0.130***

(-14.18) (-16.34)

Leverage 0.433*** 0.448***

(9.76) (10.05)

Return on assets -1.371*** -1.395***

(-15.50) (-15.66)

Operating cash flow volatility 0.287 0.270

(1.61) (1.51)

Tangibility -0.101*** -0.098***

(-4.55) (-4.43)

Z-score -0.043*** -0.044***

(-11.10) (-11.33)

MB -0.016*** -0.017***

(-8.39) (-9.11)

Credit_ Spread 0.215*** 0.214***

(9.47) (9.44)

Term_ Spread 0.150*** 0.151***

(13.35) (13.45)

44

Intercept 5.468*** 5.547***

(50.20) (51.52)

No. of observations 7852 7852

Adjusted R2 0.702 0.700

Test of β4=β5

F 24.13*** 9.52***

45

Table 3: Relation between Directorship Importance and Loan Maturity

This table presents the OLS estimation results in Panel A and Poisson regression results in Panel B of the

impact of directorship importance on the loan maturity. The following model is used:

Ln(Maturity) (Maturity) = β0 + β1Percent_Ranked_High (Majority_Ranked_High) +

β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The dependent variable is Ln (Maturity)

in Panel A and Maturity in Panel B. All models include year, industry, loan purpose, and loan type fixed

effects. The t-statistics and z-statistics reported in parentheses are based on standard errors that are

heteroskedasticity robust. To conserve space, we do not report the coefficient estimates for the year,

industry, loan purpose, and loan type dummies. *, **, and *** indicate statistical significance at the 10%,

5%, and 1% levels (two-tailed), respectively. All variables are defined in Appendix A.

46

Panel A: Directorship Importance and Loan Maturity (Ln (Maturity))

(1) (2) (3) (4) (5) (6)

Ln(Maturity) Ln(Maturity) Ln(Maturity) Ln(Maturity) Ln(Maturity) Ln(Maturity)

Percent_Ranked_ High 0.051*** 0.048** 0.052**

(2.58) (2.26) (2.45)

Percent_Ranked_Low 0.003 -0.007 -0.003

(0.17) (-0.39) (-0.17)

Majority_Ranked_High 0.046*** 0.050*** 0.051***

(3.20) (3.41) (3.48)

Majority_Ranked_Low -0.009 -0.015 -0.015

(-0.56) (-1.00) (-1.00)

Majority Independent -0.032** -0.029** -0.031** -0.031** -0.028** -0.030**

(-2.49) (-2.28) (-2.45) (-2.41) (-2.23) (-2.39)

CEO Ownership 0.093 0.131 0.111 0.080 0.133 0.109

(0.62) (0.89) (0.76) (0.53) (0.90) (0.75)

CEO Ownership Square -0.067 -0.175 -0.146 -0.046 -0.178 -0.143

(-0.26) (-0.69) (-0.57) (-0.17) (-0.70) (-0.56)

Independent Ownership 0.060 0.031 0.015 0.058 0.030 0.014

(0.78) (0.40) (0.21) (0.75) (0.39) (0.19)

Ln (Facility amount) 0.063*** 0.062***

(9.69) (9.65)

Number of total covenants 0.002 0.0010

(1.04) (0.95)

Number of lenders 0.001 0.0010

(1.15) (1.17)

Performance Pricing 0.017* 0.019*

(1.70) (1.84)

Ln (Total Assets) -0.010** -0.049*** -0.009** -0.048***

(-2.29) (-8.34) (-2.26) (-8.39)

Leverage 0.149*** 0.117*** 0.151*** 0.121***

(4.83) (3.86) (4.91) (3.99)

Return on assets 0.560*** 0.441*** 0.565*** 0.445***

(8.10) (6.45) (8.16) (6.50)

Operating cash flow volatility -0.463*** -0.484*** -0.455*** -0.476***

(-3.45) (-3.66) (-3.39) (-3.60)

Tangibility -0.021 -0.022 -0.022 -0.022

(-1.37) (-1.44) (-1.39) (-1.45)

Z-score -0.012*** -0.011*** -0.012*** -0.011***

(-4.48) (-4.18) (-4.52) (-4.21)

MB -0.003*** -0.003*** -0.003*** -0.004***

(-2.59) (-2.85) (-2.61) (-2.88)

Credit_ Spread -0.139*** -0.126*** -0.139*** -0.125***

(-7.58) (-7.09) (-7.56) (-7.07)

Term_ Spread -0.033*** -0.029*** -0.033*** -0.029***

(-4.33) (-3.86) (-4.35) (-3.88)

Intercept 2.497*** 2.701*** 2.638*** 2.500*** 2.697*** 2.633***

(34.04) (32.34) (31.79) (34.08) (32.40) (31.87)

No. of observations 8236 8236 8228 8236 8236 8228

Adjusted R2 0.706 0.716 0.722 0.706 0.716 0.722

Test of β1=β2

F 2.88* 3.52* 3.52* 6.36** 8.92*** 9.35***

47

Panel B: Directorship Importance and Loan Maturity (Maturity)

(1) (2) (3) (4) (5) (6)

Maturity Maturity Maturity Maturity Maturity Maturity

Percent_Ranked_High 0.055*** 0.047** 0.050***

(3.24) (2.50) (2.73)

Percent_Ranked_Low 0.006 -0.003 -0.000

(0.41) (-0.22) (-0.01)

Majority_Ranked_High 0.057*** 0.055*** 0.055***

(4.18) (3.98) (3.95)

Majority_Ranked_Low -0.001 -0.008 -0.008

(-0.09) (-0.62) (-0.63)

Majority Independent -0.028** -0.025** -0.025** -0.027** -0.023** -0.024**

(-2.55) (-2.28) (-2.31) (-2.41) (-2.17) (-2.20)

CEO Ownership 0.002 0.039 0.033 -0.005 0.045 0.036

(0.02) (0.34) (0.30) (-0.04) (0.39) (0.32)

CEO Ownership Square 0.017 -0.082 -0.070 0.030 -0.090 -0.073

(0.09) (-0.42) (-0.36) (0.15) (-0.46) (-0.38)

Independent Ownership 0.046 0.018 -0.000 0.046 0.017 -0.001

(0.71) (0.27) (-0.00) (0.71) (0.26) (-0.02)

Ln (Facility amount) 0.057*** 0.057***

(11.34) (11.28)

Number of total covenants 0.002 0.002

(1.59) (1.47)

Number of lenders 0.000 0.000

(0.73) (0.77)

Performance Pricing -0.009 -0.008

(-1.09) (-0.96)

Ln (Total Assets) -0.007** -0.040*** -0.007** -0.039***

(-2.04) (-8.34) (-2.00) (-8.41)

Leverage 0.140*** 0.113*** 0.140*** 0.114***

(5.77) (4.73) (5.80) (4.81)

Return on assets 0.438*** 0.346*** 0.445*** 0.352***

(7.96) (6.36) (8.12) (6.49)

Operating cash flow volatility -0.409*** -0.424*** -0.400*** -0.416***

(-3.71) (-3.92) (-3.63) (-3.85)

Tangibility -0.023* -0.026** -0.024* -0.026**

(-1.86) (-2.07) (-1.87) (-2.07)

Z-score -0.010*** -0.009*** -0.010*** -0.009***

(-4.41) (-4.21) (-4.49) (-4.27)

MB -0.002** -0.003** -0.002** -0.003**

(-2.22) (-2.41) (-2.23) (-2.41)

Credit_ Spread -0.145*** -0.131*** -0.145*** -0.130***

(-9.95) (-9.21) (-9.92) (-9.18)

Term_ Spread -0.033*** -0.031*** -0.033*** -0.031***

(-5.54) (-5.27) (-5.57) (-5.30)

Intercept 2.498*** 2.682*** 2.617*** 2.499*** 2.677*** 2.612***

(48.64) (43.43) (42.10) (48.55) (43.67) (42.33)

No. of observations 8236 8236 8228 8236 8236 8228

Pseudo R2 0.465 0.474 0.480 0.465 0.474 0.480

Test of β1=β2

Chi2 4.47** 4.08** 4.28** 9.63*** 11.06*** 11.22***

48

Table 4: Relation between Directorship Importance and Number of Covenants

This table presents the Poisson regression results of the impact of multiple-directorship on the number of

covenants. The following model is used: Number of Total Covenants (Number of General Covenants or

Number of Financial Covenants) = β0 + β1Percent_Ranked_High (Majority_Ranked_High) +

β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The dependent variable is Number of total

covenants in Panel A, Number of general covenants in Panel B, and Number of financial covenants in Panel

C. All models include year, industry, loan purpose, and loan type fixed effects. The z-statistics reported in

the parentheses are based on standard errors that are heteroskedasticity robust. To conserve space, we do

not report the coefficient estimates for the year, industry, loan purpose, and loan type dummies. *, **, and

*** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively. All variables

are defined in Appendix A.

49

Panel A: Directorship Importance and Number of Total Covenants

(1) (2) (3) (4) (5) (6)

Number of

Total

Covenants

Number of

Total

Covenants

Number of

Total

Covenants

Number of

Total

Covenants

Number of

Total

Covenants

Number of

Total

Covenants

Percent_Ranked_High -0.359*** -0.063 -0.079*

(-7.27) (-1.24) (-1.85)

Percent_Ranked_Low 0.108*** 0.005 -0.017

(2.98) (0.15) (-0.52)

Majority_Ranked_High -0.179*** -0.026 -0.096**

(-4.21) (-0.62) (-2.51)

Majority_Ranked_Low 0.075** -0.001 -0.017

(2.42) (-0.04) (-0.62)

Majority Independent -0.049* -0.024 -0.015 -0.052* -0.024 -0.017

(-1.82) (-0.93) (-0.64) (-1.94) (-0.93) (-0.75)

CEO Ownership 1.187*** 0.788** 0.296 1.247*** 0.784** 0.277

(3.52) (2.32) (0.98) (3.69) (2.31) (0.91)

CEO Ownership Square -2.239*** -1.526* -0.568 -2.291*** -1.507* -0.538

(-2.64) (-1.82) (-0.77) (-2.68) (-1.80) (-0.74)

Independent Ownership 0.170 -0.102 0.090 0.194 -0.102 0.090

(0.85) (-0.49) (0.47) (0.98) (-0.49) (0.47)

Ln (Facility amount) 0.054*** 0.054***

(4.70) (4.73)

Ln(Maturity) 0.015 0.016

(0.59) (0.64)

Prior Leader -0.016 -0.016

(-0.94) (-0.92)

Number of lenders 0.012*** 0.012***

(9.54) (9.54)

Secured 0.828*** 0.830***

(37.31) (37.36)

Ln (Total Assets) -0.155*** -0.145*** -0.159*** -0.147***

(-15.70) (-11.90) (-17.45) (-12.49)

Leverage 0.442*** 0.204*** 0.442*** 0.205***

(7.46) (3.83) (7.45) (3.86)

Return on assets 0.124 0.591*** 0.118 0.580***

(1.04) (5.39) (0.99) (5.28)

Operating cash flow volatility -1.463*** -1.316*** -1.465*** -1.328***

(-4.89) (-5.02) (-4.90) (-5.06)

Tangibility -0.104*** -0.059* -0.104*** -0.062*

(-2.83) (-1.82) (-2.85) (-1.91)

Z-score -0.024*** -0.003 -0.025*** -0.003

(-4.30) (-0.73) (-4.41) (-0.67)

MB -0.001 0.000 -0.002 -0.000

(-0.56) (0.04) (-0.71) (-0.06)

Credit_ Spread 0.019 0.026 0.019 0.025

(0.56) (0.90) (0.57) (0.88)

Term_ Spread 0.043*** 0.029* 0.044*** 0.029*

(2.62) (1.90) (2.64) (1.89)

Intercept 1.121*** 2.136*** 1.408*** 1.089*** 2.163*** 1.417***

(6.34) (11.41) (7.81) (6.09) (11.63) (7.91)

No. of observations 8452 8452 8228 8452 8452 8228

Pseudo R2 0.161 0.178 0.259 0.159 0.178 0.260

Test of β1=β2

Chi2 60.05*** 1.18 1.37 23.57*** 0.22 2.88*

50

Panel B: Directorship Importance and Number of General Covenants

(1) (2) (3) (4) (5) (6)

Number of

General

Covenants

Number of

General

Covenants

Number of

General

Covenants

Number of

General

Covenants

Number of

General

Covenants

Number of

General

Covenants

Percent_Ranked_High -0.359*** -0.060 -0.073

(-6.20) (-0.99) (-1.47)

Percent_Ranked_Low 0.148*** 0.0110 -0.021

(3.61) (0.27) (-0.58)

Majority_Ranked_High -0.135*** 0.021 -0.062

(-2.76) (0.43) (-1.38)

Majority_Ranked_Low 0.088** -0.010 -0.031

(2.53) (-0.27) (-1.03)

Majority Independent -0.032 -0.002 0.015 -0.033 0.000 0.013

(-1.03) (-0.05) (0.54) (-1.06) (0.01) (0.49)

CEO Ownership 1.110*** 0.796** 0.208 1.164*** 0.795** 0.192

(2.82) (2.00) (0.59) (2.97) (2.00) (0.55)

CEO Ownership Square -1.730* -1.174 -0.054 -1.753* -1.152 -0.024

(-1.85) (-1.25) (-0.07) (-1.88) (-1.24) (-0.03)

Independent Ownership 0.157 -0.116 0.127 0.178 -0.121 0.121

(0.63) (-0.45) (0.56) (0.72) (-0.48) (0.54)

Ln (Facility amount) 0.063*** 0.064***

(4.89) (4.92)

Ln(Maturity) 0.028 0.028

(0.97) (0.99)

Prior Leader -0.011 -0.011

(-0.58) (-0.56)

Number of lenders 0.011*** 0.011***

(8.39) (8.41)

Secured 1.076*** 1.077***

(39.21) (39.22)

Ln (Total Assets) -0.155*** -0.126*** -0.161*** -0.129***

(-13.61) (-9.32) (-15.43) (-9.92)

Leverage 0.534*** 0.249*** 0.533*** 0.251***

(7.70) (3.94) (7.70) (3.97)

Return on assets -0.067 0.475*** -0.073 0.462***

(-0.49) (3.79) (-0.53) (3.70)

Operating cash flow volatility -1.665*** -1.493*** -1.659*** -1.498***

(-4.63) (-4.80) (-4.61) (-4.82)

Tangibility -0.087** -0.040 -0.087** -0.044

(-2.06) (-1.07) (-2.06) (-1.16)

Z-score -0.034*** -0.007 -0.036*** -0.007

(-4.95) (-1.19) (-5.14) (-1.22)

MB -0.002 0.000 -0.002 -0.000

(-0.55) (0.14) (-0.79) (-0.01)

Credit_ Spread 0.051 0.061* 0.052 0.060*

(1.33) (1.89) (1.35) (1.88)

Term_ Spread 0.040** 0.019 0.040** 0.019

(2.09) (1.13) (2.11) (1.11)

Intercept 0.669*** 1.661*** 0.632*** 0.635*** 1.705*** 0.654***

(3.21) (7.57) (2.95) (3.02) (7.82) (3.07)

No. of observations 8452 8452 8228 8452 8452 8228

Pseudo R2 0.158 0.175 0.272 0.155 0.175 0.272

Test of β1=β2

Chi2 52.04*** 0.91 0.70 13.61*** 0.25 0.31

51

Panel C: Directorship Importance and Number of Financial Covenants

(1) (2) (3) (4) (5) (6)

Number of

Financial

Covenants

Number of

Financial

Covenants

Number of

Financial

Covenants

Number of

Financial

Covenants

Number of

Financial

Covenants

Number of

Financial

Covenants

Percent_Ranked_High -0.358*** -0.066 -0.081*

(-7.51) (-1.35) (-1.81)

Percent_Ranked_Low 0.042 -0.005 -0.013

(1.16) (-0.13) (-0.38)

Majority_Ranked_High -0.245*** -0.099** -0.142***

(-5.72) (-2.30) (-3.45)

Majority_Ranked_Low 0.053* 0.013 0.006

(1.71) (0.41) (0.22)

Majority Independent -0.077*** -0.058** -0.056** -0.083*** -0.060** -0.059**

(-2.91) (-2.21) (-2.25) (-3.14) (-2.30) (-2.40)

CEO Ownership 1.345*** 0.795** 0.458 1.413*** 0.785** 0.437

(3.89) (2.35) (1.40) (4.04) (2.31) (1.34)

CEO Ownership Square -3.206*** -2.200** -1.546* -3.298*** -2.185** -1.520*

(-3.50) (-2.54) (-1.81) (-3.53) (-2.53) (-1.79)

Independent Ownership 0.216 -0.037 0.077 0.242 -0.031 0.083

(1.16) (-0.18) (0.38) (1.30) (-0.15) (0.41)

Ln (Facility amount) 0.034*** 0.034***

(2.80) (2.82)

Ln(Maturity) -0.010 -0.008

(-0.38) (-0.30)

Prior Leader -0.024 -0.024

(-1.34) (-1.33)

Number of lenders 0.014*** 0.014***

(10.44) (10.44)

Secured 0.480*** 0.483***

(22.39) (22.51)

Ln (Total Assets) -0.155*** -0.172*** -0.156*** -0.171***

(-15.73) (-13.19) (-16.88) (-13.57)

Leverage 0.276*** 0.114** 0.274*** 0.114**

(4.55) (1.99) (4.55) (2.00)

Return on assets 0.469*** 0.776*** 0.466*** 0.770***

(3.83) (6.48) (3.82) (6.43)

Operating cash flow volatility -1.172*** -1.075*** -1.187*** -1.096***

(-4.04) (-3.89) (-4.09) (-3.96)

Tangibility -0.135*** -0.098*** -0.135*** -0.100***

(-3.67) (-2.82) (-3.68) (-2.87)

Z-score -0.013** -0.001 -0.013** -0.000

(-2.41) (-0.26) (-2.35) (-0.09)

MB -0.001 -0.000 -0.001 -0.000

(-0.39) (-0.16) (-0.37) (-0.12)

Credit_ Spread -0.039 -0.038 -0.040 -0.039

(-1.17) (-1.17) (-1.20) (-1.21)

Term_ Spread 0.048*** 0.043*** 0.048*** 0.043***

(2.99) (2.76) (3.00) (2.75)

Intercept 0.0500 1.132*** 0.869*** 0.0210 1.132*** 0.856***

(0.34) (6.57) (4.99) (0.14) (6.62) (4.96)

No. of observations 8452 8452 8228 8452 8452 8228

Pseudo R2 0.104 0.115 0.143 0.103 0.115 0.143

Test of β1=β2

Chi2 47.22*** 1.04 1.51 32.49*** 4.44** 8.86***

52

Table5: Relation between Directorship Importance and Loan Contract Terms: Changes Analysis

This table presents estimation results relating changes in directorship importance (from t-2 to t year) to

changes in bank loan contracting (from t-1 to t+1 year). The following model is used: Bank Loan

Contracting = β0 + β1Percent_Ranked_High (Majority_Ranked_High) + β2Percent_Ranked_Low

(Majority_Ranked_Low) + Controls + ε. The t-statistics reported in parentheses are based on standard

errors that are heteroskedasticity robust. *, **, and *** indicate statistical significance at the 10%, 5%, and

1% levels (two-tailed), respectively. All variables are defined in Appendix A.

(1) (2) (3) (4) (5)

Ln(All-in-

drawn) Maturity Number of

Total

Covenants

Number of

General

Covenants

Number of

Financial

Covenants

Percent_Ranked_High -0.189** 0.135* -0.455 -0.264 -0.202

(-2.29) (1.78) (-1.42) (-1.13) (-1.43)

Percent_Ranked_Low 0.096 0.102 0.913** 0.666** 0.253

(0.94) (1.22) (2.19) (2.15) (1.50)

Controls in Equation (1) YES YES YES YES YES

No. of observations 1557 1687 1675 1675 1675

Test of β1=β2

F/ Chi2 4.98** 0.09 7.37*** 6.93*** 4.71**

53

Table 6: Relation between Directorship Importance and Loan Contract Terms: Propensity Score

Matching

This table presents estimation results of the impact of directorship importance on bank loan contracting

after explicitly controlling for firm size differences using a propensity score matched control sample. Panel

A presents the probit estimation results of the following first-stage propensity score model: High_Low = β0

+ β1Ln(MV) + β2 Leverage + β3 Return on assets + β4 Operating earnings volatility + β5 Percent Female +

β6 Majority Independent + ε. Panel B reports results of the following model using this matched sample to

examine the relations between directorship importance and loan contract terms: Bank Loan Contracting = β0 + β1High_Low + Controls + ε. In Panel B, Column (1) reports OLS regression results, and Columns (2)-

(5) report Poisson regression results. In the first stage, all models include year and industry fixed effects. To

conserve space, we do not report the coefficient estimates for the year and industry dummies. The t- and z-

statistics reported in parentheses are based on standard errors that are heteroskedasticity robust. *, **, and

*** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively. All variables

are defined in Appendix A.

Panel A: First Stage Propensity Score Matching

First Stage:

High_Low

Ln(MV) 1.549***

(14.97)

Leverage -0.409

(-0.73)

Return on assets 1.022

(0.93)

Operating earnings volatility 3.529**

(2.18)

Percent Female -2.680***

(-3.68)

Majority Independent -0.540**

(-2.14)

Intercept -10.166***

(-10.51)

Industry fixed effect Included

Year fixed effect Included

No. of observations 1569

Pseudo R2 0.590

54

Panel B: Directorship Importance and Loan Contract Terms: Propensity Score Matched Control

Sample

(1) (2) (3) (4) (5)

Ln(All-in-

drawn) Maturity Number of

Total Covenants

Number of

General Covenants

Number of

Financial Covenants

High_Low -0.118** 0.032 -0.236*** -0.232** -0.251***

(-2.37) (1.25) (-2.90) (-2.38) (-3.08)

Controls in Equation (1) YES YES YES YES YES

Intercept 5.806*** 2.819*** -0.435 -0.772 -1.886**

(10.88) (12.26) (-0.55) (-0.78) (-2.47)

No. of observations 591 614 614 614 614

55

Table 7: Relation between Directorship Importance of Audit Committee and Loan Contract Terms

This table presents estimation results of the impact of audit committee directorship importance on bank

loan items based on the following model: Bank Loan Contracting =β0 + β1 Audit Percent_Ranked_High

(Audit Majority_Ranked_High) + β2 Audit Percent_Ranked_Low (Audit Majority_Ranked_Low) + Controls

+ ε. Panel A reports the relation between percentage of audit committee directorship importance and loan

contract terms. Panel B reports the relation between majority of audit committee directorship importance

and loan contract terms. Column (1) reports OLS regression results, and Columns (2)-(5) report Poisson

regression results. All models include year, industry, loan purpose, and loan type fixed effects. The t-

statistics reported in parentheses are based on standard errors that are heteroskedasticity robust. To

conserve space, we do not report the coefficient estimates for the year, industry, loan purpose, and loan type

dummies. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed),

respectively. All variables are defined in Appendix A.

Panel A: Relation between Percentage of Audit Committee Directorship Importance and Loan

Contract Terms

(1) (2) (3) (4) (5)

Ln(All-in-

drawn) Maturity Number of

Total

Covenants

Number of

General

Covenants

Number of

Financial

Covenants

Audit Percent_Ranked_High -0.042 0.022 -0.103*** -0.095** -0.108***

(-1.61) (1.47) (-2.90) (-2.28) (-2.83)

Audit Percent_Ranked_Low 0.062*** -0.018 0.008 0.008 0.004

(2.90) (-1.41) (0.28) (0.24) (0.13)

Controls in Equation (1) YES YES YES YES YES

No. of observations 7362 7714 7714 7714 7714

Test of β1=β2

F/ Chi2 8.95*** 3.67* 5.05** 3.19* 4.60**

Panel B: Relation between Majority of Audit Committee Directorship Importance and Loan

Contract Terms

(1) (2) (3) (4) (5)

Ln(All-in-

drawn) Maturity Number of

Total

Covenants

Number of

General

Covenants

Number of

Financial

Covenants

Audit Majority_Ranked_High -0.038* 0.031** -0.073** -0.068* -0.077**

(-1.66) (2.55) (-2.24) (-1.76) (-2.18)

Audit Majority_Ranked_Low 0.050** -0.017 -0.018 -0.038 0.012

(2.55) (-1.46) (-0.66) (-1.23) (0.40)

Controls in Equation (1) YES YES YES YES YES

No. of observations 7362 7714 7714 7714 7714

Test of β1=β2

F/ Chi2 7.97*** 7.48*** 1.55 0.33 3.54*

56

Table 8: Relationship between Directorship Importance and Syndicate Concentration

This table presents OLS estimation results of the impact of directorship importance on syndicate

concentration based on the following model: Syndicate Concentration = β0 + β1Percent_Ranked_High

(Majority_Ranked_High) + β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The

dependent variable is Syndicate Concentration. All models include year, industry, loan purpose, and loan

type fixed effects. The t-statistics reported in parentheses are based on standard errors that are

heteroskedasticity robust. To conserve space, we do not report the coefficient estimates for the year,

industry, loan purpose, and loan type dummies. *, **, and *** indicate statistical significance at the 10%,

5%, and 1% levels (two-tailed), respectively. All variables are defined in Appendix.

(1) (2) (3) (4)

Syndicate

Concentration

Syndicate

Concentration

Syndicate

Concentration

Syndicate

Concentration

Percent_Ranked_High -0.041*** -0.003

(-7.03) (-0.54)

Percent_RankedLow -0.001 0.003

(-0.16) (0.42)

Majority_Ranked_High -0.024*** -0.007*

(-6.69) (-1.81)

Majority_Ranked_Low 0.004 0.006

(0.73) (1.09)

Controls Board

characteristics

Controls in

Equation (1)

Board

characteristics

Controls in

Equation (1)

No. of observations 8444 8147 8444 8147

Test of β1=β2

F/ Chi2 20.55*** 0.42 17.13*** 3.40*

57

Table 9: Relation between Directorship Importance and Transaction Fees: Annual Fees

This table presents OLS estimation results of the impact of directorship importance on annual fee based on

the following model: Ln(Annual Fee) = β0 + β1Percent_Ranked_High (Majority_Ranked_High) +

β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The dependent variable is Ln(Annual

Fee). All models include year, industry, loan purpose, and loan type fixed effects. The t-statistics reported

in parentheses are based on standard errors that are heteroskedasticity robust. To conserve space, we do not

report the coefficient estimates for the year, industry, loan purpose, and loan type dummies. *, **, and ***

indicate statistical significance at the 10%, 5%, and 1% levels (two-tailed), respectively. All variables are

defined in Appendix A.

(1) (2) (3) (4)

Ln(Annual

Fee)

Ln(Annual

Fee)

Ln(Annual

Fee)

Ln(Annual

Fee)

Percent_RankedHigh -0.640*** -0.108**

(-14.08) (-2.51)

Percent_Ranked_Low 0.295*** 0.069

(6.18) (1.58)

Majority_Ranked_High -0.276*** -0.017

(-10.07) (-0.73)

Majority_Ranked_Low 0.158*** 0.030

(4.17) (0.92)

Controls Board

characteristics

Controls in

Equation (1)

Board

characteristics

Controls in

Equation (1)

No. of observations 2708 2659 2708 2659

Test of β1=β2

F/ Chi2 205.25*** 7.81*** 89.08*** 1.41

58

Table 10: Relation between Directorship Importance and Bonds Rating

This table presents OLS estimation results of the impact of directorship importance on bond rating based on

the following model: Bond Rating = β0 + β1Percent_Ranked_High (Majority_Ranked_High) +

β2Percent_Ranked_Low (Majority_Ranked_Low) + Controls + ε. The dependent variable is Bond Rating.

All models include year, industry fixed effects. The t-statistics reported in parentheses are based on

standard errors that are heteroskedasticity robust. To conserve space, we do not report the coefficient

estimates for the year, industry dummies. *, **, and *** indicate statistical significance at the 10%, 5%,

and 1% levels (two-tailed), respectively. All variables are defined in Appendix A.

(1) (2) (3) (4)

Bond Rating Bond Rating Bond Rating Bond Rating

Percent_Ranked_High 4.958*** 0.629*** (46.14) (6.13)

Percent_Ranked_Low -2.722*** -0.136

(-23.36) (-1.41) Majority_Ranked_High 1.956*** 0.323***

(35.58) (6.46) Majority_Ranked_Low -1.594*** 0.098

(-16.14) (1.29) Majority Independent 0.560*** 0.285*** 0.716*** 0.298***

(6.22) (4.34) (8.03) (4.57)

CEO Ownership -18.248*** -10.516*** -22.173*** -10.396*** (-17.04) (-12.36) (-19.75) (-12.12)

CEO Ownership Square 24.540*** 13.787*** 30.381*** 13.328*** (11.67) (7.76) (14.14) (7.39)

Independent Ownership 1.462*** 3.461*** 1.663*** 3.623***

(3.28) (8.05) (3.26) (8.27) Idiosyncratic Risk -104.446*** -105.258***

(-28.30) (-28.48) Ln (Bonds Maturity) 0.109*** 0.109***

(3.02) (3.04) Ln (Total Assets) 1.217*** 1.251***

(53.66) (60.77)

Leverage -2.024*** -2.047*** (-11.58) (-11.74)

Return on assets 4.744*** 4.842*** (9.76) (9.96)

Firm Risk -2.936** -2.666**

(-2.13) (-1.97)

Z-score 0.597*** 0.602***

(25.49) (25.71)

MB 0.016** 0.018***

(2.34) (2.67) Intercept 20.445*** 10.861*** 21.775*** 10.688***

(77.92) (21.38) (70.61) (20.96)

No. of observations 21199 11377 21199 11377 Adjusted R2 0.345 0.718 0.271 0.718

Test of β1=β2 F 2740.31*** 29.41*** 1004.89*** 5.89**