Finance Research Letters xxx (xxxx) xxxx

Contents lists available at ScienceDirect

Finance Research Letters

journal homepage: www.elsevier.com/locate/frl

Does inside debt help mitigate agency problems? The case with
investment inefficiency and payout policies
Asligul Erkana, , Trung Nguyena
⁎

a
Department of Finance, College of Business, East Carolina University, Bate Building, Greenville 27858, NC, USA

ARTICLE INFO ABSTRACT

JEL Classification: Although the literature has established that CEO inside debt induces risk aversion, it remains
G30 inconclusive whether these CEOs increase firm payout or not. While some studies show that high
G34 inside debt CEOs increase firm payout to avoid risky projects, others document that they reduce
G35 payout to preserve funds. This study addresses this contradiction by examining a sample of U.S.
G39
firms over the 2006–2018 period. Consistent with the agency theory, our results show that CEOs
with larger inside debt adjust payout policies to mitigate prior overinvestment. Specifically, they
Keywords:
increase payout in the subsequent period if their firm is already overinvested.
Agency theory
Executive compensation
Inside debt
Payout policy
Overinvestment

1. Introduction

Managerial compensation is a governance mechanism designed to alleviate agency problems. According to the agency theory,
stocks and options (hereafter inside equity) help mitigate the conflicts of interests between managers and shareholders, and motivate
managers to take on greater risk to increase shareholder wealth. In contrast, pension and deferred compensation (hereafter inside
debt), due to their unfunded and unsecured nature, help curb managerial risk appetite and prevent the expropriation of debt holders
(Jensen and Meckling, 1976; Edmans and Liu, 2011). Consistently, empirical research has shown that CEOs compensated with inside
debt adopt more conservative corporate policies. Notably, risk-averse CEOs with larger proportion of inside debt are associated with
lower probability of default (Sundaram and Yermack, 2007), lower R&D, firm risk and leverage (Cassell et al., 2012), but larger cash
holding (Liu et al., 2014), greater capital expenditures and asset tangibility (Lu-Andrews and Yu-Thompson, 2015), more usage of
short-maturity debt (Dang and Phan, 2016), and higher corporate social performance (Wu and Lin, 2019).
In the context of payout policies, however, there have been theoretical and empirical contradictions documented in the literature
that warrant further investigation. According to the wealth transfer hypothesis, inside equity creates incentives to provide share-
holders with liquidity while shifting risk to creditors, whereas inside debt discourages policies (e.g., paying out dividends) that result
in asset substitution (Edmans and Liu, 2011). Consistent with this risk-shifting view of payout policies, Srivastav et al. (2014) show
that bank CEOs with larger proportion of inside debt reduce payouts to preserve cash and lower bankruptcy risk. Using a hand-
collected sample of both financial and non-financial firms, Eisdorfer et al. (2015) also find that CEOs with greater inside debt reduce
dividend payouts but argue that the purpose is to preserve funds for their future pension benefits. In contrast, proponents of the risk-

⁎
Corresponding author.
E-mail address: [email protected] (A. Erkan).

https://doi.org/10.1016/j.frl.2020.101560
Received 20 August 2019; Received in revised form 21 April 2020; Accepted 22 April 2020
1544-6123/ © 2020 Elsevier Inc. All rights reserved.

Please cite this article as: Asligul Erkan and Trung Nguyen, Finance Research Letters, https://doi.org/10.1016/j.frl.2020.101560
A. Erkan and T. Nguyen Finance Research Letters xxx (xxxx) xxxx

reducing view argue that firm payouts are conservative policies that help prevent managers from engaging in value-destroying
projects that will hurt creditors as well (Grullon et al., 2002; DeAngelo et al., 2006). Thus, inside debt CEOs may prefer increasing
firm payouts over undertaking risky investments. Supporting this alternative explanation, Caliskan and Doukas (2015) find that
inside debt CEOs have higher propensity to pay dividends or repurchase shares. In this study, we aim to resolve the aforementioned
conflicting evidence by examining the effect of prior investment inefficiency (i.e., over or underinvestment) on the relation between
inside debt and payout policies.
According to the agency theory, investment inefficiency is considered a manifestation of existing conflict of interest among
managers, shareholders and debt holders where managers forego profitable investments or undertake uneconomic projects due to
divergent risk preferences or moral hazard (Jensen, 1986; Myers and Majluf, 1984). Hence, properly governed firms/managers are
expected to correct for pre-existing investment inefficiencies through future adjustments to corporate policies (Biddle et al., 2009;
Lara et al., 2016). Consistent with this view, Lara et al. (2016) argue that accounting conservatism improves investment efficiency
and find that conservative firms will invest more (less) in settings prone to underinvestment (overinvestment). More importantly,
investment, financing and payout decisions have been shown to be interdependent given market imperfections (e.g., Myers, 1984;
Grullon et al., 2002; Lambrecht and Myers, 2017), which implies that payout policies can be used in place of investment policies to
amend prior investment inefficiency. Building on this tranche of literature, therefore, we argue that CEOs compensated with inside
debt will adjust their firms’ payout policies in different ways to combat already existing agency problems associated with investment
inefficiency.
Specifically, we conjecture that higher inside debt CEOs at already overinvested firms will increase payouts. For firms that are
prone to underinvestment, however, CEOs compensated with greater inside debt are expected to retain the funds within the firm.
Unlike previous studies that examine and find a negative effect of CEO inside debt on overinvestment agency problems
(Eisdorfer et al., 2013; Yu-Thompson and Zhao, 2017), we use investment inefficiency as a pre-existing condition and investigate how
CEOs with different compensation packages adjust their payout policies.1 That is, our study complements Eisdorfer et al. (2013) and
Yu-Thompson and Zhao (2017) by demonstrating the channels through which inside debt CEOs remedy overinvestment (under-
investment).
We test our empirical predictions on a U.S. sample of 10,168 firm-year observations over the 2006–2018 period, which covers the
most recent time period and encompasses all industries that makes our results more generalizable. In the main analysis, we follow the
literature in our choice of the main variables, CEO relative leverage (e.g., Cassell et al., 2012), net and total payouts (e.g.,
Srivastav et al., 2014) and firm overinvestment (Yu-Thompson and Zhao, 2017), as well as control for firm- and year-fixed effects.
Consistent with the agency theory, we find that inside debt CEOs increase (decrease) payout in the subsequent period when their
firms are prone to overinvestment (underinvestment), which suggests that inside debt is an effective governance mechanism that
mitigates existing agency problems at corporations. The results hold under a battery of robustness checks such as addressing potential
endogeneity concerns with the instrumental variable approach or an industry-wide measure of overinvestment (underinvestment);
employing alternative measures of CEO inside debt and firm payout; as well as using clustered standard errors at the firm level. Our
study contributes to the literature by proposing a possible explanation for the contradicting results regarding the impact of inside
debt on firm payout policies. In addition, we address the reservation towards the effectiveness of inside debt as a governance
mechanism (e.g., Anantharaman et al., 2013) by showing that CEOs, in fact, adjust their payout policies to mitigate the existing
agency problems at their firms. Finally, we add to the payout literature by providing results that are supportive of the view of payout
as a governance tool (Easterbrook, 1984). The rest of the paper is as follows. Section 2 describes research design and data. Section 3
discusses empirical results and robustness tests. Section 4 concludes the paper.

2. Research design and data

2.1. Baseline model

Consistent with previous studies, we employ the following model specification to test our hypothesis. We expect the effect of
inside debt on firm payout policies to be significantly negative for underinvested firms (β1), yet significantly positive for overinvested
firms (β1 + β2).

PAYOUTi, t =
0 + 1 CEO _RLEVi, t 1 + 2 CEO _RLEVi, t 1*OVER _INVi, t 1 + 3 OVER _INVi, t 1 + 4 CEO _AGEi, t 1+
5 CEO _V 2Di, t 1 + 6 CEO _TENUREi, t 1 + 7 SIZEi, t 1 + 8 AGEi, t 1 + 9 LEVi, t 1 + 10 MTBi, t 1+
K T
11 CASHi, t 1 + 12 ROAi, t 1 + i i =1
FIRMi + t t =1
YEARt + i, t (1)

We capture firm payout policies, the dependent variables, using (a) total dividends plus net purchase of common and preferred
stocks scaled by total assets (NET_PAYOUT), and (b) total dividends plus gross purchase of common and preferred stocks scaled by
total assets (TOT_PAYOUT) (Srivastav et al., 2014). CEO_RLEV, the main independent variable, is CEO inside debt-to-inside equity

1
Although it is not the primary objective of the study, in an unreported analysis, we examined and verified the documented negative relation
between inside debt and investment inefficiency for our sample. We find consistent evidence with Eisdorfer et al. (2013) and Yu-Thompson and
Zhao (2017). The result is available upon request.

2
A. Erkan and T. Nguyen Finance Research Letters xxx (xxxx) xxxx

ratio scaled by firm debt-to-equity ratio after logarithmic transformation (Cassell et al., 2012). As noted above, inside debt includes
pension and deferred compensation while inside equity consists of stocks and stock options. As suggested by the extant literature
(e.g., Yu-Thompson and Zhao, 2017), we use a dummy (OVER_INV) to distinguish between overinvested and underinvested firms
according to the residuals estimated from the investment model below.2
S T
INVi, t = 0 + 1 MTBi, t 1 + 2 LEVi, t 1 + 3 SIZEi, t 1 + 4 ROAi, t 1 + s INDs + t YEARt + µi, t
s =1 t =1 (2)
In Eq. (2), INV is gross capital expenditures scaled by total assets at the beginning of the year (Eisdorfer et al., 2013; Yu-
Thompson and Zhao, 2017). According to this research tranche, the residual from Eq. (2) captures the extent to which a firm deviates
from its optimal (or expected) capital investment level. Hence, the dummy OVER_INV takes the value of one if the residual in Eq. (2) is
positive (i.e., overinvested), and zero otherwise (i.e., underinvested).
To account for other firm characteristics, we add firm size (SIZE), firm age (AGE), firm leverage (LEV), firm growth opportunities
(MTB), firm cash holdings (CASH), and firm return on assets (ROA) to Eq. (1) (Srivastav et al., 2014). In terms of other CEO variables,
we include CEO age after log transformed (CEO_AGE), CEO vega to CEO delta ratio (CEO_V2D), and CEO tenure after log transformed
(CEO_TENURE) in our baseline model (Caliskan and Doukas, 2015; Yu-Thompson and Zhao, 2017). Besides, we also control for firm
(FIRM) and year (YEAR) fixed effects. Table 1 provides the detailed descriptions of the variables.

2.2. Data

The sample is constructed by merging data from Execucomp, Compustat and CRSP. Since inside debt information was made
available in 2006, the sample only contains U.S. firms for the period from 2006 to 2018. Following the literature, we exclude
observations with non-positive book value of common equity (CEQ) (e.g., Nikolic and Yan, 2014). The final sample comprises 10,168
firm-year observations.
Table 2 reports means, standard deviations, and Pearson correlation coefficients among the variables used in our analyses.
Specifically, the means of NET_PAYOUT and TOT_PAYOUT are 0.0287 and 0.0407, respectively, suggesting that for every dollar in
total assets, the average firm will have a net (total) payout of 2.87 (4.07) cents. In addition, more than 86% of the observations are
considered overinvested (OVER_INV). The mean log of firm size (SIZE) is around 8.1734 while the mean log of firm age (AGE) is
approximately 3.2510. Similar to other studies, in our sample, the mean of LEVERAGE is 0.2341 and the mean of MTB is 2.9303. Cash
and short-term investments of the average firm is about 12.76% of its total assets (CASH). For every dollar invested in total assets, the
average firm earns roughly 3.75 cents in income before extraordinary items (ROA). Regarding CEO characteristics, the average of
CEO_RLEV is 0.5018 while those of CEO_AGE, CEO_V2D, and CEO_TENURE are 4.0133, 0.1693, and 1.7805, respectively.
The correlation between NET_PAYOUT and TOT_PAYOUT is above 0.80 and statistically significant at 1%. Other pair-wise cor-
relations presented in the table suggest that multicollinearity is less likely to be a concern. This is further confirmed by the corre-
sponding variance inflated factors (VIF) of lower than 2 (not tabulated for brevity, but available upon request).

3. Empirical results

In this section, we report the regression results of our baseline model as shown in Eq. (1). In column (1) of Table 3, the coefficient
of CEO_RLEV is −0.0074 and statistically significant at 1%, indicating that for underinvested firms, CEOs with higher inside debt
compensation will reduce net payout to shareholders. However, for overinvested firms, the net effect of CEO_RLEV on firm payout
policies turns to a positive value (−0.0074 + 0.0133 = 0.0059). The corresponding F-statistic of the joint-test also suggests a p-value
of less than 1%. That is, CEOs with higher inside debt compensation at overinvested firms tend to increase net payout to shareholders.
Furthermore, as shown in column (2), the empirical evidence remains strongly consistent when we replace NET_PAYOUT with
TOT_PAYOUT. Collectively, the reported findings confirm the notion that inside debt helps alleviate agency problems, especially in
the context of overinvestment.
We also find that younger CEOs with higher vega-to-delta ratio and less tenure tend to retain funds within the firms rather than
pay out to shareholders. Similar to previous studies, we also document that bigger firms with lower leverage, higher growth op-
portunities, greater cash holdings, and higher return of assets are associated with higher payout in the next period.
In our extended analyses, we decompose net payout into dividend payout and net repurchase as in Srivastav et al. (2014), as well
as recalculate CEO relative leverage using pension only or deferred compensation only as in Wu and Lin (2019). As shown in column
(3) and (4), inside debt managers appear to correct for prior investment inefficiency issues using share repurchases more. Consistent
with Wu and Lin (2019), the effect of inside debt on firm payout policy given the pre-existing investment inefficiency is more
pronounced when only deferred compensation is used than when only pension is used. We tackle the potential endogeneity concern
in our study using two alternative approaches. First, we employ instrumental variable regressions in which industry-average of CEO
relative leverage and maximum state income tax rate are used as instruments (Cassell et al., 2012; Anantharaman et al., 2013).
Second, instead of using firm-specific information, we identify firms with prior overinvestment (underinvestment) issues based on
industry-level data proposed by previous studies (e.g., Lara et al., 2016). As shown in Tables 4 and 5, the empirical evidence holds

2
Industry fixed effects are captured using 2-digit SIC code. In addition, using sales growth instead of market-to-book ratio as in Biddle et al. (2009)
does not qualitatively change our overall findings (available upon request).

3
A. Erkan and T. Nguyen Finance Research Letters xxx (xxxx) xxxx

Table. 1
Variable descriptions.
Variable Description with COMPUSTAT Xpressfeed (if applicable)

NET_PAYOUT Total dividends (DVT) plus purchase of common and preferred stocks (PRSTKC) minus sale of common and preferred stocks (SSTK) scaled
by total assets (AT)
TOT_PAYOUT Total dividends (DVT) plus purchase of common and preferred stocks (PRSTKC) scaled by total assets (AT)
DIV Total dividends (DVT) scaled by total assets (AT)
NET_REP Purchase of common and preferred stocks (PRSTKC) minus sale of common and preferred stocks (SSTK) scaled by total assets (AT)
INV Capital expenditures (CAPX) scaled by total assets (AT) at the beginning of the year
OVER_INV A dummy equal to one if the residual from the firm-specific capital investment model is greater than zero, and zero otherwise
S T
INVi, t = 0 + 1 MTBi, t 1 + 2 LEVi, t 1 + 3 SIZEi, t 1 + 4 ROAi, t 1 + s s = 1 INDs + t t = 1 YEARt + µ i, t
SIZE Natural logarithm of total assets (AT)
AGE Natural logarithm of the number of years the firm has appeared in the COMPUSTAT database (YEAR1)
LEV Sum of long-term debts (DLTT) and short-term debts (DLC) divided by total assets (AT)
MTB Market value of equity (PRCC_F*CSHO) to book value of equity (CEQ)
CASH Cash and short-term investments (CHE) divided by total assets (AT)
ROA Income before extraordinary items (IB) divided by total assets (AT)
CEO_RLEV CEO inside debt to CEO inside equity scaled by the ratio of firm debt to equity after logarithmic transformation
CEO_AGE Natural logarithm of CEO age (AGE)
CEO_V2D The ratio of CEO vega to CEO delta
CEO_TENURE Natural logarithm of the number of years a CEO has been in the position (BECAMECEO)
IND_RLEV Industry-year average of CEO relative leverage based on SIC classification
IND_INV Industry-year average of total investments which equal capital expenditures (CAPX) plus R&D expenses (XRD) plus acquisitions (AQC)
minus net sales of property, plant and equipment (SPPE) scaled by total assets (AT) at the beginning of the year
IND_GROWTH Industry-year average of firm sales growth which equal the percentage change in sales (SALE) from the previous period to the current
period
IND_OVER_INV A dummy equal to one if the residual from the following industry-wide investment model belongs to the top quintile, and zero otherwise
IND _INVs, t = 1 + 2 IND _GROWTHs,t 1 + s, t
IND_UNDER_INV A dummy equal to one if the residual from the following industry-wide investment model belongs to the bottom quintile, and zero
otherwise
IND _INVs, t = 1 + 2 IND _GROWTHs,t 1 + s, t
TAXRATE_WAGES Maximum state income tax rate on wages retrieved from the National Bureau of Economic Research (NBER) https://users.nber.org/
~taxsim/state-rates/

Table. 2
Descriptive statistics and correlation matrix. This table reports the means and the standard deviations of as well as the correlations among the
variables used in the main empirical analysis. The final sample comprises 10,168 firm-year observations over the 2006–2018 period. Detailed
variable descriptions are provided in Table 1. ***, **, and * are used to denote significance at 1%, 5%, and 10% levels, respectively.
Variable Index Mean STD 1 2 3 4 5 6

NET_PAYOUTt 1 0.0287 0.0725 1.0000

TOT_PAYOUTt 2 0.0407 0.0619 0.8303*** 1.0000
OVER_INVt-1 3 0.8675 0.3390 0.0968*** 0.0420*** 1.0000
SIZEt-1 4 8.1734 1.7321 0.1082*** −0.0020 0.3344*** 1.0000
AGEt-1 5 3.2510 0.5987 0.1097*** 0.0631*** 0.1392*** 0.3231*** 1.0000
LEVt-1 6 0.2341 0.1654 −0.0650*** −0.0740*** −0.0150 0.1082*** 0.0469*** 1.0000
MTBt-1 7 2.9303 3.2823 0.2396*** 0.3435*** 0.0011 −0.0520*** −0.0040 0.1672***
CASHt-1 8 0.1276 0.1402 0.0615*** 0.2025*** 0.0133 −0.2480*** −0.1640*** −0.2040***
ROAt-1 9 0.0375 0.0966 0.4137*** 0.3358*** 0.0659*** 0.0669*** 0.0684*** −0.0980***
CEO_RLEVt-1 10 0.5018 0.8522 0.1727*** 0.1772*** 0.0505*** 0.0556*** 0.1988*** −0.2520***
CEO_AGEt-1 11 4.0133 0.1179 0.0059 −0.0190* 0.0539*** 0.1398*** 0.1290*** −0.0050
CEO_V2Dt-1 12 0.1693 0.3039 0.0568*** 0.0544*** 0.0709*** 0.1266*** 0.0994*** 0.0029
CEO_TENUREt-1 13 1.7805 0.8335 −0.0120 −0.0140 −0.0130 −0.0530*** −0.0580*** −0.0370***
Variable Index 7 8 9 10 11 12 13
MTBt-1 7 1.0000
CASHt-1 8 0.1607*** 1.0000
ROAt-1 9 0.2118*** 0.0073 1.0000
CEO_RLEVt-1 10 0.0328*** 0.0244** 0.1411*** 1.0000
CEO_AGEt-1 11 −0.0520*** −0.1070*** 0.0185* 0.1150*** 1.0000
CEO_V2Dt-1 12 0.0087 0.0049 0.0605*** 0.1199*** −0.0440*** 1.0000
CEO_TENUREt-1 13 −0.0030 0.0028 0.0284*** 0.0124 0.3759*** −0.1640*** 1.0000

after correcting for the endogeneity concern.

We also perform the following tests to check the sensitivity of our findings (not tabulated but available upon request). According
to Wei and Yermack (2011), CEO relative incentive may do a better job in capturing whether CEOs are induced to engage in risk-
shifting strategies that transfer value between debt and equity as predicted by the agency theory. To ensure that our results are not
sensitive to this particular issue, we perform a similar regression analysis to Eq. (1), but employ the log of CEO relative incentive as

4
 Table. 3
Baseline regressions and extended analyses. This table reports the regression results of firm payout policies on CEO relative leverage, overinvestment dummy, and their interaction controlling for other
CEO and firm characteristics. In column (1), (5), and (7), the dependent variable (NET_PAYOUT) is computed as total dividends plus purchase of common and preferred stocks minus sale of common and
A. Erkan and T. Nguyen

preferred stocks scaled by total assets. In column (2), (6), and (8), the dependent variable (TOT_PAYOUT) is estimated as total dividends plus purchase of common and preferred stocks scaled by total
assets. In column (3), the dependent variable (DIV) is total dividends scaled by total assets. In column (4), the dependent variable (NET_REP) is purchase of common and preferred stocks minus sale of
common and preferred stocks scaled by total assets. In column (5) and (6), CEO relative leverage is calculated using only CEO pension value, whereas in column (7) and (8), the estimation of CEO relative
leverage relies only on CEO deferred compensation. Detailed variable descriptions are provided in Table 1. All specification models also include firm- and year-fixed effects. Standard errors are reported in
parentheses. ***, **, and * are used to denote significance at 1%, 5%, and 10% levels, respectively.
Variable NET_PAYOUTt TOT_PAYOUTt DIVt NET_REPt NET_PAYOUTt TOT_PAYOUTt NET_PAYOUTt TOT_PAYOUTt
Pension only Deferred compensation only
(1) (2) (3) (4) (5) (6) (7) (8)

CEO_RLEVt-1 −0.0074*** −0.0070*** −0.0006 −0.0068*** −0.0072* −0.0063* −0.0084*** −0.0080***

(0.0025) (0.0021) (0.0008) (0.0024) (0.0043) (0.0036) (0.0028) (0.0023)
CEO_RLEVt-1 * OVER_INVt-1 0.0133*** 0.0125*** 0.0028*** 0.0105*** 0.0093** 0.0084** 0.0153*** 0.0144***
(0.0026) (0.0022) (0.0008) (0.0025) (0.0044) (0.0037) (0.0031) (0.0026)
OVER_INVt-1 −0.0009 −0.0026 −0.0016 0.0007 0.0029 0.0010 0.0004 −0.0014
(0.0036) (0.0030) (0.0011) (0.0035) (0.0036) (0.0030) (0.0036) (0.0030)
CEO_AGEt-1 −0.0282** −0.0219** −0.0038 −0.0245** −0.0257** −0.0196** −0.0263** −0.0199**
(0.0110) (0.0092) (0.0034) (0.0106) (0.0111) (0.0093) (0.0110) (0.0092)
CEO_V2Dt-1 −0.0043* −0.0038** −0.0006 −0.0037* −0.0040* −0.0036** −0.0042* −0.0037**
(0.0022) (0.0018) (0.0007) (0.0021) (0.0022) (0.0018) (0.0022) (0.0018)

5
CEO_TENUREt-1 0.0024** 0.0019* 0.0001 0.0023* 0.0024** 0.0019* 0.0023* 0.0018*
(0.0012) (0.0010) (0.0004) (0.0012) (0.0012) (0.0010) (0.0012) (0.0010)
SIZEt-1 0.0173*** 0.0038** −0.0017*** 0.0190*** 0.0169*** 0.0034** 0.0173*** 0.0037**
(0.0020) (0.0017) (0.0006) (0.0019) (0.0020) (0.0017) (0.0020) (0.0017)
AGEt-1 −0.0056 −0.0089 0.0045* −0.0101 −0.0054 −0.0087 −0.0055 −0.0088
(0.0081) (0.0068) (0.0025) (0.0078) (0.0081) (0.0068) (0.0081) (0.0068)
LEVt-1 −0.0936*** −0.0862*** −0.0052** −0.0885*** −0.1000*** −0.0920*** −0.0953*** −0.0878***
(0.0076) (0.0064) (0.0023) (0.0073) (0.0075) (0.0063) (0.0075) (0.0063)
MTBt-1 0.0013*** 0.0014*** 0.0002*** 0.0011*** 0.0013*** 0.0014*** 0.0013*** 0.0014***
(0.0002) (0.0002) (0.0001) (0.0002) (0.0002) (0.0002) (0.0002) (0.0002)
CASHt-1 0.0996*** 0.0808*** 0.0137*** 0.0859*** 0.0989*** 0.0801*** 0.1000*** 0.0809***
(0.0095) (0.0079) (0.0029) (0.0091) (0.0095) (0.0080) (0.0095) (0.0079)
ROAt-1 0.0880*** 0.0334*** 0.0027 0.0853*** 0.0880*** 0.0334*** 0.0882*** 0.0335***
(0.0077) (0.0064) (0.0023) (0.0074) (0.0077) (0.0065) (0.0077) (0.0064)
INTERCEPT 0.0160 0.1286*** 0.0256* −0.0096 0.0084 0.1220*** 0.0076 0.1211***
(0.0498) (0.0417) (0.0152) (0.0479) (0.0500) (0.0419) (0.0498) (0.0417)
Net effect of inside debt at firms with overinvestment (F-stat) 23.71*** 29.77*** 34.90*** 10.20*** 1.51 2.14 24.45*** 29.54***
Firm FE Yes Yes Yes Yes Yes Yes Yes Yes
Year FE Yes Yes Yes Yes Yes Yes Yes Yes
R2 (%) 55.85% 57.59% 58.03% 50.98% 55.70% 57.40% 55.86% 57.58%
N 10,168 10,168 10,168 10,168 10,164 10,164 10,166 10,166
Finance Research Letters xxx (xxxx) xxxx
A. Erkan and T. Nguyen Finance Research Letters xxx (xxxx) xxxx

Table. 4
Addressing potential endogeneity with instrumental variable analysis. This table reports the instrumental variable regression analysis of firm
payout policies on CEO relative leverage and investment inefficiency. Panel A presents the first-stage regressions of CEO relative leverage and
its interaction with overinvestment dummy on a set of exogenous instruments including industry-year average of CEO relative leverage
(IND_RLEV), maximum state tax rates on wages (TAXRATE_WAGE) and their interactions with overinvestment, as suggested by
Wooldridge (2010). Sanderson–Windmeijer χ2 and F statistics are reported for the under-identification and weak instrument tests, respectively.
Panel B shows the second-stage regression results of payout variables on the predicted values of CEO relative leverage and that of the in-
teraction term between CEO relative leverage and overinvestment. Sargan χ2 statistic and Sargan–Hansen χ2 statistic are reported for the over-
identification and endogeneity tests, respectively. Other controls from the baseline model including firm- and year-fixed effects are also used in
both stages of the analysis. Detailed variable descriptions are provided in Table 1. Standard errors are reported in parentheses. ***, **, and *
are used to denote significance at 1%, 5%, and 10% levels, respectively.
Panel A: First-stage regressions
Variable CEO_RLEVt-1 CEO_RLEVt-1*OVER_INVt-1
(1) (2)

IND_RLEVt-1 0.9140*** −0.0747**

(0.0378) (0.0360)
IND_RLEVt-1 * OVER_INVt-1 −0.2129*** 0.7792***
(0.0418) (0.0397)
TAXRATE_WAGEt-1 −0.0144 −0.0187**
(0.0096) (0.0091)
TAXRATE_WAGEt-1 * OVER_INVt-1 0.0026 0.0082
(0.0074) (0.0070)
Other controls Yes Yes
Firm FE Yes Yes
Year FE Yes Yes
Underidentification test (Sanderson–Windmeijer χ2-stat) 4673.37*** 2961.78***
Weak instruments test (Sanderson–Windmeijer F-stat) 1313.72*** 832.58***
R2 (%) 76.56% 76.66%
N 9,542 9,542

Panel B: Second-stage regressions

Variable NET_PAYOUTt TOT_PAYOUTt
(1) (2)

CEO_RLEVt-1 (PREDICTED) −0.0146*** −0.0073**

(0.0044) (0.0037)
CEO_RLEVt-1 * OVER_INVt-1 (PREDICTED) 0.0206*** 0.0130***
(0.0052) (0.0043)
Other controls Yes Yes
Firm FE Yes Yes
Year FE Yes Yes
Overidenfication test (Sargan χ2-stat) 0.96 0.10
Endogeneity test (Sargan–Hansen χ2-stat) 5.11* 0.10
Net effect of inside debt at firms with overinvestment (F-stat) 3.18* 4.24**
R2 (%) 6.98% 6.22%
N 9,542 9,542

the main independent variable. To check the sensitivity of our results to clustering effects, we also rerun the regression analysis with
clustered standard errors (Petersen, 2009). Nonetheless, we continue to find that higher inside debt CEOs at overinvested (under-
invested) firms are associated with higher (lower) payouts. In sum, the empirical evidence confirms the effectiveness of inside debt
compensation.

4. Conclusion

According to the agency theory, inside debt compensation can help mitigate agency problems in corporations by altering risk-
taking behavior. However, there are theoretical and empirical inconsistencies regarding the effect of inside debt on firm payout. Our
paper is motivated to address this issue. Drawing from investment efficiency literature, we argue and empirically show that high
inside debt CEOs increase (decrease) payouts in the subsequent period if their firms have overinvestment (underinvestment) pro-
blems in the prior period. Our study contributes to the literature by shedding light on the aforementioned contradiction, and showing
that inside debt is an effective mechanism that helps alleviate agency problems. That is, inside debt CEOs adjust payouts to mitigate
investment inefficiency at their firms.

6
A. Erkan and T. Nguyen Finance Research Letters xxx (xxxx) xxxx

Table. 5
Regression results with industry-wide measures of investment inefficiency. This table reports the regression results of firm payout policies on CEO
relative leverage, industry-wide overinvestment/underinvestment dummies, and their interactions controlling for other CEO and firm character-
istics. The IND_OVER_INV (IND_UNDER_INV) dummy takes a value of one if the residual estimated from the following industry-year regression
belongs to the top (bottom) quintile, and zero otherwise (Biddle et al., 2009; Lara et al., 2016). IND _INVs, t = 1 + 2 IND _GROWTHs, t 1 + s, t .In
which, IND_INV is the industry-year average of firm total investments that equal capital expenditures plus R&D expenses plus acquisitions minus net
sales of property, plant and equipment scaled by total assets at the beginning of the year. IND_GROWTH is the industry-year average of firm sales
growth. Detailed variable descriptions are provided in Table 1. Standard errors are reported in parentheses. ***, **, and * are used to denote
significance at 1%, 5%, and 10% levels, respectively.
Variable NET_PAYOUTt TOT_PAYOUTt
(1) (2)

CEO_RLEVt-1 0.0015 0.0014

(0.0014) (0.0012)
CEO_RLEVt-1 * IND_UNDER_INVt-1 −0.0090*** −0.0079***
(0.0029) (0.0024)
CEO_RLEVt-1 * IND_OVER_INVt-1 0.0076*** 0.0066***
(0.0017) (0.0014)
IND_UNDER_INVt-1 0.0057* 0.0043*
(0.0031) (0.0025)
IND_OVER_INVt-1 −0.0041** −0.0019
(0.0020) (0.0017)
CEO_AGEt-1 −0.0288** −0.0246***
(0.0113) (0.0093)
CEO_V2Dt-1 −0.0035 −0.0031*
(0.0022) (0.0018)
CEO_TENUREt-1 0.0021* 0.0017*
(0.0012) (0.0010)
SIZEt-1 0.0184*** 0.0044***
(0.0020) (0.0017)
AGEt-1 −0.0006 −0.0039
(0.0082) (0.0068)
LEVt-1 −0.0915*** −0.0851***
(0.0077) (0.0064)
MTBt-1 0.0012*** 0.0012***
(0.0003) (0.0002)
CASHt-1 0.0969*** 0.0779***
(0.0096) (0.0079)
ROAt-1 0.0867*** 0.0304***
(0.0077) (0.0064)
INTERCEPT −0.0072 0.1171***
(0.0508) (0.0421)
Net effect of inside debt for firms in industries prone to underinvestment (F-stat) 7.39*** 7.90***
Net effect of inside debt for firms in industries prone to overinvestment (F-stat) 36.25*** 41.00***
Firm FE Yes Yes
Year FE Yes Yes
R2 (%) 56.25% 58.13%
N 9,777 9,777

Acknowledgement statement

We would like to thank Samuel Vigne (the editor-in-chief), the handling editor, and two anonymous reviewers for their invaluable
comments and suggestions throughout the revision process.

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.frl.2020.101560.

References

Anantharaman, D., Fang, V., Gong, G., 2013. Inside debt and the design of corporate debt contracts. Manag. Sci. 60, 1260–1280.
Biddle, G.C., Hilary, G., Verdi, R.S., 2009. How does financial reporting quality relate to investment efficiency? J. Account. Econ. 48, 112–131.
Caliskan, D., Doukas, J., 2015. CEO risk preferences and dividend policy decisions. J. Corp. Financ. 35, 18–42.
Cassell, C., Huang, S., Manuel, S., Stuart, M., 2012. Seeking safety: the relation between CEO inside debt holdings and the riskiness of firm investment and financial
policies. J. Financ. Econ. 103, 588–610.
Dang, V.A., Phan, H.V., 2016. CEO inside debt and corporate debt maturity structure. J. Bank Financ. 70, 38–54.
DeAngelo, H., DeAngelo, L., Stulz, R.M., 2006. Dividend policy and the earned/contributed capital mix: a test of the life-cycle theory. J. Financ. Econ. 81, 227–254.
Easterbrook, F., 1984. Two agency-cost explanations of dividends. Am. Econ. Rev. 74, 650–659.
Edmans, A., Liu, Q., 2011. Inside Debt. Rev. Financ. 15, 75–102.

7
A. Erkan and T. Nguyen Finance Research Letters xxx (xxxx) xxxx

Eisdorfer, A., Giaccotto, C., White, R., 2013. Capital structure, executive compensation, and investment efficiency. J. Bank Financ. 37, 549–562.
Eisdorfer, A., Giaccotto, C., White, R., 2015. Do corporate managers skimp on shareholders' dividends to protect their own retirement funds? J. Corp. Financ. 30,
257–277.
Grullon, G., Michaely, R., Swaminathan, B., 2002. Are dividend changes a sign of firm maturity? J. Bus. 75, 387–424.
Jensen, M., Meckling, W., 1976. Theory of the firm: managerial behavior, agency cost, and ownership structure. J. Financ. Econ. 3, 305–360.
Jensen, M., 1986. Agency costs of free cash flow, corporate finance, and takeovers. Am. Econ. Rev. 76, 323–329.
Lambrecht, B.M., Myers, S.C., 2017. The dynamics of investment, payout and debt. Rev. Financ. Stud. 30, 3759–3800.
Lara, J.M.G., Osma, B.G., Penalva, F., 2016. Accounting conservatism and firm investment efficiency. J. Account. Econ. 61, 221–238.
Liu, Y., Mauer, D., Zhang, Y., 2014. Firm cash holdings and CEO inside debt. J. Bank. Financ. 42, 83–100.
Lu-Andrews, R., Yu-Thompson, Y., 2015. CEO inside debt, asset tangibility, and investment. Int. J. Manag. Financ. 11, 451–479.
Myers, S.C., 1984. The capital structure puzzle. J. Financ. 39, 574–592.
Myers, S.C., Majluf, N.S., 1984. Corporate financing and investment decisions when firms have information that investors do not have. J. Financ. Econ. 13, 187–221.
Nikolic, B., Yan, X.S., 2014. Investor overconfidence, misvaluation, and corporate decisions. J. Financ. Econ. 2, 52–104.
Petersen, M.A., 2009. Estimating standard errors in finance panel data sets: comparing approaches. Rev. Financ. Stud. 22, 435–480.
Srivastav, A., Armitage, S., Hagendorff, J., 2014. CEO inside debt holdings and risk shifting: evidence from bank payout policies. J. Bank. Financ. 47, 41–53.
Sundaram, R., Yermack, D., 2007. Pay me later: inside debt and its role in managerial compensation. J. Financ. 62, 1551–1588.
Wei, C., Yermack, D., 2011. Investor reactions to CEOs inside debt incentives. Rev. Financ. Stud. 24, 3813–3840.
Wooldridge, J.M., 2010. Econometric Analysis of Cross Section and Panel Data. MIT press.
Wu, T.H., Lin, M.C., 2019. Relationship of CEO inside debt and corporate social performance: a data envelopment analysis approach. Financ. Res. Lett. 29, 308–314.
Yu-Thompson, Y., Zhao, S., 2017. CEO inside debt and overinvestment. J. Account. Financ. 17, 83–99.