1. bookAHEAD OF PRINT
Informacje o czasopiśmie
License
Format
Czasopismo
eISSN
2444-8656
Pierwsze wydanie
01 Jan 2016
Częstotliwość wydawania
2 razy w roku
Języki
Angielski
access type Otwarty dostęp

Corporate social responsibility fulfilment, product-market competition and debt risk: Evidence from China

Data publikacji: 31 Mar 2022
Tom & Zeszyt: AHEAD OF PRINT
Zakres stron: -
Otrzymano: 13 Jul 2021
Przyjęty: 06 Dec 2021
Informacje o czasopiśmie
License
Format
Czasopismo
eISSN
2444-8656
Pierwsze wydanie
01 Jan 2016
Częstotliwość wydawania
2 razy w roku
Języki
Angielski
Abstract

Based on the financial data for Chinese A-share companies listed in the Shanghai and Shenzhen stock exchanges and the corporate social responsibility (CSR) score data published in the Hexun websitefrom 2010 to 2019, we carried out an empirical study and found that CSR fulfilment helps to restrain debt risk faced by firms, and that product-market competition can promote the inhibitory effect of CSR fulfilment on debt risk. The further study found that CSR fulfilment of monetary capital layer and human capital layer on debt risk will be significantly stronger than that of social capital layer.

Keywords

Introduction

China’s rapid economic growth over the past 40 years has brought about remarkable results, not least in making it one of the fastest growing economies among the countries of the world. In the past, China’s economic growth was achieved by relying on the high household savings rate and the high investment growth rate, and all this was due to the high debt operation. Also, in recent years, China’s debt risk has been exposed due to a series of changes in the macroeconomic environment, for example, the disappearance of demographic bonus, the slowdown of economic growth rate, overcapacity, financial crises and so on; further, default events such as city investment debt and China Aviation Oil have also emerged in quick succession, resulting in an adverse impact on the debt risk. Unless stringent precautionary measures are implemented, the debt risk will be released in a short time, which is likely to evolve into a systemic financial risk and harm the real economy of China. Therefore, scientific management and control of corporate debt risk is not only conducive to the long-term development of firms but also helps to build a barrier to prevent and resolve major financial systemic risks at the micro level; further, it has become a matter of consensus among all sectors of society that this scientific approach is imperatively necessary given the incumbent position of debt risk in China. Most of the existing studies in the literature investigate the causes of corporate debt risk based on three factors: corporate characteristics, internal corporate governance and external macro environment. The fulfilment of corporate social responsibility (CSR) helps to enhance corporate reputation, obtain the support of stakeholders and gain competitive advantage in the fierce product-market competition. It has a positive impact on corporate profitability and cash flow, and ultimately affects the debt risk faced by firms. However, few scholars have conducted in-depth research on the relationship between CSR and debt risk based on the Chinese background. With the advent of the ‘new normal’, the degree of industry competition and the firm behaviour in China’s product market show new characteristics. The fierce market competition has greatly affected companies’ internal financing ability and profit space. In order to obtain external funds at lower cost, managers must release more signals to the market to alleviate the information asymmetry with stakeholders, and establish a good social image and market image for firms through the fulfilment of CSR [1]. The impact of degree of competition in the product market on the relationship between CSR fulfilment and debt risk is worthy of further exploration.

The possible marginal contributions of this paper are as follows. First, it helps managers to further identify the influencing factors of debt default risk, and enriches the literature research in the field of CSR and debt risk. Second, for the first time, this paper brings CSR fulfilment, product-market competition and debt risk into the same research framework, and this arrangement provides a feasible way to deeply analyse the internal relationship prevailing among them. Third, this paper further explores the effect of the stratified influence mechanism of CSR fulfilment on corporate debt risk, and also, an understanding of this causative phenomenon is helpful for policy makers to reasonably guide CSR fulfilment and helps to prevent and resolve systemic financial risks, which has certain practical and theoretical significance.

The rest of this paper is organised as follows. The second part is the literature review and research hypothesis. The third part is the research design. The fourth part is the empirical analysis. The fifth part is a furtherance of the analysis. The last part contains the conclusion and policy suggestions.

Literature review and research hypothesis
The impact of CSR fulfilment on debt risk

This paper reviews the research achievements of domestic and foreign scholars in recent years. It is found that there are four main reasons that can explain why the fulfilment of social responsibility has a negative impact on debt risk: reducing the cost of debt financing, extending the debt maturity and reducing the strictness of debt contract, and improving the credit rating so as to have a restraining effect on debt risk. First, in terms of reducing the cost of debt financing: Goss and Roberts [2] conducted a sample survey of 3996 corporate loans in the United States, and found that the loan interest rate of firms with better CSR is 7–18 basis points lower than that of firms only focusing on CSR. Second, in terms of extending debt maturity and reducing the strictness of debt contracts, Shi and Sun [3] found that the more firms invest in CSR, the less restrictive the clauses in their debt contracts. Third, in recent years, studies have focused more on the relationship between CSR and credit rating. Attig et al. [4] showed that credit rating agencies tend to give higher credit ratings to firms that perform well in social responsibility. The studies mentioned above all lend credence to the view that firms playing an active part in CSR can obtain the benefit of reduced debt risk owing to their good corporate image, lower debt financing cost, more relaxed credit conditions and higher credit rating. According to this inference, we deduce the first research hypothesis:

H1: The fulfilment of CSR can inhibit the debt risk faced by firms.

The moderating role of the degree of competition in the product market in the relationship between CSR fulfilment and debt risk

With the advent of the ‘new normal’, the degree of industry competition and the firm behaviour in China’s product market show new characteristics. On the one hand, product-market competition urges firms to seek better internal governance; a highly competitive market will force firms to reduce product costs, seek lower financing costs, improve internal governance [5] and reduce the degree of information asymmetry [6]. On the other hand, fierce competition in the product market will aggravate firm risk. Wu Haomin et al. [7] show that the significant increase of idiosynthetic risk of listed companies in China is attributable to the intensification of product-market competition. Han Zhongxue and Zhou Tingting [8] found that for all manufacturing listed companies in China, product-market competition has a significant predatory risk effect on cash holdings. Competition is one of the most urgent problems for firms to face when they are established in the market. To some extent, firms’ fulfilment of CSR is affected by the competitors’ CSR input level. When stakeholders evaluate the fulfilment of CSR, they will also refer to the competitors’ degree of CSR fulfilment in the same industry. Thus, we propose the second hypothesis of this paper:

H2: The degree of product-market competition regulates the relationship between CSR fulfilment and debt risk.

Research design
Sample selection and data sources

This paper extracts data for the Chinese A-share listed companies in the Shanghai and Shenzhen stock exchanges from 2010 to 2019 as research samples, and carries out the following processing steps:

Remove the samples of listed companies with abnormal trading status such as ST, *ST, Pt, etc.

Exclude samples with missing main variables.

Exclude listed companies belonging to the financial and insurance sectors.

To avoid the influence of extreme outliers, conduct winsorization to achieve tail reduction on all continuous variables at the level of 1%.

Through the steps outlined above, annual observations of 23,660 companies were finally obtained. The CSR data used in this paper were obtained from the CSR rating database Hexun, and the financial data from the CSMAR database.

Model design and variable definition

In order to test hypothesis 1, we design model 1, which is represented as follows: FDRi,t=α0+α1CSRi,t+α2Controlsi,t+Industry+Year+εi,t {\rm{FDR}}_{i,t} = {\alpha _0} + {\alpha _1}{\rm{CSR}}_{i,t} + {\alpha _2}{\rm{Controls}}_{i,t} + \sum {\rm{Industry}} + \sum {\rm{Year}} + {\varepsilon _{i,t}}

In the above model, the main explanatory variable is CSR, which is the comprehensive score of CSR published by Hexun from 2010 to 2019. The rating system integrates the CSR fulfilment in five aspects: shareholder, employee, rights and interests of suppliers, customers and consumers, environment and social of listed Companies. The full score is 100; the higher the score, the better is the fulfilment of CSR.

FDR is used as an explained variable to measure the debt risk of firms. According to previous studies, there are mainly two ways to estimate the debt risk of firms. The first approach is to use accounting data to measure the debt risk, and the second way is to measure the debt risk by combining market data and accounting data. Z_score [9] and O_score [11] are among the commonly used accounting data estimation methods. Estimation methods based on the combination of market data and accounting data include the default option model mentioned by Black and Scholes [12], the KMV model used by Merton [13] to measure default distance and a simple risk model mentioned in the study of Shumway [14]. In the past, a large number of studies in the literature used the combination of market data and accounting data to predict the debt risk of firms [15, 16]. However, when Agarwal and Taffler [17] conducted cross-border analysis, they found that the Z_score model was superior to risk-based and market-based prediction models in the accuracy of debt risk prediction. In a recent study, Altman et al. [18] confirmed the effectiveness of Z_score as a predictor of debt risk. Based on the above inferences drawn from the literature, this paper uses a forecasting model based on accounting data to estimate the debt risks faced by firms.

In model 1, FDR is the explained variable to measure the debt risk of firms. We mainly use two accounting measurement methods to predict the debt risk of firms, namely Z_score [19] as calculated by model 2 and O_score [11] as calculated by model 3. Z_score=1.2×(Operatingfunds/Totalassets)+1.4×(Retainedearnings/Totalassets)+3.3×(Earningsbeforeinterestandtax/Totalassets)+0.6×(Valueofequity/Valueofliability)+1×(Salesrevenue/Totalassets) \matrix{{{\rm{Z}}\_{\rm{score}}} \hfill & {= 1.2 \times ({\rm{Operating}}\,{\rm{funds}}/{\rm{Total}}\,{\rm{assets}}) + 1.4 \times ({\rm{Retained}}\,{\rm{earnings}}/{\rm{Total}}\,{\rm{assets}})} \hfill \cr {} \hfill & {+ 3.3 \times ({\rm{Earnings}}\,{\rm{before}}\,{\rm{interest}}{\kern 1pt} {\rm{and}}\,{\rm{tax}}/{\rm{Total}}\,{\rm{assets}}) + 0.6 \times ({\rm{Value}}\,{\rm{of}}\,{\rm{equity}}/{\rm{Value}}\,{\rm{of}}\,{\rm{liability}})} \hfill \cr {} \hfill & {+ 1 \times ({\rm{Sales}}\,{\rm{revenue}}/{\rm{Total}}\,{\rm{assets}})} \hfill \cr}

The higher the calculated value of Z_score, the better the financial status and the lower the debt risk of the firm. For the sake of explanation, Z_score has been negative in the follow-up empirical test. O_scorei,t=1.320.407×logTAi,t+0.63×(TLi.t/TAi,t)1.43×(WCi.t/TAi,t)+0.076×(CLi.t/CAi,t)1.72×(OEMEGi,t)2.37×(NLi,t/TAi,t)1.83×(OIBDi.t/TAi,t)+0.285×(EBITTWOi,t)0.521×(NIi,tNIi,t1)/(|NIi,t||NIi,t1|) \matrix{{{\rm{O}}\_{\rm{score}}_{i,t}} \hfill & {=- 1.32 - 0.407 \times {\log}{{\rm{TA}}_{i,t}} + 0.63 \times ({\rm{TL}}_{i.t}/{\rm{TA}}_{i,t}) - 1.43 \times ({\rm{WC}}_{i.t}/{\rm{TA}}_{i,t})} \hfill\cr{} \hfill & {+ 0.076 \times ({\rm{CL}}_{i.t}/{\rm{CA}}_{i,t}) - 1.72 \times ({\rm{OEMEG}}_{i,t}) - 2.37 \times ({\rm{NL}}_{i,t}/{\rm{TA}}_{i,t}) - 1.83 \times ({\rm{OIBD}}_{i.t}/{\rm{TA}}_{i,t})} \hfill\cr{} \hfill & {+ 0.285 \times ({\rm{EBITTWO}}_{i,t}) - 0.521 \times ({\rm{NI}}_{i,t} - {\rm{NI}}_{i,t - 1})/(|{\rm{NI}}_{i,t}| - |{\rm{NI}}_{i,t - 1}|)} \hfill\cr}

In model 3, logTAi,t is natural logarithm of total assets; TLi.t/TAi,t is total liabilities divided by total assets; WCi.t/TAi,t is working capital divided by total assets; CLi.t/CAi,t is current liabilities divided by current assets; OEMEGi,t is a dummy variable, and if the total liabilities of the firm are greater than the total assets, this takes the value 1, and otherwise 0; NLi,t/TAi,t is net profits divided by total assets; OIBDi.t/TAi,t is operating income before depreciation divided by total assets; EBITTWOi,t is a dummy variable, and if firm net profit has been negative for 2 consecutive years, the index takes the value 1, and otherwise 0; NIi,t and NIi,t−1 represent the firm’s net profit for the current year and the previous year, respectively. The higher the value of O_score, the greater the risk of bankruptcy faced by the firm.

Based on model 1, model 4 is designed to test hypothesis 2, as under: FDRi,t=β0+β1CSRi,t+β2HHIi,t+β3CSRi,t×HHIi,t+β4Controlsi,t+Industry+Year+εi,t {\rm{FDR}}_{i,t} = {\beta _0} + {\beta _1}{\rm{CSR}}_{i,t} + {\beta _2}{\rm{HHI}}_{i,t} + {\beta _3}{\rm{CSR}}_{i,t} \times {\rm{HHI}}_{i,t} + {\beta _4}{\rm{Controls}}_{i,t} + \sum {\rm{Industry}} + \sum {\rm{Year}} + {\varepsilon _{i,t}}

In model 4, the explanatory variable is the cross product of CSR and HHI, where HHI is the degree of market competition. Based on the method suggested by Peress [20], the present study uses the Herfindahl–Hirschman index (HHI) to portray the degree of industry competition. Among others, one calculation formula used for HHI is the following: HHI=i=1N(XiX)2 {\rm{HHI}} = \sum\limits_{i = 1}^N {\left({{{{X_i}} \over X}} \right)^2} where N is the total number of firms in the industry, Xi is the annual sales revenue of firm i and X is the total annual sales revenue of firms in the market. The lower the HHI, the stronger the competitiveness is in the industry, and vice versa. For the sake of explanation, the HII has been allowed to assume a negative value in the empirical tests that follow in this paper.

In models 1 and 4, we also include the following sets of control variables, which probably affect FDR. Firm’s characteristics variables: firm’s scale (Size), firm’s age (Age). Firm’s financial characteristics variables: profitability (ROA), cash flow volatility (VCF), management expense ratio (Overhead), non-debt tax shield (NDT), market-to-book ratio of equity (MTB), whether there is a dividend payout during the current year (Div), noncurrent assets to total assets ratio (PPE), shareholding ratio of the top five shareholders (Lncon), growth rate of operating income (Growth) and stock returns (RET). Board characteristic variables: Ratio of independent directors (Indratio), CEO duality (Duality). We also control for ∑Industry (∑Year), which is a set of industry (year) dummy variables. Definitions of other relevant main variables are shown in Table 1.

Variable definitions

Variables classification Variable names Variable definitions
Explained variables FDR Z_score as calculated in model 2.
O_score as calculated in model 3.
Explanatory variables CSR CSR score is obtained from the CSR reports of listed companies published in Hexun.
Control variables Size Natural logarithm of total assets at the end of the year.
ROA ROA is equal to pretax operating income divided by total assets.
Age Age is equal to inspection year minus listing year.
VCF Three year volatility of the ratio of cash flow to total assets at the end of the year.
Overhead Overhead is the management expense that happened during the current year divided by total assets.
NDTS NDTS is the ratio of current year depreciation and amortisation to total assets.
MTB MTB is the growth rate of firms, and it is the ratio of the market value of equity at the end of the year to the book value of equity at the end of the year.
Div Div represents weather the enterprise has had dividend payouts in the current year; if yes it takes the value of 1 and 0 otherwise.
PPE The ratio of total noncurrent assets to total assets.
Lncon Lncon is the combined shareholding of the top five shareholders.
Growth Growth is the growth rate of operating income that equals change in sales revenue of the current year compared to the previous year, and then divided by the previous year’s operating income.
RET RET is the firm’s average monthly stock return during the year.
Indratio The proportion of independent directors to the total number of directors.
Duality Duality is a dummy variable that equals 1 if chairman and the general manager are in one position and 0 otherwise.

CSR, corporate social responsibility; MTB, market-to-book ratio of equity; NDTS, non-debt tax shield.

Analysis of empirical results
Variable descriptive statistics

Table 2 presents descriptive statistics of the main variables. Altman [19] pointed out that when the Z_score is >2.99, it means that the financial status of the firm is in good health and the probability of debt risk is low. When Z_score is <1.81, it indicates that the firm has latent debt risk. When Z_score is between 1.81 and 2.99, it is called the ‘grey area’, which indicates that the financial situation of firm is in an unstable state. The main variable that is explained in this paper, Z_score, has a mean value of 6.7777, the maximum value is 55.5422 and the minimum value is 0.1785; further, the 25% percentile is 2.0313, the median value is 3.7837 and the standard deviation is 8.8820. O_score’s mean value is −11.5968, the maximum value is −5.0486 and the minimum value is −27.3073; further, the 25% percentile is −13.3983, the median value is −10.7945 and the standard deviation is 3.9466. This shows that the debt risk faced by different firms is quite different, and that at least 25% of firms are faced with debt risk. The full score of CSR fulfilment published in the Hexun database is 100. According to these data, the average and median value of CSR are 24.8365 and 22.1300, respectively, with the maximum value of 74.5200 and the minimum value of −3.1500, indicating that the degree of CSR fulfilment in China needs to be further improved. With regard to firm characteristics, the mean value of Size is 22.1602, the mean value of ROA is 0.0394, the mean value of VCF is 0.0453, the mean value of Growth is 0.1967 and the mean value of PPE is 0.2186, which indicates that the firms in this sample have a large proportion of fixed assets, low asset profitability and weak solvency, and face a certain level of default risk.

Descriptive statistics

Variables N Minimum Mean p25 Median p75 Std. Dev Maximum

Z_score 23660 0.1785 6.7777 2.0313 3.7837 7.4065 8.8820 55.5422
O_score 23660 −27.3073 −11.5968 −13.3983 −10.7945 −8.8855 3.9466 −5.0486
CSR 23660 −3.1500 24.8365 16.5800 22.1300 27.8000 16.0305 74.5200
Size 23660 19.7510 22.1602 21.2268 21.9870 22.8968 1.2861 25.8845
ROA 23660 −0.2157 0.0394 0.0147 0.0368 0.0657 0.0557 0.1962
Age 23660 0.0000 10.0511 4.0000 9.0000 16.0000 7.0084 25.0000
VCF 23660 0.0022 0.0453 0.0189 0.0339 0.0586 0.0392 0.2164
Overhead 23660 0.0041 0.0447 0.0248 0.0394 0.0581 0.0282 0.1544
NDTS 23660 0.0007 0.0238 0.0119 0.0209 0.0325 0.0157 0.0763
MTB 23660 0.5996 3.5910 1.7365 2.6862 4.3403 3.0321 19.1986
Div 23660 0.0000 0.7555 1.0000 1.0000 1.0000 0.4298 1.0000
PPE 23660 0.0022 0.2186 0.0889 0.1845 0.3128 0.1656 0.7144
Lncon 23660 0.2021 0.5356 0.4230 0.5381 0.6489 0.1523 0.8804
Growth 23660 −0.5395 0.1967 −0.0090 0.1159 0.2795 0.4604 3.1335
RET 23660 −0.0686 0.0102 −0.0188 0.0052 0.0319 0.0421 0.1592
Indratio 23660 0.3333 0.3748 0.3333 0.3333 0.4286 0.0536 0.5714
Duality 23660 0.0000 0.2575 0.0000 0.0000 1.0000 0.4373 1.0000

CSR, corporate social responsibility; MTB, market-to-book ratio of equity; NDTS, non-debt tax shield.

Analysis of regression results
The impact of CSR on corporate debt risk

In columns 1 and 4 of Table 3, we only control the industry and annual fixed effect for regression. In columns 2 and 5, we not only control the industry and annual fixed effect but also add control variables for multiple regressions. The results show that the regression coefficients of CSR and corporate debt risk Z_score and O_score are significantly negative at the 1% level; that is to say, the higher the level of CSR fulfilment, the lower the debt risk faced by firms, which supports hypothesis 1. In addition, the regression results of the control variables show that Size, Age, PPE and Growth are positively correlated with debt risk. However, ROA, Div, Lncon, RET and Indratio are negatively correlated with debt risk.

CSR fulfilment and debt risk, the moderating effect of product-market competition (Z_score and HHI are negative)

Variables Z_score O_score

(1) (2) (3) (4) (5) (6)

CSR −0.0169*** (−4.66) −0.0099*** (−2.73) −0.0164*** (−3.99) −0.0398*** (−25.38) −0.0073*** (−4.61) −0.0103*** (−5.70)
HHI 14.6110*** (10.46) 4.2472*** (6.93)
CSR*HHI −0.0911*** (−3.02) −0.0435*** (−3.28)
Size 2.0913*** (41.76) 2.0756*** (41.54) 0.8739*** (39.86) 0.8697*** (39.68)
ROA −36.6600*** (−35.63) −36.1459*** (−35.19) −26.9191*** (−59.76) −26.7790*** (−59.45)
Age 0.0488*** (5.77) 0.0522*** (6.19) 0.0215*** (5.82) 0.0225*** (6.08)
VCF 10.7848*** (8.71) 10.7521*** (8.71) −0.5821 (−1.07) −0.5788 (−1.07)
Overhead 4.3665** (2.15) 5.8658*** (2.89) −16.1690*** (−18.21) −15.7934*** (−17.77)
NDTS 15.1886*** (3.21) 14.8717*** (3.15) −20.6733*** (−9.99) −20.8409*** (−10.08)
MTB −0.6836*** (−33.75) −0.6920*** (−34.21) 0.2599*** (29.32) 0.2581*** (29.10)
Div −0.2356* (−1.81) −0.2202* (−1.70) −0.3078*** (−5.41) −0.3027*** (−5.33)
PPE 4.1707*** (8.84) 4.1867*** (8.90) 3.7310*** (18.06) 3.7347*** (18.10)
Lncon −4.8438*** (−14.28) −4.6771*** (−13.81) −2.7487*** (−18.51) −2.7058*** (−18.22)
Growth 1.3257*** (12.81) 1.3227*** (12.82) 0.2641*** (5.83) 0.2626*** (5.80)
RET −12.8841*** (−7.89) −11.9855*** (−7.35) −1.9071*** (−2.67) −1.6842** (−2.35)
Indratio −4.2738*** (−4.95) −4.3505*** (−5.05) −0.7872** (−2.08) −0.8201** (−2.17)
Duality −0.5299*** (−4.84) −0.5266*** (−4.82) 0.0088 (0.18) 0.0095 (0.20)
Constant −7.2702*** (−14.07) −46.4121*** (−39.28) −44.2865*** (−37.10) −10.0410*** (−44.92) −28.1203*** (−54.36) −27.5263*** (−52.57)
Year/Industry Yes Yes Yes Yes Yes Yes
N 23660 23660 23659 23660 23660 23659
Adj-R2 0.0977 0.3779 0.3813 0.1443 0.3960 0.3972
F 86.4271 327.5768 318.0266 133.9788 353.5853 339.8737

Note: T values are in parentheses;

represent significance at the levels of 1%, 5% and 10%, respectively. CSR, corporate social responsibility; HHI, Herfindahl-Hirschman index; MTB, market-to-book ratio of equity; NDTS, non-debt tax shield.

The moderating role of the degree of product-market competition in the relationship between CSR fulfilment and debt risk

The regression results of columns 3 and 6 (Table 3) show that the regression coefficients of debt risk Z_score and O_score and CSR * HHI are significant at the 1% level, which indicates that the degree of product-market competition has a significant moderating effect on the relationship between CSR fulfilment and debt risk, i.e. the more fierce the product-market competition, the more the inhibitory effect of CSR fulfilment on debt risk can be promoted, and hypothesis 2 is verified.

Robustness checks
Replace key variables

CSR and debt risk Z_score and O_score are the main variables in this paper, and these may be affected by the inherent differences of the industry. m_CSR, m_Z_score and m_O_score are the industry-adjusted CSR, Z_score and O_score, measured as CSR, Z_score and O_score less the mean values of CSR, Z_score and O_score for all firms in the same industry, respectively.

The results are shown in Table 4. The m_Z_score and m_O_score are significantly negatively correlated with the m_CSR, which is consistent with the regression results in Table 3 concerning the main test analysis, indicating that the above conclusions are robust.

Replacement of main variables (Z_score is negative)

Variables Adjusted by industry annual mean value

m_Z_score m_O_score

(1) (2)

m_CSR −0.0126*** (−3.49) −0.0090*** (−5.68)
Controls Yes Yes
Constant −40.0503*** (−33.60) −17.4561*** (−33.27)
Year/Industry Yes Yes
N 23660 23660
Adj-R2 0.3019 0.3023

Note: The control variables in Table 4 are the same as those in Table 3, and to avoid repetition, the results of control variables are omitted.

T values are in parentheses;

represent significance at the levels of 1%, 5% and 10%, respectively. CSR, corporate social responsibility.

Dynamic test

In conformity with the suggestion of Chen Caiyun [10], we test the robustness of our results by using change in CSR and change in debt risk in the current year and the next year; whole control variables are changes in the current year, and the results are shown in columns 1 and 3 (Table 5). The CSR fulfilment change in the current period and the next period show a negative relationship with the Z_score and O_score changes in the next period at the 1% level, as can be seen in columns 2 and 4 (Table 5). However, the CSR fulfilment change in the next period does not seem to affect the Z_score and O_score changes in the current period.

Dynamic test (Z_score is negative)

Variables ΔZ_scoret+1 ΔZ_scoret ΔO_scoret+1 ΔO_scoret

(1) (2) (3) (4)

ΔCSRt+1 −0.0131*** (−4.40) 0.0017 (0.59) −0.0208*** (−18.17) −0.0007 (−0.73)
ΔCSRt −0.0163*** (−5.27) −0.0061*** (−5.13)
ΔSize 0.9445*** (5.55) 3.1999*** (18.70) −0.0751 (−1.15) 2.0405*** (33.96)
ΔROA 3.1905*** (3.69) −16.2485*** (−19.27) 5.6848*** (17.16) −15.4860*** (−52.30)
ΔAge 0.3425 (0.43) −1.3515* (−1.69) −0.0981 (−0.32) −0.1314 (−0.47)
ΔVCF 0.6081 (0.45) −2.8218** (−2.09) 0.2477 (0.48) −0.9724** (−2.05)
Δoverhead 12.4276*** (4.08) −7.0720** (−2.31) 4.4970*** (3.85) −10.6482*** (−9.90)
ΔNDTS −10.4221 (−1.44) −13.0303* (−1.79) 3.2840 (1.18) −26.0014*** (−10.15)
ΔMTB 0.0874*** (4.09) −0.4907*** (−22.85) −0.0156* (−1.90) 0.1604*** (21.28)
ΔDiv −0.0005 (−0.01) −0.2662*** (−2.72) 0.1689*** (4.49) −0.3473*** (−10.10)
ΔPPE −0.4916 (−0.75) 5.2796*** (7.98) 0.0463 (0.18) 3.4835*** (14.99)
ΔLncon −3.5143*** (−4.52) −6.5829*** (−8.42) −0.8265*** (−2.77) −3.6601*** (−13.33)
ΔGrowth −0.0156 (−0.23) 0.4864*** (7.23) −0.0286 (−1.12) −0.3487*** (−14.77)
ΔRET −1.1422 (−1.17) −18.2114*** (−18.52) −1.1966*** (−3.19) −2.1178*** (−6.13)
ΔIndratio −0.6330 (−0.57) 0.8220 (0.74) −0.2544 (−0.60) 0.2334 (0.60)
ΔDuality 0.1502 (1.07) −0.1674 (−1.18) −0.0247 (−0.46) −0.0028 (−0.06)
Constant −0.3142 (−0.36) 1.4759* (1.68) 0.2355 (0.71) 0.1100 (0.36)
Year/industry Yes Yes Yes Yes
N 17164 17164 17164 17164
r2 0.0872 0.2325 0.0528 0.2586
Adj-R2 0.0850 0.2306 0.0504 0.2568

Note: T values are in parentheses;

represent significance at the levels of 1%, 5% and 10%, respectively.

CSR, corporate social responsibility.

This shows that changes in fulfilment of CSR in the current period and the next period can affect the change in debt risk in the next period, but the change in fulfilment of CSR in the next period cannot affect the change in debt risk in the current period, which eliminates the endogenous problem caused by the reverse causal relationship between the fulfilment of CSR and debt risk. The above research conclusion is still robust.

The regression results of independent variable lag 1 year

The interaction between the dependent variable and the independent variable is also a key factor leading to endogeneity. Since the lag term of the independent variable has occurred, it is ‘predetermined’ (– i.e. from the perspective of the current period, its value has been fixed). The current independent variable is related to it, but the current dependent variable has nothing to do with it. Therefore, the endogenous problem caused by the interaction between dependent variables and independent variables can be properly dealt with by the 1-year lag treatment of the independent variables. Therefore, this paper will replace the original independent variables in model 1 with the 1-year lag treatment of the independent variables, and then carry out the robustness test. The results are shown in columns 1 and 2 (Table 6), where a CSR lag of 1 year shows a negative relationship with the debt risk Z_score and O_score at the 1% level, indicating that after considering the endogenous problems that may be caused by the interaction of the dependent variables and the independent variables, the results are still consistent with the results of the previous main test analysis, which is robust.

Regression results of independent variable lag for one period (Z_score is negative)

Variables Z_score O_score

(1) (2)

CSRt−1 −0.0174*** (−4.82) −0.0095*** (−5.67)
Sizet−1 2.1373*** (39.83) 0.7652*** (30.88)
ROAt−1 −28.8522*** (−25.90) −19.5183*** (−37.95)
Aget−1 0.0127 (1.39) 0.0117*** (2.76)
VCFt−1 8.3963*** (6.47) −0.2604 (−0.43)
overheadt−1 0.6466 (0.31) −15.3937*** (−15.87)
NDTSt−1 12.3036** (2.43) −17.8428*** (−7.63)
MTBt−1 −0.4280*** (−20.11) 0.2122*** (21.59)
Divt−1 0.0453 (0.33) −0.1449** (−2.27)
PPEt−1 3.1000*** (6.15) 3.0327*** (13.04)
Lncont−1 −4.6787*** (−13.07) −2.7233*** (−16.48)
Growtht−1 0.9543*** (8.88) 0.2054*** (4.14)
RETt−1 −6.0977*** (−3.59) −3.2185*** (−4.10)
Indratiot−1 −4.6850*** (−5.11) −0.4761 (−1.12)
Dualityt−1 −0.4769*** (−4.09) 0.0440 (0.82)
Constant −46.6619*** (−37.39) −25.7190*** (−44.64)
Year/Industry Yes Yes
N 20363 20363
r2 0.3128 0.3038
Adj-R2 0.3114 0.3023

Note: T values are in parentheses;

represent significance at the levels of 1%, 5% and 10%, respectively.

CSR, corporate social responsibility; MTB, market-to-book ratio of equity; NDTS, non-debt tax shield.

Propensity score matching (PSM)

Considering that the firm’s CSR fulfilment may be inherent to its environment and its own characteristics, the characteristic variables of enterprises with good CSR fulfilment and those with poor CSR fulfilment are significantly different. If these characteristic variables also affect debt risk, then the relationship between firm CSR fulfilment and debt risk observed in the above probabilistic analysis model based on the full sample may affect interference factors, and so the PSM method is used to solve the endogenous problem. This method controls the significant differences in corporate characteristics between the firms with good and poor CSR fulfilment. In order to realise the PSM method, we use matching samples with similar firm characteristics to test the impact of CSR fulfilment on debt risk. This paper matches the firms with higher scores of CSR (higher than the industry median value) with those firms with lower scores (lower than the industry median value). We implement this procedure using the probit method (model 6), and control the following firm’s characteristic variables: Size, Age, ROA, VCF and SOE (Nature of ownership is a dummy variable; if a firm’s ultimate controller is the government, it take the value of 1, and 0 otherwise); at the same time, we also control industry and year fixed effects. The dummy variable of CSR fulfilment is the dependent variable in the probit regression; if the score of CSR fulfilment is higher than the industry median value, then it takes the value of 1, and otherwise the value of 0. We use the nearest neighbour method to match the firms with high scores of CSR (treatment firms) with the firms with low scores of CSR (control firms), and conduct the sampling without putting back. The PSM method produces a matching sample, including 8599 firm annual observations. DumCSR=γ0+γ1Sizei,t+γ3ROAi,t+γ4Agei,t+γ5VCFi,t+γ6SOEi,t+Industry+Year+εi,t {\rm{Dum}}_{{\rm{CSR}}} = {\gamma _0} + {\gamma _1}{\rm{Size}}_{i,t} + {\gamma _3}{\rm{ROA}}_{i,t} + {\gamma _4}{\rm{Age}}_{i,t} + {\gamma _5}{\rm{VCF}}_{i,t} + {\gamma _6}{\rm{SOE}}_{i,t} + \sum {\rm{Industry}} + \sum {\rm{Year}} + {\varepsilon _{i,t}}

Table 7 shows the test results of running the main regression with matched samples. In columns 1 and 2, CSR and debt risk Z_score and O_score are negatively correlated, and these are significant at the level of 1%. This shows that the firms that integrate CSR fulfilment into their business activities and business strategies will face lower debt risk, which is consistent with the results of the previous full sample model regression, thus proving that the conclusion of this study is robust.

PSM (Z_score is negative)

Variables Z_score O_score

(1) (2)

CSR −0.0166*** (−2.99) −0.0093*** (−3.79)
Controls Yes Yes
Constant −46.7106*** (−23.97) −28.4039*** (−33.11)
Year/Industry Yes Yes
N 8599 8599
r2 0.3815 0.3802
Adj-R2 0.3783 0.3771

Note: The control variables in Table 7 are the same as those in Table 3. To avoid repetition, the results of control variables are omitted. T values are in parentheses;

represent significance at the levels of 1%, 5% and 10%, respectively.

CSR, corporate social responsibility; PSM, propensity score matching.

To sum up, after considering the measurement error of variables, the endogenous problems caused by simultaneity and the interference factors of different CSR fulfilments with the same corporate characteristics, all kinds of robustness test results show that good CSR fulfilment can effectively reduce the debt risk faced by firms. The research conclusion of hypothesis 1 is thus robust.

Robustness test of moderate effect

In order to further test the moderating effect of product-market competition on the relationship between CSR fulfilment and debt risk, this paper divides the samples into a high product-market competition level group and a low product-market competition level group, and on this basis, the results of subsamples’ regression of model 1 are shown in Table 8. In the high product-market competition level group, columns 1 and 3 (Table 8) show that the estimated coefficients of debt risk Z_score and O_score are −0.0236 and −0.0157, respectively, and they are significant at the level of 1%. In the low product-market competition level group, columns 2 and 4 (Table 8), the estimated coefficients of debt risk Z_score and O_score are −0.0030 and −0.0001, respectively, but they are not statistically significant, indicating that the higher the degree of competition in the product market, the more obvious the inhibitory effect of CSR fulfilment on debt risk. The group test results support hypothesis H2.

Grouping by product-market competition intensity (Z_score and HHI are negative)

Variables Z_score O_score

High competition level Low competition level High competition level Low competition level

(1) (2) (3) (4)

CSR −0.0236*** (−4.45) −0.0030 (−0.60) −0.0157*** (−6.55) −0.0001 (−0.02)
Controls Yes Yes Yes Yes
Constant −51.6314*** (−31.24) −43.2057*** (−28.27) −34.0105*** (−45.51) −26.3189*** (−39.05)
Year/Industry Yes Yes Yes Yes
N 11830 11829 11830 11829
r2 0.3654 0.3780 0.3770 0.3403
Adj-R2 0.3641 0.3767 0.3757 0.3390

Note: The control variables in Table 8 are the same as those in Table 3. To avoid repetition, the results of control variables are omitted. T values are in parentheses; and

represent significance at the levels of 1%, 5% and 10%, respectively.

CSR, corporate social responsibility; HHI, Herfindahl-Hirschman index.

Further analysis
The different layer CSR fulfilment effect on debt risk

According to Table 3, although we arrive at the conclusion that the fulfilment of CSR will have an inhibitory effect on a firm’s debt risk, according to the CSR scoring system of Hexun, the fulfilment of CSR includes five dimensions: shareholder; employee; rights and interests of suppliers, customers and consumers; environmental; and CSR. In order to test the difference of the inhibitory effect of different dimensions of CSR on corporate debt risk, this paper, based on the inferences drawn by Junshan and Xudong [21], divides the above-mentioned five dimensions of CSR fulfilment into three layers: monetary capital layer (ShareholderR), human capital layer (EmployeeR) and social capital layer (ExternalR). Among them, the monetary capital level is expressed by the score of the shareholder responsibility dimension, the human capital level is expressed by the score of the employee responsibility dimension and the social capital level is expressed by the sum of the three dimensions of supplier, customer and consumer rights responsibility, and environmental responsibility and CSR. Tables 9 and 10 provide the regression results of different layers of CSR fulfilment and debt risk Z_score and O_score, respectively. In columns 1 and 2 (Tables 9 and 10), it can be seen that the fulfilment of CSR in monetary capital layer and human capital layer is negatively correlated with debt risk, and the estimated coefficients are significant at the level of 1%, indicating that the debt risk faced by firms with good CSR fulfilment in these two layers is significantly lower. Debt risk is largely caused by the external financing constraints of firms. From the perspective of monetary capital layer, the fulfilment of CSR at the layer of monetary capital helps firms to better obtain the support of shareholders, creditors and other capital providers, and obtain financing through better channels and more easier credit policies, without regard to the aspect of reducing the cost of equity financing [22] or that of reducing the cost of debt financing. In addition, it also plays a significant role in extending the debt maturity [2], thus alleviating the debt risk faced by firms. From the perspective of human capital layer, employees are one of the key stakeholders of the firm, and human capital is regarded as one of the most critical factors for a firm to succeed in an increasingly competitive environment; to a large extent, these results are in line with the conclusion of Bae et al. [23] that CSR fulfilment in terms of employee treatment fairness has a positive impact on corporate capital structure. From column 3 (Table 9), it can be seen that fulfilment of CSR in social capital layer (ExternalR) and Z_score are negatively correlated, and the estimated coefficient is −0.0067, which is not significant. In column 3, Table 10 displays that the regression results for the fulfilment of CSR in social capital layer (ExternalR) and O_score are also negatively correlated, with an estimated coefficient of −0.0063, although it is significant to the extent of 1%; but it is still remote compared to the coefficient of monetary capital layer CSR fulfilment in column 1 and the coefficient of human capital layer CSR fulfilment in column 2 (Table 10). It can be inferred that the inhibitory effect of CSR fulfilment of social capital on debt risk is significantly weaker than those of monetary capital and human capital. The regression results shown in column 4 (Table 9) also provide direct evidence in support of this conclusion. This shows that the CSR at the social capital layer has little inhibitory effect on the debt risk faced by firms. Column 4 (Tables 9 and 10) shows the regression results of all layers of CSR fulfilment on corporate debt risk, to assess the net impact of CSR fulfilment on corporate debt risk at each level. The regression results are similar to those in columns 1 and 2 (Tables 9 and 10). Similarly, the estimated coefficient of CSR fulfilment of social capital layer is not statistically significant, and this conclusion is consistent with the results of Godfrey et al. [24]. The reason is that the fulfilment of CSR to the main stakeholders (such as fund providers and employees) will create more favourable resource exchange among these groups, which can effectively alleviate the financing constraints of firms, enhance the value of firms and prevent negative economic consequences. It seems that the fulfilment of CSR to the secondary stakeholders cannot create more favourable financing conditions for firms, so as to reduce the debt risk faced by firms [25].

Dimensions of CSR and debt risk (Z_score is negative)

Variables Z_score

(1) (2) (3) (4)

ShareholderR −0.0775*** (−4.90) −0.0738*** (−4.58)
EmployeeR −0.0463*** (−2.79) −0.0643** (−2.46)
ExternalR −0.0067 (−1.38) 0.0121 (1.58)
Controls Yes Yes Yes Yes
Constant −46.1237*** (−39.79) −46.5026*** (−39.17) −46.0650*** (−39.11) −46.5855*** (−39.24)
Year/Industry Yes Yes Yes Yes
N 23660 23660 23660 23660
Adj-R2 0.3783 0.3779 0.3777 0.3784
F 328.1829 327.5895 327.3737 314.1067

Note: The control variables in Table 9 are the same as those in Table 3. To avoid repetition, the results of control variables are omitted. T values are in parentheses; and

represent significance at the levels of 1%, 5% and 10%, respectively.

CSR, corporate social responsibility.

Dimensions of CSR and debt risk O_score

Variables O_score

(1) (2) (3) (4)

ShareholderR −0.0775*** (−4.90) −0.0738*** (−4.58)
EmployeeR −0.0463*** (−2.79) −0.0643** (−2.46)
ExternalR −0.0067 (−1.38) 0.0121 (1.58)
Controls Yes Yes Yes Yes
Constant −27.8747*** (−54.94) −28.0681*** (−53.99) −27.9260*** (−54.14) −28.1411*** (−54.16)
Year/Industry Yes Yes Yes Yes
N 23,660 23,660 23,660 23,660
Adj-R2 0.3783 0.3779 0.3777 0.3784
F 328.1829 327.5895 327.3737 314.1067

Note: The control variables in Table 10 are the same as those in Table 3. To avoid repetition, the results of control variables are omitted. T values are in parentheses;

represent significance at the levels of 1%, 5% and 10%, respectively.

CSR, corporate social responsibility.

Research conclusions and policy recommendations

Based on the CSR score data published in Hexun from 2010 to 2019 for A-share listed companies in the Shanghai and Shenzhen stock exchanges, this paper makes an empirical study on the impact of CSR fulfilment on debt risk faced by firms under the background of China’s economic system transformation, and assesses whether product-market competition can strengthen the effect of CSR fulfilment on debt risk. First, we find that CSR is helpful to restrain debt risk; this conclusion is still valid after fully considering the endogenous problems caused by variable measurement error and missing important variables. Second, product-market competition can improve the internal governance of firms, reduce the degree of information asymmetry and promote the inhibitory effect of CSR on debt risk; this conclusion is still valid after grouping the degree of product-market competition. Third, there are significant differences in the impact of different layers of CSR fulfilment on the debt risk faced by firms. The inhibition effect of CSR fulfilment in monetary capital layer and human capital layer on debt risk is significantly stronger than that in social capital layer. Therefore, firms should fully understand the positive significance of fulfilment of CSR, strive to promote the integration of CSR fulfilment and firm strategy, give full play to the positive role of CSR in the survival and development of firms, promote the harmonious unity of social interests and firm interests through the implementation of CSR strategy and ultimately realise sustainable development.

Regression results of independent variable lag for one period (Z.score is negative)

Variables Z_score O_score

(1) (2)

CSRt−1 −0.0174*** (−4.82) −0.0095*** (−5.67)
Sizet−1 2.1373*** (39.83) 0.7652*** (30.88)
ROAt−1 −28.8522*** (−25.90) −19.5183*** (−37.95)
Aget−1 0.0127 (1.39) 0.0117*** (2.76)
VCFt−1 8.3963*** (6.47) −0.2604 (−0.43)
overheadt−1 0.6466 (0.31) −15.3937*** (−15.87)
NDTSt−1 12.3036** (2.43) −17.8428*** (−7.63)
MTBt−1 −0.4280*** (−20.11) 0.2122*** (21.59)
Divt−1 0.0453 (0.33) −0.1449** (−2.27)
PPEt−1 3.1000*** (6.15) 3.0327*** (13.04)
Lncont−1 −4.6787*** (−13.07) −2.7233*** (−16.48)
Growtht−1 0.9543*** (8.88) 0.2054*** (4.14)
RETt−1 −6.0977*** (−3.59) −3.2185*** (−4.10)
Indratiot−1 −4.6850*** (−5.11) −0.4761 (−1.12)
Dualityt−1 −0.4769*** (−4.09) 0.0440 (0.82)
Constant −46.6619*** (−37.39) −25.7190*** (−44.64)
Year/Industry Yes Yes
N 20363 20363
r2 0.3128 0.3038
Adj-R2 0.3114 0.3023

CSR fulfilment and debt risk, the moderating effect of product-market competition (Z.score and HHI are negative)

Variables Z_score O_score

(1) (2) (3) (4) (5) (6)

CSR −0.0169*** (−4.66) −0.0099*** (−2.73) −0.0164*** (−3.99) −0.0398*** (−25.38) −0.0073*** (−4.61) −0.0103*** (−5.70)
HHI 14.6110*** (10.46) 4.2472*** (6.93)
CSR*HHI −0.0911*** (−3.02) −0.0435*** (−3.28)
Size 2.0913*** (41.76) 2.0756*** (41.54) 0.8739*** (39.86) 0.8697*** (39.68)
ROA −36.6600*** (−35.63) −36.1459*** (−35.19) −26.9191*** (−59.76) −26.7790*** (−59.45)
Age 0.0488*** (5.77) 0.0522*** (6.19) 0.0215*** (5.82) 0.0225*** (6.08)
VCF 10.7848*** (8.71) 10.7521*** (8.71) −0.5821 (−1.07) −0.5788 (−1.07)
Overhead 4.3665** (2.15) 5.8658*** (2.89) −16.1690*** (−18.21) −15.7934*** (−17.77)
NDTS 15.1886*** (3.21) 14.8717*** (3.15) −20.6733*** (−9.99) −20.8409*** (−10.08)
MTB −0.6836*** (−33.75) −0.6920*** (−34.21) 0.2599*** (29.32) 0.2581*** (29.10)
Div −0.2356* (−1.81) −0.2202* (−1.70) −0.3078*** (−5.41) −0.3027*** (−5.33)
PPE 4.1707*** (8.84) 4.1867*** (8.90) 3.7310*** (18.06) 3.7347*** (18.10)
Lncon −4.8438*** (−14.28) −4.6771*** (−13.81) −2.7487*** (−18.51) −2.7058*** (−18.22)
Growth 1.3257*** (12.81) 1.3227*** (12.82) 0.2641*** (5.83) 0.2626*** (5.80)
RET −12.8841*** (−7.89) −11.9855*** (−7.35) −1.9071*** (−2.67) −1.6842** (−2.35)
Indratio −4.2738*** (−4.95) −4.3505*** (−5.05) −0.7872** (−2.08) −0.8201** (−2.17)
Duality −0.5299*** (−4.84) −0.5266*** (−4.82) 0.0088 (0.18) 0.0095 (0.20)
Constant −7.2702*** (−14.07) −46.4121*** (−39.28) −44.2865*** (−37.10) −10.0410*** (−44.92) −28.1203*** (−54.36) −27.5263*** (−52.57)
Year/Industry Yes Yes Yes Yes Yes Yes
N 23660 23660 23659 23660 23660 23659
Adj-R2 0.0977 0.3779 0.3813 0.1443 0.3960 0.3972
F 86.4271 327.5768 318.0266 133.9788 353.5853 339.8737

Dynamic test (Z.score is negative)

Variables ΔZ_scoret+1 ΔZ_scoret ΔO_scoret+1 ΔO_scoret

(1) (2) (3) (4)

ΔCSRt+1 −0.0131*** (−4.40) 0.0017 (0.59) −0.0208*** (−18.17) −0.0007 (−0.73)
ΔCSRt −0.0163*** (−5.27) −0.0061*** (−5.13)
ΔSize 0.9445*** (5.55) 3.1999*** (18.70) −0.0751 (−1.15) 2.0405*** (33.96)
ΔROA 3.1905*** (3.69) −16.2485*** (−19.27) 5.6848*** (17.16) −15.4860*** (−52.30)
ΔAge 0.3425 (0.43) −1.3515* (−1.69) −0.0981 (−0.32) −0.1314 (−0.47)
ΔVCF 0.6081 (0.45) −2.8218** (−2.09) 0.2477 (0.48) −0.9724** (−2.05)
Δoverhead 12.4276*** (4.08) −7.0720** (−2.31) 4.4970*** (3.85) −10.6482*** (−9.90)
ΔNDTS −10.4221 (−1.44) −13.0303* (−1.79) 3.2840 (1.18) −26.0014*** (−10.15)
ΔMTB 0.0874*** (4.09) −0.4907*** (−22.85) −0.0156* (−1.90) 0.1604*** (21.28)
ΔDiv −0.0005 (−0.01) −0.2662*** (−2.72) 0.1689*** (4.49) −0.3473*** (−10.10)
ΔPPE −0.4916 (−0.75) 5.2796*** (7.98) 0.0463 (0.18) 3.4835*** (14.99)
ΔLncon −3.5143*** (−4.52) −6.5829*** (−8.42) −0.8265*** (−2.77) −3.6601*** (−13.33)
ΔGrowth −0.0156 (−0.23) 0.4864*** (7.23) −0.0286 (−1.12) −0.3487*** (−14.77)
ΔRET −1.1422 (−1.17) −18.2114*** (−18.52) −1.1966*** (−3.19) −2.1178*** (−6.13)
ΔIndratio −0.6330 (−0.57) 0.8220 (0.74) −0.2544 (−0.60) 0.2334 (0.60)
ΔDuality 0.1502 (1.07) −0.1674 (−1.18) −0.0247 (−0.46) −0.0028 (−0.06)
Constant −0.3142 (−0.36) 1.4759* (1.68) 0.2355 (0.71) 0.1100 (0.36)
Year/industry Yes Yes Yes Yes
N 17164 17164 17164 17164
r2 0.0872 0.2325 0.0528 0.2586
Adj-R2 0.0850 0.2306 0.0504 0.2568

PSM (Z.score is negative)

Variables Z_score O_score

(1) (2)

CSR −0.0166*** (−2.99) −0.0093*** (−3.79)
Controls Yes Yes
Constant −46.7106*** (−23.97) −28.4039*** (−33.11)
Year/Industry Yes Yes
N 8599 8599
r2 0.3815 0.3802
Adj-R2 0.3783 0.3771

Variable definitions

Variables classification Variable names Variable definitions
Explained variables FDR Z_score as calculated in model 2.
O_score as calculated in model 3.
Explanatory variables CSR CSR score is obtained from the CSR reports of listed companies published in Hexun.
Control variables Size Natural logarithm of total assets at the end of the year.
ROA ROA is equal to pretax operating income divided by total assets.
Age Age is equal to inspection year minus listing year.
VCF Three year volatility of the ratio of cash flow to total assets at the end of the year.
Overhead Overhead is the management expense that happened during the current year divided by total assets.
NDTS NDTS is the ratio of current year depreciation and amortisation to total assets.
MTB MTB is the growth rate of firms, and it is the ratio of the market value of equity at the end of the year to the book value of equity at the end of the year.
Div Div represents weather the enterprise has had dividend payouts in the current year; if yes it takes the value of 1 and 0 otherwise.
PPE The ratio of total noncurrent assets to total assets.
Lncon Lncon is the combined shareholding of the top five shareholders.
Growth Growth is the growth rate of operating income that equals change in sales revenue of the current year compared to the previous year, and then divided by the previous year’s operating income.
RET RET is the firm’s average monthly stock return during the year.
Indratio The proportion of independent directors to the total number of directors.
Duality Duality is a dummy variable that equals 1 if chairman and the general manager are in one position and 0 otherwise.

Dimensions of CSR and debt risk (Z.score is negative)

Variables Z_score

(1) (2) (3) (4)

ShareholderR −0.0775*** (−4.90) −0.0738*** (−4.58)
EmployeeR −0.0463*** (−2.79) −0.0643** (−2.46)
ExternalR −0.0067 (−1.38) 0.0121 (1.58)
Controls Yes Yes Yes Yes
Constant −46.1237*** (−39.79) −46.5026*** (−39.17) −46.0650*** (−39.11) −46.5855*** (−39.24)
Year/Industry Yes Yes Yes Yes
N 23660 23660 23660 23660
Adj-R2 0.3783 0.3779 0.3777 0.3784
F 328.1829 327.5895 327.3737 314.1067

Dimensions of CSR and debt risk O.score

Variables O_score

(1) (2) (3) (4)

ShareholderR −0.0775*** (−4.90) −0.0738*** (−4.58)
EmployeeR −0.0463*** (−2.79) −0.0643** (−2.46)
ExternalR −0.0067 (−1.38) 0.0121 (1.58)
Controls Yes Yes Yes Yes
Constant −27.8747*** (−54.94) −28.0681*** (−53.99) −27.9260*** (−54.14) −28.1411*** (−54.16)
Year/Industry Yes Yes Yes Yes
N 23,660 23,660 23,660 23,660
Adj-R2 0.3783 0.3779 0.3777 0.3784
F 328.1829 327.5895 327.3737 314.1067

Descriptive statistics

Variables N Minimum Mean p25 Median p75 Std. Dev Maximum

Z_score 23660 0.1785 6.7777 2.0313 3.7837 7.4065 8.8820 55.5422
O_score 23660 −27.3073 −11.5968 −13.3983 −10.7945 −8.8855 3.9466 −5.0486
CSR 23660 −3.1500 24.8365 16.5800 22.1300 27.8000 16.0305 74.5200
Size 23660 19.7510 22.1602 21.2268 21.9870 22.8968 1.2861 25.8845
ROA 23660 −0.2157 0.0394 0.0147 0.0368 0.0657 0.0557 0.1962
Age 23660 0.0000 10.0511 4.0000 9.0000 16.0000 7.0084 25.0000
VCF 23660 0.0022 0.0453 0.0189 0.0339 0.0586 0.0392 0.2164
Overhead 23660 0.0041 0.0447 0.0248 0.0394 0.0581 0.0282 0.1544
NDTS 23660 0.0007 0.0238 0.0119 0.0209 0.0325 0.0157 0.0763
MTB 23660 0.5996 3.5910 1.7365 2.6862 4.3403 3.0321 19.1986
Div 23660 0.0000 0.7555 1.0000 1.0000 1.0000 0.4298 1.0000
PPE 23660 0.0022 0.2186 0.0889 0.1845 0.3128 0.1656 0.7144
Lncon 23660 0.2021 0.5356 0.4230 0.5381 0.6489 0.1523 0.8804
Growth 23660 −0.5395 0.1967 −0.0090 0.1159 0.2795 0.4604 3.1335
RET 23660 −0.0686 0.0102 −0.0188 0.0052 0.0319 0.0421 0.1592
Indratio 23660 0.3333 0.3748 0.3333 0.3333 0.4286 0.0536 0.5714
Duality 23660 0.0000 0.2575 0.0000 0.0000 1.0000 0.4373 1.0000

Replacement of main variables (Z.score is negative)

Variables Adjusted by industry annual mean value

m_Z_score m_O_score

(1) (2)

m_CSR −0.0126*** (−3.49) −0.0090*** (−5.68)
Controls Yes Yes
Constant −40.0503*** (−33.60) −17.4561*** (−33.27)
Year/Industry Yes Yes
N 23660 23660
Adj-R2 0.3019 0.3023

Grouping by product-market competition intensity (Z.score and HHI are negative)

Variables Z_score O_score

High competition level Low competition level High competition level Low competition level

(1) (2) (3) (4)

CSR −0.0236*** (−4.45) −0.0030 (−0.60) −0.0157*** (−6.55) −0.0001 (−0.02)
Controls Yes Yes Yes Yes
Constant −51.6314*** (−31.24) −43.2057*** (−28.27) −34.0105*** (−45.51) −26.3189*** (−39.05)
Year/Industry Yes Yes Yes Yes
N 11830 11829 11830 11829
r2 0.3654 0.3780 0.3770 0.3403
Adj-R2 0.3641 0.3767 0.3757 0.3390

Zhang Zhengyong. Market Competition in Product, the Corporate Governance and the CSR Information Disclosure: Empirical Evidence from CSR Bulletin of Chinese Listed Companies [J]. Journal of Shanxi University of Finance and economics, 2012, 34 (04): 67–76. ZhangZhengyong Market Competition in Product, the Corporate Governance and the CSR Information Disclosure: Empirical Evidence from CSR Bulletin of Chinese Listed Companies [J] Journal of Shanxi University of Finance and economics 2012 34 04 67 76 Search in Google Scholar

Goss A, Roberts G S. The impact of CSR on the cost of bank loans [J]. Journal of Banking & Finance, 2011, 35(7):1794–1810. GossA RobertsG S The impact of CSR on the cost of bank loans [J] Journal of Banking & Finance 2011 35 7 1794 1810 10.1016/j.jbankfin.2010.12.002 Search in Google Scholar

Shi G, Sun J. Corporate Bond Covenants and CSR Investment [J]. Journal of Business Ethics, 2015, 131(2):285–303. ShiG SunJ Corporate Bond Covenants and CSR Investment [J] Journal of Business Ethics 2015 131 2 285 303 10.1007/s10551-014-2272-y Search in Google Scholar

Attig N, Ghoul SE, Guedhami O, et al. CSR and Credit Ratings[J]. Journal of Business Ethics, 2013, 117(4):679–694. AttigN GhoulSE GuedhamiO CSR and Credit Ratings[J] Journal of Business Ethics 2013 117 4 679 694 10.1007/s10551-013-1714-2 Search in Google Scholar

Stigler, George. The economies of scale [J]. Journal of Law and Economics, 1958, 1: 54–71. StiglerGeorge The economies of scale [J] Journal of Law and Economics 1958 1 54 71 10.1086/466541 Search in Google Scholar

Holmstrom, B. Moral hazard in teams [J], Bell Journal of Economics, 1982, 13 (2):324–340. HolmstromB Moral hazard in teams [J] Bell Journal of Economics 1982 13 2 324 340 10.2307/3003457 Search in Google Scholar

Wu Haomin, Yang Xingquan, Wei Hui. Product Market Competition and Firm’s Stock Idiosyncratic Risk: Based on the Empirical Evidence of Chinese Listed Companies [J]. Economic Research Journal, 2012 (06): 101–115. WuHaomin YangXingquan WeiHui Product Market Competition and Firm’s Stock Idiosyncratic Risk: Based on the Empirical Evidence of Chinese Listed Companies [J] Economic Research Journal 2012 06 101 115 Search in Google Scholar

Han Zhongxue, Zhou Tingting. Product Market Competition, Financing constraints and Cash Holdings: An Empirical Study Based on Enlisted Companies in China’s Manufacture Industry [J]. Nankai Business Review, 2011 (4): 149–160. HanZhongxue ZhouTingting Product Market Competition, Financing constraints and Cash Holdings: An Empirical Study Based on Enlisted Companies in China’s Manufacture Industry [J] Nankai Business Review 2011 4 149 160 Search in Google Scholar

Altman EI. FINANCIAL RATIOS, DISCRIMINANT ANALYSIS AND THE PREDICTION OF CORPORATE BANKRUPTCY[J]. Journal of Finance, 1968, 23(4):589–609. AltmanEI FINANCIAL RATIOS, DISCRIMINANT ANALYSIS AND THE PREDICTION OF CORPORATE BANKRUPTCY[J] Journal of Finance 1968 23 4 589 609 10.1111/j.1540-6261.1968.tb00843.x Search in Google Scholar

Chen Caiyun, Tang Xiangxi. Innovation Investment, Tax Avoidance and Corporate Risk [J]. Modern finance and Economics (Journal of Tianjin University of Finance and Economics), 2019, 39 (11): 14–33. ChenCaiyun TangXiangxi Innovation Investment, Tax Avoidance and Corporate Risk [J] Modern finance and Economics (Journal of Tianjin University of Finance and Economics) 2019 39 11 14 33 Search in Google Scholar

Ohlson JA. Financial ratios and the probabilistic prediction of bankruptcy [J]. Journal of Accounting Research, 1980, 18 (1):109–131. OhlsonJA Financial ratios and the probabilistic prediction of bankruptcy [J] Journal of Accounting Research 1980 18 1 109 131 10.2307/2490395 Search in Google Scholar

Black F, Scholes MS. The Pricing of Options and Corporate Liabilities[J]. Journal of Political Economy, 1973, 81(3):637–654. BlackF ScholesMS The Pricing of Options and Corporate Liabilities[J] Journal of Political Economy 1973 81 3 637 654 10.1142/9789814759588_0001 Search in Google Scholar

Merton R C. On the pricing of corporate debt: the risk structure of interest rates[J]. Journal of Finance, 1974, 29(2):449–470. MertonR C On the pricing of corporate debt: the risk structure of interest rates[J] Journal of Finance 1974 29 2 449 470 10.1142/9789814759588_0003 Search in Google Scholar

Shumway T. Forecasting Bankruptcy More Accurately: A Simple Hazard Model[J]. The Journal of Business, 2001, 74(1):101–124. ShumwayT Forecasting Bankruptcy More Accurately: A Simple Hazard Model[J] The Journal of Business 2001 74 1 101 124 10.1086/209665 Search in Google Scholar

Campbell JY, Hilscher J, Szilagyi J. In Search of Distress Risk[J]. Journal of Finance, 2008, 63(6):2899–2939. CampbellJY HilscherJ SzilagyiJ In Search of Distress Risk[J] Journal of Finance 2008 63 6 2899 2939 10.3386/w12362 Search in Google Scholar

Richardson G, Lanis R, Taylor GL. Financial distress, outside directors and corporate tax aggressiveness spanning the global financial crisis: An empirical analysis[J]. Journal of Banking & Finance, 2015, 52(mar.):112–129. RichardsonG LanisR TaylorGL Financial distress, outside directors and corporate tax aggressiveness spanning the global financial crisis: An empirical analysis[J] Journal of Banking & Finance 2015 52 mar. 112 129 10.1016/j.jbankfin.2014.11.013 Search in Google Scholar

Agarwal V, Taffler R. Comparing the performance of market-based and accounting-based bankruptcy prediction models[J]. Journal of Banking & Finance, 2008, 32(8):1541–1551. AgarwalV TafflerR Comparing the performance of market-based and accounting-based bankruptcy prediction models[J] Journal of Banking & Finance 2008 32 8 1541 1551 10.1016/j.jbankfin.2007.07.014 Search in Google Scholar

Altman EI, Iwanicz-Drozdowska M, Laitinen EK, et al. Financial Distress Prediction in an International Context: A Review and Empirical Analysis of Altman’s Z-Score Model[J]. Journal of International Financial Management & Accounting, 2017, 28(2):131–171. AltmanEI Iwanicz-DrozdowskaM LaitinenEK Financial Distress Prediction in an International Context: A Review and Empirical Analysis of Altman’s Z-Score Model[J] Journal of International Financial Management & Accounting 2017 28 2 131 171 10.1111/jifm.12053 Search in Google Scholar

Altman, Edward I, Corporate Distress Prediction Models in a Turbulent Economic and Basel Environment. 2002, NYU Working Paper No. S-CDM-02-11. AltmanEdward I Corporate Distress Prediction Models in a Turbulent Economic and Basel Environment 2002 NYU Working Paper No. S-CDM-02-11. Search in Google Scholar

Missing citation “Peress 2010” Missing citation “Peress 2010” Search in Google Scholar

Duan Junshan, Zhuang Xudong. Research on the Hierarchical Influencing Mechanism of CSR Fulfillment on Firm Innovation Ability: Empirical Evidence Based on Chinese Listed Companies [J]. South China Journal of Economics, 2020 (08): 49–64. DuanJunshan ZhuangXudong Research on the Hierarchical Influencing Mechanism of CSR Fulfillment on Firm Innovation Ability: Empirical Evidence Based on Chinese Listed Companies [J] South China Journal of Economics 2020 08 49 64 Search in Google Scholar

Dhaliwal, Dan, S, et al. Voluntary Nonfinancial Disclosure and the Cost of Equity Capital: The Initiation of CSR Reporting. [J]. Accounting Review, 2011, 86(1):59–100. DhaliwalDan, S Voluntary Nonfinancial Disclosure and the Cost of Equity Capital: The Initiation of CSR Reporting. [J] Accounting Review 2011 86 1 59 100 10.2308/accr.00000005 Search in Google Scholar

Bae KH, Rang JK, Wang J. Employee treatment and firm leverage: A test of the stakeholder theory of capital structure [J]. Journal of Financial Economics, 2011, 100(1):130–153. BaeKH RangJK WangJ Employee treatment and firm leverage: A test of the stakeholder theory of capital structure [J] Journal of Financial Economics 2011 100 1 130 153 10.1016/j.jfineco.2010.10.019 Search in Google Scholar

Godfrey PC, Merrill CB, Hansen JM. The relationship between CSR and shareholder value: an empirical test of the risk management hypothesis [J]. Strategic Management Journal, 2010, 30(4):425–445. GodfreyPC MerrillCB HansenJM The relationship between CSR and shareholder value: an empirical test of the risk management hypothesis [J] Strategic Management Journal 2010 30 4 425 445 10.1002/smj.750 Search in Google Scholar

Freeman RE, Harrison JS, Wicks A. 2008. Managing for Stakeholders: Survival, Reputation, and Success. Yale University Press: New Haven, CT. FreemanRE HarrisonJS WicksA 2008 Managing for Stakeholders: Survival, Reputation, and Success Yale University Press New Haven, CT Search in Google Scholar

Aghion P, Griffith R. Competition and growth: reconciling theory and evidence. [M]. MIT Press, 2008. AghionP GriffithR Competition and growth: reconciling theory and evidence. [M] MIT Press 2008 Search in Google Scholar

Flammer C. Does product market competition foster CSR? Evidence from trade liberalization [J]. Strategic Management Journal, 2015, 36(10):1469–1485. FlammerC Does product market competition foster CSR? Evidence from trade liberalization [J] Strategic Management Journal 2015 36 10 1469 1485 10.1002/smj.2307 Search in Google Scholar

Freeman RE. 1984. Strategic Management: A Stakeholder Approach. Pitman: Boston, MA. FreemanRE 1984 Strategic Management: A Stakeholder Approach Pitman Boston, MA Search in Google Scholar

Friedman M. The CSR of Business is to Increase its Profits [J]. New York Times Magazine, 1970, 13(33):173–178. FriedmanM The CSR of Business is to Increase its Profits [J] New York Times Magazine 1970 13 33 173 178 Search in Google Scholar

Polecane artykuły z Trend MD

Zaplanuj zdalną konferencję ze Sciendo