Working Capital Management and Bank Mergers

Abstract

This study investigates the consequences of bank mergers on non-financial borrowers’ working capital management. We find evidence that bank mergers increase corporate cash holdings and decrease receivables and investments in inventories by reducing bank credit availability. Bank mergers also decrease trade credit used through the reduction in bank credit availability. These findings are new contributions to the literature, suggesting that borrowing firms find it more difficult to manage working capital after bank mergers occur and that bank-dependent firms find it more difficult to manage working capital than their non-dependent counterparts after mergers.

1. Introduction

The existing literature has extensively studied the consequences of banks’ mergers and acquisitions (M&As) on bank performance, efficiency, and market power (Amel et al. 2004). On the borrower side, the effects on banks’ credit availability or lending policies have been studied (Berger et al. 1995; Bonaccorsi di Patti and Gobbi 2007). Bank mergers have attracted academic attention because they are more than just the merger of multiple banks (usually two) into a single bank. New management changes the basic policies on credit, deposits, and other service activities, and organizational changes involve employee turnover, officer rearrangement, and branch downsizing. The number of bank M&A transactions in Japan have increased significantly over the past two decades. These deals have not only reshaped the competitive landscape but have also had a profound impact on borrowers, affecting their financial management policies and strategies. This study raises the research question: what are the consequences of bank mergers on borrower working capital management? Despite extensive research on the financial impacts of bank mergers, little attention has been paid to how these mergers affect corporate account receivables, a critical component of working capital management. This study determines the consequences of bank mergers on borrowers’ working capital management and addresses whether mergers weaken the bank function of liquidity provision, bank information technology, or both.

The literature on working capital management is extensive (Emery 1987; Brennan et al. 1988; Long et al. 1993; Fazzari and Petersen 1993; Petersen and Rajan 1994, 1997; Summers and Wilson 2000; Burkart and Ellingsen 2004; Bates et al. 2009; Molina and Preve 2009; Hill et al. 2010; Kieschnick et al. 2013; Francis et al. 2014; Aktas et al. 2015; Chen and Kieschnick 2018; Kieschnick and Rotenberg 2016; Kling et al. 2014). This body of work has investigated various aspects of working capital, from its role as a source of internal funds to the strategies firms employ in managing it, particularly in response to external financial constraints and opportunities. Moreover, significant literature has examined the effects of bank mergers, highlighting their potential to improve the merged entity’s profitability and market position, manage funding liquidity risk, and affect credit availability and borrowing costs.

Our study empirically investigates the relationship between bank mergers and nonfinancial firms’ working capital management and sheds some new light on working capital management in the banking merger literature. This study provides the following empirical findings. (i) Bank mergers decrease the rate of change in total bank borrowings. (ii) Additionally, bank mergers increase cash holdings and other current assets while decreasing account receivables, inventories, trade credit used, and other current liabilities. (iii) Bank-dependent firms tend to hold more cash and inventories than non-bank-dependent firms when bank mergers occur. The first analysis finds that bank mergers result in a decline in the growth of bank borrowings, which means that bank mergers decrease the bank credit availability of firm borrowers. The second analysis investigates the relationship between bank mergers and working capital management, considering the endogenous effect of bank mergers on the growth rate of bank borrowings and employing the panel data quasi-maximum likelihood method (Papke and Wooldridge 2008). We find that bank mergers increase corporate cash holdings and other current assets and decrease the receivables, inventories, trade credit used, and other current liabilities through a decrease in the growth rate of bank borrowing. These findings are new contributions to the banking merger literature and suggest that bank mergers weaken the liquidity-providing function of banks and weaken the information technology of banks. Specifically, this study makes a significant contribution to the field by being the first to investigate the relationship between bank mergers and corporate account receivables. Also, this study offers insights into how bank consolidations influence the liquidity and credit management strategies of corporations, contributing to a deeper understanding of the broader economic implications of bank mergers. In the third analysis, we allow the effect of bank mergers on working capital through bank borrowings to depend on the bank dependency. Chen and Kieschnick (2018) found that bank-dependent firms’ working capital responses to changes in bank credit availability differ from those of less bank-dependent firms in some areas. Our evidence shows that bank mergers have a greater impact on bank-dependent firms’ corporate cash holdings and inventories than on non-bank-dependent firms. As a robustness check, we use the propensity score matching with difference-in-difference estimation method (PSM-DID) to identify the bank merger effect. We show that our results are robust to the potential endogeneity of bank mergers.

The subsequent sections are organized as follows. Section 2 reviews the literature and develops hypotheses, Section 3 describes the data and variables and states the econometric specifications and methodology, and Section 4 provides empirical evidence. First, we report the impact of bank mergers on bank credit growth. Second, we investigate the relationship between working capital and bank mergers by considering the endogeneity of bank borrowings. Third, we investigate the effects of changes in bank credit and bank mergers on bank-dependent and non-bank-dependent firms. Finally, we provide the PSM-DID results as a robustness check. Section 5 concludes the paper.

2. Literature Review and Hypothesis Development

A significant portion of the literature has examined the effects of bank mergers. Among them, Went (2003) showed that bank mergers improve the profitability, market position, and customer service of the merged institution. Similarly, Ly and Shimizu (2018) investigated the management of funding liquidity risk by a multi-bank holding company (MBHC) through its internal liquidity market, providing evidence that the diversification effect occurs after the bank merger and that member banks benefit from diversifying risk when a new entrant joins.

Although a large amount of the literature has examined the effects of bank mergers, the specific effect on borrowers, particularly concerning their working capital management, has been relatively under-researched. One influential study examining the relationship between bank mergers and borrower performance was conducted by Degryse et al. (2011). They examined the bank–borrower relationship after a bank merger. Following a bank merger, soft information available at the target bank may be lost if key employees leave or move within the merging bank (Degryse et al. 2011, p. 1114). Hence, borrowers are more harmed when the relationship is dropped after a merger than otherwise. For example, the asset growth of droppers after a merger is lower than that of stayers and switchers. Additionally, Pasiouras et al. (2007) determined that acquired banks are less profitable and less efficient in managing expenses. Fraisse et al. (2018) indicated that merging banks reduce their lending in local markets where their market shares overlapped before the merger relative to local markets in which at least one of the merging banks had a small market share. Montgomery and Takahashi (2018) found evidence that a merger announcement by a firm’s main bank results in a contraction in credit supply from the merging bank. Also, the strategic financial management of corporate borrowers may need to be adjusted in light of a bank merger. This includes adjusting investment plans, cash flow management, and risk management strategies to accommodate new banking practices or to mitigate the adverse impact of the merger on financing conditions (Ogada et al. 2017). Dzhagityan (2018) showed that mergers and acquisitions serve as strategic interventions within the banking sector, potentially offering both opportunities and challenges to corporate borrowers. They can be pivotal in reshaping the competitive landscape and the availability of financial resources for borrowers. This suggests that bank mergers might weaken the liquidity-providing function of banks, compelling firms to adjust their financial management strategies, including investment plans and risk management approaches, to mitigate the adverse effects of reduced financing availability (Ogada et al. 2017).

The impact of bank mergers on the cost of borrowing can be significant. Interest rates on loans may decrease if the merger leads to increased efficiency and competition within the local market. However, if the merger significantly decreases competition, the cost of borrowing may increase as the consolidated bank exploits its market power (Gachigo et al. 2022). This observation is supported by Bonaccorsi di Patti and Gobbi (2007), who examined the impact of bank M&As on outstanding credit. They found evidence that corporate borrowers experience a reduction in outstanding credit and that the reduction depends on the share of credit borrowed from the merged banks. Additionally, relationship termination has a more significant adverse impact on credit volumes.

Moreover, the analysis extends into how mergers affect loan contracts and the broader competitive landscape of the banking sector, influencing interest rate margins and bank stock prices. Sapienza (2002) examined the effects of bank mergers on loan contracts. After the in-market merger, loan interest rates charged by the consolidated banks decrease substantially. After out-of-market mergers, i.e., mergers between banks previously operating in different provinces, the decrease in interest rates is not as significant; hence, in-market mergers generate higher efficiency than out-of-market mergers. Additionally, Chortareas et al. (2012) investigated the determinants of interest rate margins, and they found that increased competition in the banking sector tends to reduce interest rate margins. Asimakopoulos and Athanasoglou (2013) examined the impact of bank M&A announcements on bank stock prices. Bank merger announcements are positively associated with the stock price of the target bank, while they do not create value for the acquirer’s shareholders. Furthermore, Karceski et al. (2005) examined the impact of bank merger announcements on borrowers’ stock prices. Bank merger announcements are associated with stock price declines for the target borrowers, especially smaller target borrowers in large bank mergers, and, to a lesser extent, stock price increases for acquiring borrowers. These results suggest that merged banks tend to adopt practices that favor acquiring borrowers over target borrowers. Additionally, they found evidence suggesting that firms with low switching costs may switch banks, while similar firms with high switching costs remain locked into their current relationship.

Among the literature on working capital management, Brennan et al. (1988) argued that firms with market power should optimally engage in vendor financing if credit customers have lower reservation prices than cash customers or if adverse selection makes it infeasible to write credit contracts that separate customers according to their credit risk. Fazzari and Petersen (1993) suggested that working capital is considered a source of internal funds. Hill et al. (2010) argued that sales growth, the uncertainty of sales, costly external financing, and financial distress encourage firms to pursue more aggressive working capital strategies, while firms with greater internal financing capacity and superior capital market access employ more conservative working capital policies. Petersen and Rajan (1997) argued that financially constrained firms use trade credit when credit from financial institutions is unavailable. Additionally, Kieschnick and Rotenberg (2016) found that small firms tend to hold excess cash to hedge business risks compared to large firms, suggesting that small and large firms have very different working capital management strategies.

Although these studies focus on borrowers’ credit, investment, stock price, and loan rate, little existing research has investigated the relationship between bank mergers and borrowers’ working capital management. As an exception, Fraisse et al. (2018) investigated the effect of a merger between two large banks on trade credit used. Accompanied by a reduction in lending after the bank merger, an insignificant substituting result of trade credit for bank credit is found. They argue that firms have local suppliers or customers affected by the same credit supply shock as themselves. Furthermore, Francis et al. (2014) focused on the cash policy of nonfinancial corporations after the merger of banks and found that the intensity of in-market bank mergers is negatively related to corporate cash holdings.

Our study builds on these insights, focusing specifically on the under-researched area of bank mergers’ implications for borrower firms’ working capital management. By integrating the existing findings with our empirical investigation, we aim to shed new light on how bank consolidations impact borrowers’ financial strategies, providing a comprehensive understanding that bridges the gap in the current literature. Specifically, we extend the work of Chen and Kieschnick (2018), who investigated how changes in the availability of bank credit can influence borrower firms’ working capital management, and analyzed its relationship with bank mergers. We not only analyze the impact of bank mergers on changes in bank credit but also clarify the link between bank credit and working capital management and calculate the merger effect to clarify the impact of bank mergers on borrowing firms’ working capital management. We consider that bank mergers affect bank credit availability, and changes in bank credit availability are closely related to working capital management. We link these two streams of the literature, i.e., bank mergers’ literature and nonfinancial firms’ working capital management literature, and shed new light on corporate working capital management in the context of bank mergers.

Bank M&As negatively shock the credit supply because they impact bank–client relationships. As emphasized by Singh et al. (2023), it is crucial to understand the intricate linkages between the various elements of the banking system, of which the customer relationship is a key element. The information generated by the relationships is difficult to transfer to the post-merger banks because bank M&As lead to the loss of knowledge accumulated within each merging bank due to organizational changes, employee turnover, and branch downsizing after the bank merger (Bonaccorsi di Patti and Gobbi 2007). We expect bank mergers to negatively affect changes in firms’ total bank borrowings. However, because of bank mergers, bank size changes affect the bank loan supply (Stein 2002), and the loan market structure impacts changes in bank size on credit availability (Berger et al. 2007; Jayaratne and Wolken 1999). Furthermore, bank mergers reduce banks’ information-gathering costs and increase bank efficiency, which increases overall market efficiency (Focarelli et al. 2002). Thus, we develop the first hypothesis about the effect of bank mergers on changes in bank credit availability:

Hypothesis 1.

A bank merger decreases the rate of change in bank borrowings.

The second analysis investigates each component of working capital management. The working capital comprises cash holdings (CASH), receivables (RCVs), inventories (INVs), other current assets (OCAs), trade credit used (accounts payable) (TC), and other current liabilities (OCLs).

Harford et al. (2014) found that refinancing risk is a key determinant of cash holdings, and firms mitigate refinancing risk by increasing their cash holdings. Kling et al. (2014) also found evidence that a short-term liquidity shock leads to more cash holdings. That is, if bank credit increases and the probability that a firm can pay its debts rises, the firm should reduce its cash holdings, consistent with Hypothesis 1 explained by Chen and Kieschnick (2018). If our Hypothesis 1 (i.e., bank merger results in a decrease in the rate of change in total bank borrowings) holds, the bank merger results in an increase in corporate cash holdings through a decrease in the rate of change in total bank borrowings. To the best of our knowledge, no existing literature has studied the impacts of bank mergers on corporate cash holdings except for Francis et al. (2014). Francis et al. (2014) suggested that the merged bank can pass on its synergies and efficiency gains to bank customers. After the bank merger, corporate borrowers benefit from lower interest rates and improved access to external capital; hence, there is less incentive to hoard cash. For the impact of bank mergers on corporate cash holdings, we develop the following hypothesis:

Hypothesis 2.

A bank merger increases corporate cash holdings (CASH).

No research has investigated the relationship between bank mergers and corporate account receivables.

Yang (2011) tested the relationship between account receivables and bank credit, finding a complementary effect, i.e., a positive effect, between bank credit and account receivables. Lin and Chou (2015) also found a positive relationship between account receivables and bank credit. According to Hypothesis 3 explained by Chen and Kieschnick (2018), all else being equal, an increase in bank credit would increase the credit that the firm extends to its customers. In addition, firms with better access to credit may offer more trade credit to their customers (Petersen and Rajan 1997). Furthermore, Love et al. (2007) found that account receivables decrease after a crisis, and they argue that the reduction in account receivables is driven by the reduction in the supply of bank credit; hence, we predict that if bank credit is less (more) available after the bank merger, the borrowing firm provides less (more) credit to its customers. We expect that bank mergers are negatively associated with corporate account receivables.

Regarding account receivables, we develop the following hypothesis:

Hypothesis 3.

A bank merger decreases account receivables (RCVs).

Inventories and other current assets can be regarded as a substitute for cash (Bates et al. 2009) or a source of internal funds, i.e., a readily reversible store of liquidity (Fazzari and Petersen 1993, pp. 329, 335). If bank credit becomes readily available, firms meet the lower refinancing risk and do not need to hold too much cash; therefore, firms would increase their investment in inventories to substitute cash holdings. Moreover, according to Chen and Kieschnick (2018), increased bank credit availability lowers the cost of bank financing, allowing firms to increase investments in inventories. As a result, we expect that bank mergers lead to reduced inventories due to decreased bank credit availability.

Next, we consider the following hypotheses:

Hypothesis 4.

A bank merger decreases inventories (INVs).

Among other current asset accounts, reverse trade credit or supply chain financing is also included (Chen and Kieschnick 2018). Prior reports in the literature suggest that this form of financing is a substitute for bank financing or trade credit, and then we can expect that firms can increase this form of financing as bank credit becomes less available due to bank mergers.

Hypothesis 5.

A bank merger increases other current assets (OCAs).

No studies have examined the relationship between bank mergers and TC. Since firms often use either bank credit or TC to finance their investments, these sources of financing may be complements or substitutes (Chen and Kieschnick 2018, p. 582). Burkart and Ellingsen (2004) suggested that TC and bank credit are complements for financially constrained firms, and Uesugi and Yamashiro (2008) argued that TC and bank loans are complementary for small firms in Japan. These analyses suggest that trade credit and bank credit cannot be substitutes, but they are complementary. That is, the increase in bank loans can increase the use of TC; however, there is evidence that firms use more TC when credit from banks is unavailable (Petersen and Rajan 1997). Furthermore, Yang (2011) found a substitute effect between bank credit and accounts payable. According to Hypothesis 1, we predict that bank mergers result in a reduction in the growth of total bank borrowings; therefore, we expect that bank mergers result in a decrease in TC used.

Mateut (2014) found that firms’ use of advanced prepayment by customer firms (OCL) positively correlates with the prepayments received from customers (RCV). Therefore, we expect a positive correlation between bank credit and the company’s other current liabilities accounts. That is, bank mergers reduce OCLs.

We develop the following two hypotheses for the liability side of working capital:

Hypothesis 6.

The bank merger results in the reduction in trade credit (TC) used.

Hypothesis 7.

A bank merger reduces other current liabilities (OCLs).

Chen and Kieschnick (2018) suggested that working capital management by bank-dependent and non-bank-dependent firms differ significantly. They found evidence that bank-dependent firms tend to hold more cash, extend and use more TC, and hold larger inventories than less bank-dependent firms. These findings suggest that bank dependence is closely related to working capital management. The working capital responses of bank-dependent firms to changes in the availability of bank credit differ from those of less bank-dependent firms in some areas.

The last hypothesis that we consider is as follows:

Hypothesis 8.

The impacts of bank mergers on working capital through bank credit availability of bank-dependent firms are greater than those of non-bank-dependent firms.

Summarizing our development of hypotheses, we consider Hypotheses 2–7 to determine which component contributes to the reduction or increase in the working capital under Hypothesis 1. We also consider Hypothesis 8 to identify the effect of bank mergers on bank-dependent firms’ working capital management.

3. Data and Methodology

3.1. Data and Variables

Our sample of listed nonfinancial firms in Japan comes from the Nikkei NEEDS database from 2007 to 2019; financial institutions were excluded. Table 1 reports the number of sample firms in each year. The number of firms per year is 1752, and we have 22,776 firm-year observations. The summary of bank merger activities in Japan between 2007 and 2019 is reported in Appendix B. These mergers reflect the relentless efforts of the Japanese banking industry to adapt to the challenging domestic and international economic environment and to improve operational efficiency and financial stability.

Table 1 also presents the number of firm observations that experienced bank mergers. We construct two bank merger dummy variables. The first dummy variable MGR takes a value of one if one of the firm’s lenders merged or were merged and zero otherwise. The second dummy variable, TMGR, takes a value of one if the firm’s top lender merged or was merged and zero otherwise. The top lender of firm i is defined as the lender whose amount of loans to firm i is the largest among all lenders. The total number of firm-year observations that experienced any bank mergers is 591, and the number of firm-year observations that experienced top lender mergers is 217.

Table 2 provides the summary statistics and non-parametric hypothesis testing result of the balanced data set for the working capital, firm performance, and bank borrowing variables. The definitions of the variables are in Appendix A. All of the working capital variables are normalized by dividing each item by total assets.

The summary statistics in Panel A show that the mean of cash and short-term securities denoted by CASH is 17.64%. Account receivables denoted by RCVs are, on average, 20.36%. Among other components of working capital, inventories denoted by INVs are, on average, 11.47%. Other current assets are denoted by OCAs, which has a mean of 4.70%. On the current liability side, TC is the accounts payable, also called trade credit used, representing borrowings from the supplier from which a firm purchases goods; the mean of TC is 13.10%. Other current liabilities are denoted by OCLs, which has a mean of 18.37%. We do not consider income tax payable in current liabilities as Chen and Kieschnick (2018) do because this is unlikely to be affected by fluctuations in bank credit. Panel A of Table 2 also provides the summary statistics for each component of working capital by bank merger status. Most of these variables indicate significant differences in means between the subsample of MGR = 1 and that of MGR = 0. CASH and OCAs are remarkably lower for firms experiencing bank mergers than for firms that do not experience bank mergers; however, RCVs, INVs, TC, and OCLs are higher for the former than the latter.

As summarized in Table 2, we consider several determinants of working capital management. First, following Chen and Kieschnick (2018), we consider that a firm can borrow more when tangible assets, denoted by TNGs, increase because of the mitigated agency costs of debt because tangible assets serve as collateral to diminish the risk of the lender (Rajan and Zingales 1995). Additionally, tangible assets control a firm’s dependence on working capital finance.

A firm that faces higher cash flow volatility may find it optimal to have a higher inventory level due to a precautionary motive. Additionally, such a firm may have a higher level of receivables to smooth the timing of settlements of the goods that it sells to buyers (Emery 1987; Long et al. 1993). Furthermore, firms facing more uncertain cash flow have higher liquidity needs and tend to rely on TC (accounts payable) to enhance cash flow (Hill et al. 2010). Following De Veirman and Levin (2018) and Chen and Kieschnick (2018), our cash flow volatility measure is based on the residuals of regressions of the following form:

(1)${\gamma }_{i,t}={\alpha }_0+{\alpha }_{1}YEAR+{\alpha }_{2}INDUSTRY+{\alpha }_{3}SIZE+{\alpha }_{4}YEAR\times INDUSTRY+{\alpha }_{5}YEAR\times SIZE+{ϵ}_{i,t}$

where ${\gamma }_{i,t}$ represents the growth in firm i’s operating cash flow from time t − 1 to t, $YEAR$ is a vector of year dummies, $INDUSTRY$ is a vector of industry dummies, and $SIZE$ is a vector of size dummies. We obtain residuals, ${\widehat{ϵ}}_{i,t}$. Second, we estimate the standard deviation of the residual by a term proportional to the absolute value of the estimated residual as ${\widehat{\sigma }}_{i,t}=\sqrt{\pi ∕2}\left|{\widehat{ϵ}}_{i,t}\right|$ under the assumption of a normal distribution of the residuals. Finally, we define the cash flow volatility measure lnCFV as ln($\widehat{\sigma }$).

The third determinant of working capital is firm size, which can be considered a proxy for a firm’s creditworthiness; larger firms can borrow more. The variable lnTA is a natural log of the book value of total assets.

The profit margin, denoted by PM, is considered as one of the essential determinants of working capital (Chen and Kieschnick 2018). All else being equal, as the PM increases, the working capital increases because account receivables increase more than accounts payable by PM per good when selling an additional one unit of a good.

Sales growth, denoted by SG, is considered to be directly related to receivables (Petersen and Rajan 1997). Research and development (R&D) intensity, denoted by RD, is the ratio of a firm’s R&D expense to its sales. A firm with more significant R&D expense has greater costs of growth opportunities, resulting in increased financial distress costs. To avoid such costs, a firm in this position tends to hold more cash (Bates et al. 2009). We employ real GDP growth, denoted by GDP, to control for macroeconomic factors.

The other variables listed in Table 2 are related to bank borrowings, which are important in this study. Total bank borrowings (TBs) are the sum of borrowings from each bank by each firm. TBG is the percentage change in total borrowings from banks. The sample mean of TBG is −1.18%.

Comparing the means of the growth of total bank borrowings TBG between the subsample of MGR = 1 and that of MGR = 0, the mean of TBG decreases by 3.46% (= −1.09–(−4.55)) after the bank merger. The difference is not statistically significant. This evidence is consistent with Bonaccorsi di Patti and Gobbi (2007), who found that bank borrowing is negatively associated with bank mergers.

Panel B of Table 2 presents the results of the Mann–Whitney U test conducted to compare the median values of various working capital variables, firm performance variables, and bank borrowing variables between the subsample of MGR = 1 and that of MGR = 0. The number of observations (Obs) and medians are reported for each group, along with the test statistics (Z) and p-values. Overall, most of the variables indicate significant differences in medians between the subsample of MGR = 1 and that of MGR = 0.

3.2. Econometric Specifications

3.2.1. Impacts of Bank Mergers on Growth in Bank Credit

To test Hypothesis 1, we investigate the impact of bank mergers on growth in bank credit. We consider the following linear regression model to clarify the relationship between bank mergers and bank credit:

(2)$TB{G}_{i,t}={\alpha }_0+{\alpha }_{1}MG{R}_{i,t}+{\alpha }_2{X}_{i,t}+{\xi }_t+{\eta }_i+{ϵ}_{i,t}$

where $TB{G}_{i,t}$ represents the percentage change of bank credit of firm i in year t. In the specification, we use the bank merger dummy $MG{R}_{i,t}$ or the top lender merger dummy $TMG{R}_{i,t}$. Following Chen and Kieschnick (2018), we employ the disposal of non-performing loans (Disposal) and the charge-off amount of loans (Charge-off) as instrumental variables to alleviate an endogeneity issue surrounding Equation (2). These variables are considered to be correlated with a supply factor but not with a demand factor because these variables reflect the banks’ credit supply capacity, which is crucial for examining how M&As might alter lending practices.

The coefficient ${\alpha }_1$ measures the effect of bank mergers MGR on the growth in bank credit. $X$ denotes a set of control variables, and ${\xi }_t$ and ${\eta }_i$ represent year and firm fixed effects, respectively. We predict that bank mergers $MG{R}_{i,t}$ are associated with the reduction in the growth in total bank borrowings $TB{G}_{i,t}$, which is stated as ${\alpha }_1$ < 0.

3.2.2. Impacts of Bank Mergers on Working Capital Management

The second analysis investigates the relationship between the growth in bank credit and working capital management to calculate the total merger effect on working capital. Because of the doubly bounded nature of the dependent variables, we employ the panel data quasi-maximum likelihood (nonlinear) model from Papke and Wooldridge (2008) to clarify the impacts of bank mergers. We consider the following specification:

(3)$E\left(y_{i,t}|X_{i,t},TB{G}_{i,t},{\nu }_{i,t}\right)=\mathsf{\Phi }\left({\beta }_0+{\beta }_{1}TB{G}_{i,t}+{\beta }_2{X}_{i,t}+{\beta }_3\overline{X_i}+{\beta }_4{\nu }_{i,t}\right)$

where $y_{i,t}$ is the dependent variable, which is one of the working capital measures (CASH, RCV, INV, OCA, TC, and OCL) for firm i at time t. $X_{i,t}$ denotes a set of control variables. Following procedure 4.1 of Papke and Wooldridge (2008, p. 126), to estimate probit Equation (3), we first estimate Equation (2) to obtain the residuals for all (i, t) pairs ${\nu }_{i,t}$. Secondly, we use the pooled probit QMLE (quasi-maximum likelihood estimation) of $y_{i,t}$ on $X_{i,t}$, $TB{G}_{i,t}$, ${\nu }_{i,t}$ to estimate the coefficients ${\beta }_1$, ${\beta }_2$, ${\beta }_3$, and ${\beta }_4$ in Equation (3). $\overline{X_i}$ is the mean of control variables.

The coefficient ${\beta }_1$ measures the effect of growth in total borrowings $TB{G}_{i,t}$ on working capital management. A component of working capital is negatively associated with growth in total borrowings from banks if ${\beta }_1$ < 0. Considering Equation (3) together with Equation (2), the effect of a merger on a component of working capital is defined as ${\alpha }_1$ × ${\beta }_1$, which we call the total merger effect. In other words, mergers affect working capital through the change in total borrowings in our model specification.

In Equations (2) and (3), all right-hand side variables are lagged by one period to mitigate the endogeneity issue that the variables are simultaneously determined in equilibrium. All regressions include fixed effects to control for time-invariant firm characteristics. Additionally, we use year-fixed effects to control for changing circumstances over time.

3.2.3. PSM-DID Estimation Method

We use the PSM (propensity score matching)–DID (difference-in-difference) estimation method to identify the bank merger’s effect in the baseline estimation as a robustness check.

Firms’ working capital management is closely tied to bank borrowings because they tend to be determined simultaneously. Since bank borrowing is determined to equate its supply with its demand, bank mergers and borrowings on the supply side are considered to be determined simultaneously. We used the PSM-DID approach to overcome this potential endogeneity issue with a stricter method; however, the supply side information is not sufficiently properly defined when we use the MGR variable as a bank merger. We are able to use top lender’s financial ratios when we use the TMGR variable, but it is more ambiguous how we should define the financial ratios when we use MGR than when we use TMGR because the supply side impact of non-top lender lending may be offset by the top lender’s or other non-top lender’s lending behavior. In this sense, we are skeptical about the proper definition of financial ratios when we use the MGR variable and do not report those results.

In the following, we consider that the treatment is a bank merger, and the outcome is each component of the working capital. We use the abbreviation D = TMGR for treatment and $y$ for the outcome. $y^0$ represents the outcome that firm i would attain at time t in the absence of the treatment, while $y^1$ represents the outcome that firm i would attain at time t if exposed to the treatment. The average effect of the treated conditional on the covariates $X$ is

(4)$E\left(y_1^1-y_1^0|X,D=1\right)$

Under the assumption of parallel shift, we have

(5)$E\left(y_1^0-y_0^0|X,D=1\right)=E\left(y_1^0-y_0^0|X,D=0\right)$

The average effect of the treatment on the treated conditional on the covariates X is expressed as the DID form:

(6)$$\begin{gathered} E \left(y_1^1-y_1^0 | X,D=1\right)=\{E \left(y_1 | X,D=1\right)-E \left(y_1 | X,D=0\right)\} \\ -\{E \left(y_0 | X,D=1\right)-E \left(y_0 | X,D=0\right)\} \end{gathered}$$

Our PSM-DID method comprises the following two steps.

Step 1: We estimate the propensity score by probit estimation. The propensity score $\widehat{p}$ is estimated by the probit equation as

(7)$\mathrm{Pr}\left(D=1\right)=\mathsf{\Phi }\left(\mathsf{\delta }X\right)$

where $\mathsf{\Phi }$ is a cumulative normal distribution function, and $X$ is a vector of control variables. We define $\widehat{\mathsf{\delta }}$ as the estimated coefficient vector, and the estimated propensity score is defined as $\widehat{p}\equiv \mathsf{\Phi }\left(\widehat{\mathsf{\delta }}X\right)$. Using the propensity score, we match a firm with D = 1 (the treated group that experienced a bank merger) with a firm with D = 0 (the control group that did not experience a bank merger) one by one using nearest-neighbor propensity score matching (Rosenbaum and Rubin 1983).

Step 2: We consider the average treatment effects of a bank merger on the working capital components. Under the assumption of selection on observable restrictions (Abadie 2005):

(8)$E\left(y_1^0|X,D=1\right)=E\left(y_1^0|X,D=0\right)$

and the overlapping assumption, the consistent estimator of the average treatment effect becomes

(9)$E\left(y_1^1-y_1^0|X,D=1\right)={\left(\mathsf{\rho }N\right)}^{-1}\sum _{i=1}^N\frac{\left(D_{i1}-{\widehat{p}}_i\right)\mathsf{\Delta }{y}_i}{1-{\widehat{p}}_i}$

where N is the number of observations, $\mathsf{\rho }$ is the fraction of the treated, and $\mathsf{\Delta }{y}_i=y_{1,i}-y_{0,i}$ is the observed difference in $y$. The overlapping condition ensures the existence of potential matches for each treatment group among the control group. This assumption is sometimes called common support. As usual, we assume conditional mean independence $E\left(y_1^0|x,D\right)=E\left(y_1^0|x\right)$ and $E\left(y_1^1|x,D\right)=E\left(y_1^1|x\right)$. The conditional mean independence assumption is also called the unconfoundedness assumption.

We estimate the average treatment effects (ATTs) of a bank merger (TMGR) on CASH, RCV, INV, OCA, TC, and OCL. If Hypothesis 2 and 5 hold, we expect the ATTs of CASH and OCA to be positive. If Hypothesis 3, 4, 6, and 7 hold, we expect the ATTs of RCV, INV, TC, and OCL to be negative.

4. Empirical Evidence

4.1. Impacts of Bank Mergers on Growth in Bank Credit

The first analysis investigates the impacts of bank mergers on growth in bank credit. Table 3 reports the estimated results of Equation (2). From columns (1) to (3), the merger dummy MGR has a significantly negative effect on the growth rate of total bank borrowings TBG. This result is consistent with Hypothesis 1—the bank merger decreases the rate of change in total bank borrowings.

This result implies that bank mergers deteriorate the bank’s function of liquidity production because the bank merger prevents the bank from using the soft information of loan officers previously hired by either the merged or the merging bank. That is, bank mergers weaken banks’ information technology, which deteriorates the liquidity provision function. Rhoades (1998) indicates that bank mergers often result in significant cost cuts, but unexpected difficulties in integrating data processing systems and operations can hinder efficiency gains. Alternatively, the bank merger reduces the efficiency of bank management (Montgomery et al. 2014), and as a result, a less efficient bank cannot extend the loan amount after the merger.

From columns (4) to (6), we estimate the regression equation using the top lender merger dummy TMGR. Each coefficient of TMGR is significantly negative. Consolidations of top lenders are associated with a reduction in the growth of total borrowings. Montgomery and Takahashi (2018) indicate that a merger announcement by a firm’s main bank results in a decrease in credit supply from the merging bank. These results support Hypothesis 1.

As reported in Table 2, our sample of data on firms in Japan exhibits a significantly negative merger effect even in the sample mean. Our regression results are also consistent with Bonaccorsi di Patti and Gobbi (2007), who found that the bank borrowing is negatively associated with bank mergers.

Table 3 estimations include industry-fixed effects because the dependence on working capital varies across industries. Among the control variables, cash flow volatility lnCFV has a significant negative effect, while profit margin PM and R&D intensity RD have significant positive effects on the dependent variables.

4.2. Impacts of Bank Mergers on Working Capital Management through Bank Credit

The second analysis investigates the relationship between bank mergers and working capital management, considering the effect of bank mergers on the growth rate of bank borrowings.

Table 4 reports the estimated results of Equation (3).

In column (1), the growth in total borrowings TBG has a significantly negative effect on CASH, which means that firms with lower growth in bank credit hold increased amounts of cash. This result implies that when bank credit becomes less available after a merger, the borrower firms have hard access to external funding and, therefore, must hold more cash. The merger effect is 0.030, indicating that bank mergers increase corporate cash holdings through decreased growth in bank borrowings. This result supports Hypothesis 2, which suggests that borrowing firms find it more difficult to manage working capital and need to mitigate refinancing risk more carefully by holding more cash.

In column (2), the coefficient of TBG is significantly positive, indicating that firms with a lower amount of bank credit hold a lower amount of account receivables. The firm is willing to offer less credit to its customers when it borrows smaller amounts of bank loans. The merger effect is calculated as −0.015, which means that the bank mergers negatively affect account receivables through decreased growth in bank borrowings. That is, the consolidation of trading banks would affect the firm’s decisions regarding the provision of credit to its customers. This result supports Hypothesis 3, which suggests that borrowing firms find it more difficult to manage working capital and cannot extend more TC to their customers because their access to credit worsens.

In column (3), growth in total borrowings TBG has a significantly positive effect on INV, which means that a decrease in bank credit reduces the amount of investment in inventories. The total merger effect is −0.028, which means that bank mergers reduce inventories by decreasing bank borrowings. This result supports Hypothesis 4, which suggests that borrowing firms find it more difficult to manage working capital and need to hold more cash for refinancing risk consideration; therefore, they hold fewer inventories to substitute cash holdings.

In column (4), the coefficient of TBG is significantly negative. The total merger effect is calculated as 0.008, which means that bank mergers lead to an increase in OCA; this result supports Hypothesis 5.

In column (5), growth in total borrowings TBG has a significantly positive effect on trade credit used TC, i.e., accounts payable, which means that the relationship between bank credit and accounts payable is complementary. The total merger effect is calculated as −0.013, which supports Hypothesis 6—bank mergers reduce TC by decreasing bank borrowings. Since there is no substitution relationship between bank credit and TC, bank mergers make bank credit less available, and borrowing firms reduce the use of TC. Working capital management becomes increasingly difficult after a bank merger.

Column (6) shows that the growth in total borrowings TBG has a significantly positive effect on OCL, which means that extended bank credit increases other current liabilities. The merger effect is calculated as −0.037, indicating that bank mergers reduce OCL by decreasing bank borrowings; this result supports Hypothesis 7.

In panel B, we use the top lender merger dummy TMGR when estimating the first stage equation, and we use its residuals. The results are similar to panel A, except for the result of OCA, suggesting that if a firm experiences a merger of top lenders, it may react the same way in cash holdings, account receivables, accounts payable, and OCL.

The results for the control variables in panel A and panel B are similar. First, the tangibility TNG mostly has a significantly negative effect on the dependent variables. As the tangible assets serving as collateral increase, a firm can borrow more due to the mitigated agency costs of debt, which reduces working capital, at least through a decrease in precautionary demand for cash. Second, the cash flow volatility lnCFV significantly negatively affects CASH and RCV—greater cash flow volatility results in reduced cash holdings and account receivables. Third, firm size lnTA significantly negatively affects RCV and TC. As firm size increases, a firm’s creditworthiness rises, which results in the reduction in RCV and TC. Moreover, the profit margin is significantly positively associated with TC and negatively associated with OCA and OCL. Sales growth significantly negatively affects RCV and TC and positively affects OCA and OCL. R&D intensity is significantly positively associated with TC and negatively associated with OCA and OCL. Finally, GDP growth is significantly positively associated with CASH, RCV, and TC and negatively associated with OCA and OCL. Furthermore, GDP growth is significantly negatively associated with INV in panel B.

4.3. Bank Dependency and Merger Effect on Working Capital Management

Although previous evidence shows the main impacts of bank mergers on firms’ working capital financing and TBG, it is unclear how bank-dependent and non-bank-dependent firms respond differently to bank mergers. Chen and Kieschnick (2018) argued that bank-dependent firms tend to hold more in current assets and rely more on current liabilities in managing their working capital than less bank-dependent firms do and found that the working capital responses of bank-dependent firms to changes in the availability of bank credit differ from the responses of less bank-dependent firms in some areas. Therefore, we expect bank-dependent and non-bank-dependent firms to react differently to bank mergers.

We employ the panel data quasi-maximum likelihood (nonlinear) model from Papke and Wooldridge (2008) to clarify the impacts of bank mergers. We consider that small firms depend more on bank financing than large firms do. Bonaccorsi di Patti and Gobbi (2001) indicate that small businesses are more vulnerable to changes in local credit markets as they are more dependent on credit from local banks. Moreover, Mkhaiber and Werner (2021) show that small firms are largely dependent on bank credit for external funding. To identify bank-dependent and non-bank-dependent firms, we follow the study by Chen and Kieschnick (2018), which defines a firm as bank-dependent if it is in the bottom three deciles of firm size; otherwise, we consider the firm a non-bank-dependent firm. We employ this bank-dependent dummy variable, denoted by $BDe{p}_{i,t}$, in the regression analysis to compare the working capital management of bank-dependent and non-bank-dependent firms. Specifically, we relate the dummy variable BDep to the growth in bank borrowings TBG to show the different responses of bank-dependent firms to bank mergers. We consider the following specification:

(10)$E\left(y_{i,t}|X_{i,t},TB{G}_{i,t},{\mathsf{\nu }}_{i,t}\right)=\mathsf{\Phi }\left({\mathsf{\gamma }}_0+{\mathsf{\gamma }}_{1}TB{G}_{i,t}+{\mathsf{\gamma }}_{2}BDe{p}_{i,t}\times TB{G}_{i,t}+{\mathsf{\gamma }}_3{X}_{i,t}+{\mathsf{\gamma }}_4\overline{X_i}+{\mathsf{\gamma }}_5{\mathsf{\nu }}_{i,t}\right)$

where $y_{i,t}$ is one of the working capital measures (CASH, RCV, INV, OCA, TC, and OCL) for firm i at time t. $TB{G}_{i,t}$ is the growth in total bank borrowings. $BDe{p}_{i,t}$ is the dummy variable that takes one if a firm is in the bottom three deciles of firm size. $X_{i,t}$ denotes a set of control variables and $\overline{X_i}$ is the mean of control variables. ${\nu }_{i,t}$ is the residuals of all (i, t) pairs obtained by estimating Equation (2).

Table 5 reports the results of the regression analyses. The effects of bank credit on bank-dependent and non-bank-dependent firms are different. In column (1) of Panel A, the coefficient of $BDe{p}_{i,t}\times TB{G}_{i,t}$ is significantly negative, and the sum of the coefficients of $TB{G}_{i,t}$ and $BDe{p}_{i,t}\times TB{G}_{i,t}$ is −1.327 = (−1.287 − 0.040), implying that bank-dependent firms significantly decrease their cash holdings as non-bank-dependent firms do, and the negative impact of bank credit availability on corporate cash holdings of bank-dependent firms is greater than that of non-bank-dependent firms. The merger effect is positive (0.058 = (−1.327) × (−0.044)) for bank-dependent firms. The positive impact of bank mergers on the corporate cash holdings of bank-dependent firms is also greater than that of non-bank-dependent firms. The result suggests that bank-dependent firms find it more difficult to manage working capital and need to mitigate refinancing risk more carefully by holding a larger amount of cash than non-dependent counterparts after a merger.

In column (3) of Panel A, the coefficient of $BDe{p}_{i,t}\times TB{G}_{i,t}$ is significantly positive. The sum of the coefficients of $TB{G}_{i,t}$ and $BDe{p}_{i,t}\times TB{G}_{i,t}$ is 0.754 (= 0.724 + 0.030), which means that bank-dependent firms significantly increase their inventories after bank credit becomes more available as non-bank-dependent firms do, and the positive impact of bank credit availability on inventories of bank-dependent firms is greater than that of non-bank-dependent firms. The merger effect is calculated as negative (−0.033 = 0.754 × (−0.044)) for bank-dependent firms. The negative impact of bank mergers on inventories of bank-dependent firms is also greater than that of non-bank-dependent firms. The result suggests that bank-dependent firms find it more difficult to manage working capital and need to hold more cash for refinancing risk consideration; therefore, they hold fewer inventories to substitute cash holdings than their non-dependent counterparts after mergers.

However, the coefficients of $BDe{p}_{i,t}\times TB{G}_{i,t}$ in columns (2) and (4)–(6) are insignificant in both panels A and B, suggesting that bank-dependent firms do not significantly change their account receivables, other current assets, trade credit used, and other current liabilities, whereas non-bank-dependent firms significantly change these working capital measures.

4.4. PSM-DID Estimator

Table 6 reports the estimation result of the PSM-DID estimator. We use TMGR as the dependent variable. The average treatment effects (ATTs) of a bank merger on RCV, INV, and TC are significantly different from zero; in particular, the ATTs of a bank merger on RCV and TC are significant at 1%.

Below the standard errors, we report the estimation result of the probit estimation. The bank variables explain the probability that a firm experiences a bank merger. We employ the bank capital ratio denoted by BCR, the non-performing bank loans ratio denoted by BNPL, the bank deposit rate denoted by BDR, and the bank operating cost denoted by BOC as the bank variables. We also employ the disposal of non-performing loans (Disposal) and charge-off amount of loans (Charge-off) as instrumental variables to alleviate an endogeneity issue. A higher capital ratio, higher NPL ratio, and a lower deposit rate and operating cost tend to result in bank mergers.

5. Conclusions

This study raised a new question: what are the consequences of bank mergers on borrower working capital management? Since the availability of bank credit influences firms’ working capital management (Chen and Kieschnick 2018) and bank mergers influence credit availability (Bonaccorsi di Patti and Gobbi 2007), bank mergers affect bank borrowers’ working capital management.

Our study sheds new light on working capital management in the banking merger literature, providing the following empirical findings. (1) The bank merger resulted in the reduction in the rate of change in total bank borrowings. (2) The bank merger leads to an increase in corporate cash holdings and other current assets. (3) The bank borrower reduces working capital after the bank merger, mainly through the reduction in receivables, inventories, and the use of trade credit.

Our finding that bank mergers lead to a decrease in the growth rate on total bank borrowings underscores the direct influence of bank consolidations on credit availability, which suggests that bank mergers weaken the liquidity-providing function of the bank. A notable consequence of bank mergers is the resultant uptick in corporate cash holdings and other current assets. This behavior likely reflects firms’ strategies to mitigate refinancing risks in the face of reduced credit availability, aligning with the precautionary savings motive discussed in the broader financial literature. In addition, bank credit and trade credit are complementary, and bank mergers weaken the supply of bank credit, which also leads to a decrease in both account receivables and accounts payable. This finding illustrates the strategic reallocation of resources as firms navigate the tightened credit landscapes post-merger.

We still have limitations in our research. We do not consider whether firms are financially constrained or not or whether firms become more financially constrained after a merger, as the deterioration of the liquidity-providing function of the bank suggests. Furthermore, we do not consider the level of bank competitiveness, which plays a crucial role in influencing merger outcomes and bank lending dynamics. Such analysis would be a prospective direction for future research.

Author Contributions

Conceptualization, K.S.; methodology, B.N. and K.S.; software, B.N. and K.S.; validation, B.N. and K.S.; formal analysis, B.N. and K.S.; investigation, B.N. and K.S.; resources, B.N. and K.S.; data curation, B.N. and K.S.; writing—original draft preparation, K.S.; writing—review and editing, B.N.; visualization, B.N.; supervision, K.S.; project administration, K.S.; funding acquisition, B.N. and K.S. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The data that support the findings of this study are available from the authors upon reasonable request.

Acknowledgments

We are grateful for the helpful comments from participants at the JFA-PBFJ Conference (Online, Kyushu University) on 14 March 2022, the International Society for the Advancement of Financial Economics (ISAFE) Conference (Online, Ho Chi Minh University of Banking) on 6 December 2022, the Japan Finance Association (JFA) Annual Meeting (Tokyo Keizai University) on 10 September 2022, and the Nippon Finance Association (NFA) Annual Conference (Waseda University) on 20 May 2023. We would like to extend our thanks to Nobuyuki Teshima for his discussion at the JFA-PBFJ Conference, as well as to Hiromichi Iwaki for his discussion at the JFA meeting and to Cui Weihan for her valuable discussion at the NFA meeting, which led to the improvement of this manuscript. We also thank the anonymous referees for their helpful comments and suggestions that enhanced the quality of this manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Tables

Table 1

Frequency distributions of firms experiencing bank mergers in Japan.

Fiscal Year	Number of Firms
	All	MGR = 1	MGR = 0	TMGR = 1	TMGR = 0
2007	1752	10	1742	3	1749
2008	1752	151	1601	39	1713
2009	1752	27	1725	3	1749
2010	1752	0	1752	0	1752
2011	1752	16	1736	2	1750
2012	1752	362	1390	169	1583
2013	1752	0	1752	0	1752
2014	1752	0	1752	0	1752
2015	1752	0	1752	0	1752
2016	1752	0	1752	0	1752
2017	1752	10	1742	1	1751
2018	1752	6	1746	0	1752
2019	1752	9	1743	0	1752
Total	22,776	591	22,185	217	22,559

Notes: The table reports the frequency distributions of sample firms by fiscal year. Also, it provides frequency distributions of subsamples by bank merger (MGR) and by top lender merger (TMGR).

Table 2

Summary statistics and non-parametric hypothesis testing result.

Panel A: Summary Statistics
Variables	All			MGR = 1			MGR = 0			t-Test	p-Value
	Obs	Mean	S.D.	Obs	Mean	S.D.	Obs	Mean	S.D.
Working capital variables
CASH	22,776	17.64	12.81	583	12.56	7.89	22,193	17.77	12.89	9.72	0.000
RCV	22,776	20.36	13.75	583	21.28	13.76	22,193	20.34	13.74	−1.64	0.101
INV	22,776	11.47	10.23	583	13.56	11.43	22,193	11.41	10.20	−5.00	0.000
OCA	22,776	4.70	7.93	583	3.80	6.41	22,193	4.73	7.96	2.78	0.005
TC	22,776	13.10	10.82	583	14.33	10.12	22,193	13.07	10.83	−2.79	0.005
OCL	22,776	18.37	11.24	583	23.88	10.89	22,193	18.22	11.21	−12.03	0.000
Firm performance variables
TNG	22,776	0.30	0.19	583	0.35	0.18	22,193	0.30	0.19	−5.70	0.000
lnCFV	22,776	−1.06	1.46	583	−0.66	1.44	22,193	−1.07	1.46	−6.78	0.000
lnTA	22,776	10.92	1.57	583	11.02	1.38	22,193	10.92	1.58	−1.60	0.111
PM	22,776	0.05	0.19	583	0.04	0.05	22,193	0.05	0.19	1.36	0.173
SG	22,776	0.00	0.00	583	0.00	0.00	22,193	0.00	0.00	−1.31	0.191
RD	22,776	0.02	0.16	583	0.01	0.02	22,193	0.02	0.16	0.78	0.434
GDP	22,776	0.47	1.82	583	1.17	2.33	22,193	0.45	1.80	−9.40	0.000
Bank borrowing variables
TB	22,776	11.64	14.53	583	23.19	14.59	22,193	11.34	14.40	−19.61	0.000
TBG	22,776	−1.18	57.11	583	−4.55	39.32	22,193	−1.09	57.50	1.44	0.149
MGR	22,776	0.03	0.16
Notes: This table presents the means and standard deviations of the variables for sample firms and subsample firms that experienced bank mergers (MGR = 1) and those that did not (MGR = 0). It also reports the t statistics and p values for the null hypothesis that there is no difference in the sample mean between a subsample of bank mergers and no bank mergers. The definition of variables is in Appendix A. The unit of the mean is a percentage except for lnCFV and lnTA.
Panel B: Non-Parametric Hypothesis Testing.
	MGR = 1				MGR = 0
Variable	Obs		Median		Obs		Median		Z		p -Value
Working capital variables
CASH	583		11.192		22,193		14.819		9.457		0.000
RCV	583		20.079		22,193		19.142		−1.874		0.061
INV	583		11.645		22,193		9.715		−5.065		0.000
OCA	583		2.584		22,193		2.743		1.783		0.075
TC	583		12.515		22,193		10.746		−4.513		0.000
OCL	583		22.103		22,193		16.108		−13.924		0.000
Firm performance variables
TNG	583		0.325		22,193		0.279		−6.376		0.000
lnCFV	583		−0.564		22,193		−0.991		−7.661		0.000
lnTA	583		10.883		22,193		10.765		−1.979		0.048
PM	583		0.035		22,193		0.046		6.444		0.000
SG	583		0.000		22,193		0.000		−4.929		0.000
RD	583		0.004		22,193		0.003		0.021		0.983
GDP	583		2.700		22,193		0.600		−12.570		0.000
Bank borrowing variables
TB	583		21.430		22,193		5.034		−21.087		0.000
TBG	583		−4.210		22,193		0.000		5.181		0.000
Notes: The non-parametric hypothesis testing method used in this analysis is the Mann–Whitney U (Wilcoxon rank sum) test.

Table 3

The impact of bank mergers on growth in bank credit.

	(1)	(2)	(3)	(4)	(5)	(6)
	TBG	TBG	TBG	TBG	TBG	TBG
MGR	−0.043 *	−0.041 *	−0.044 *
	(0.022)	(0.022)	(0.022)
TMGR				−0.050 *	−0.049 *	−0.051 *
				(0.028)	(0.028)	(0.028)
Disposal	11.590 ***		14.956 ***	0.072 *		0.091
	(3.915)		(5.233)	(0.042)		(0.059)
Charge-off		16.287 **	−6.558		0.093	−0.039
		(6.606)	(8.529)		(0.064)	(0.081)
TNG	−0.012	−0.011	−0.012	−0.054	−0.052	−0.054
	(0.062)	(0.062)	(0.062)	(0.062)	(0.062)	(0.062)
lnCFV	−0.006 *	−0.006 *	−0.006 *	−0.006 *	−0.006 *	−0.006 *
	(0.003)	(0.003)	(0.003)	(0.003)	(0.003)	(0.003)
lnTA	−0.032	−0.032	−0.032	−0.023	−0.024	−0.023
	(0.020)	(0.020)	(0.020)	(0.019)	(0.019)	(0.019)
PM	0.292 ***	0.291 ***	0.292 ***	0.321 ***	0.321 ***	0.321 ***
	(0.068)	(0.068)	(0.068)	(0.071)	(0.071)	(0.071)
SG	−4.248	−4.242	−4.246	−3.202	−3.213	−3.206
	(5.004)	(5.002)	(5.005)	(4.944)	(4.944)	(4.944)
RD	0.323 ***	0.322 ***	0.323 ***	0.356 ***	0.356 ***	0.356 ***
	(0.075)	(0.075)	(0.075)	(0.078)	(0.078)	(0.078)
GDP	0.007	0.007	0.007
	(0.014)	(0.014)	(0.014)
Constant	0.657 **	0.655 **	0.655 **	0.512 *	0.521 *	0.512 *
	(0.315)	(0.316)	(0.315)	(0.306)	(0.307)	(0.306)
Industry fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Observations	22,776	22,776	22,776	22,776	22,776	22,776
Number of firms	1752	1752	1752	1752	1752	1752
R-squared	0.007	0.007	0.007	0.003	0.003	0.003

Notes: This table presents the regressions of the changes in bank borrowings on bank mergers dummy MGR and TMGR and the instruments (i.e., the disposal of non-performing loans and charge-off amount of loans by banks). The estimated models are panel data models with fixed effects. The definitions of the variables are provided in Appendix A. Standard errors are in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

Table 4

The impacts of bank mergers and changes in bank credit on working capital management.

Panel A
	(1)	(2)	(3)	(4)	(5)	(6)
	CASH	RCV	INV	OCA	TC	OCL
TBG	−0.682 ***	0.340 ***	0.645 ***	−0.182 **	0.285 ***	0.835 ***
	(0.057)	(0.056)	(0.064)	(0.078)	(0.058)	(0.047)
TNG	−1.352 ***	−0.400 ***	−0.285 **	−1.440 ***	−0.073	0.362 ***
	(0.103)	(0.092)	(0.115)	(0.288)	(0.091)	(0.099)
lnCFV	−0.013 ***	−0.007 ***	0.006 ***	0.011 ***	0.002	0.021 ***
	(0.002)	(0.002)	(0.002)	(0.003)	(0.002)	(0.002)
lnTA	0.026	−0.132 ***	0.088 ***	0.004	−0.099 ***	−0.004
	(0.020)	(0.018)	(0.023)	(0.049)	(0.019)	(0.024)
PM	0.071	0.114	−0.102	−0.277 ***	0.144 **	−0.493 ***
	(0.060)	(0.084)	(0.071)	(0.088)	(0.069)	(0.099)
SG	−0.975	−4.965 **	2.300	6.217 ***	−3.906 *	2.941 **
	(1.710)	(2.064)	(2.155)	(2.368)	(2.079)	(1.489)
RD	0.087	0.100	−0.106	−0.396 ***	0.771 ***	−0.544 ***
	(0.072)	(0.111)	(0.082)	(0.137)	(0.247)	(0.122)
GDP	0.005 ***	0.007 ***	−0.001	−0.003 ***	0.005 ***	−0.008 ***
	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)
m_TNG	0.187	−1.150 ***	−0.257 *	0.350	−1.098 ***	−0.283 **
	(0.118)	(0.112)	(0.135)	(0.296)	(0.115)	(0.110)
m_lnCFV	0.006	−0.042 *	0.128 ***	−0.000	−0.061 **	0.077 ***
	(0.019)	(0.022)	(0.025)	(0.037)	(0.030)	(0.023)
m_lnTA	−0.125 ***	0.140 ***	−0.079 ***	0.086 *	0.150 ***	0.013
	(0.021)	(0.020)	(0.025)	(0.052)	(0.021)	(0.025)
m_PM	1.370 ***	−1.730 ***	−0.445 **	0.183	−3.193 ***	−0.985 ***
	(0.238)	(0.265)	(0.220)	(0.656)	(0.344)	(0.202)
m_SG	1.629 ***	−1.212 *	−0.103	−0.243	0.125	−0.594
	(0.220)	(0.705)	(0.159)	(0.725)	(0.411)	(0.363)
m_RD	1.661 ***	−1.827 ***	−0.491 **	−0.369	−5.314 ***	−1.119 ***
	(0.210)	(0.222)	(0.233)	(0.713)	(0.727)	(0.226)
resTBG	0.679 ***	−0.351 ***	−0.636 ***	0.197 **	−0.311 ***	−0.803 ***
	(0.057)	(0.056)	(0.064)	(0.078)	(0.058)	(0.047)
Constant	0.330 ***	−0.403 ***	−0.985 ***	−2.353 ***	−1.169 ***	−0.819 ***
	(0.067)	(0.073)	(0.087)	(0.132)	(0.078)	(0.072)
Merger effect	0.030	−0.015	−0.028	0.008	−0.013	−0.037
Observations	22,776	22,776	22,776	22,776	22,776	22,776
Number of firms	1752	1752	1752	1752	1752	1752
Wald chi2	1975.20	865.22	324.46	173.41	575.93	875.57
Prob >chi2	0.000	0.000	0.000	0.000	0.000	0.000
Pseudo R-squared	0.044	0.036	0.011	0.032	0.036	0.009
Notes: This table presents the impacts of bank mergers and changes in bank credit on working capital management. These models were estimated using the quasi-likelihood (nonlinear) model of Papke and Wooldridge (2008) with fixed effects. A variable prefixed with m_ indicates the average value of this variable. resTBG is the residual of all (i, t) pairs obtained by estimating a simplified form of TBGi,t. Definitions of other variables are provided in Appendix A. Standard errors are in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.
Panel B
	(1)	(2)	(3)	(4)	(5)	(6)
	CASH	RCV	INV	OCA	TC	OCL
TBG	−0.516 ***	0.217 ***	0.526 ***	−0.084	0.215 ***	0.706 ***
	(0.056)	(0.052)	(0.062)	(0.077)	(0.053)	(0.047)
TNG	−1.347 ***	−0.403 ***	−0.290 **	−1.441 ***	−0.075	0.356 ***
	(0.103)	(0.092)	(0.116)	(0.289)	(0.092)	(0.099)
lnCFV	−0.012 ***	−0.007 ***	0.006 ***	0.011 ***	0.001	0.021 ***
	(0.002)	(0.002)	(0.002)	(0.003)	(0.002)	(0.002)
lnTA	0.025	−0.131 ***	0.089 ***	0.003	−0.099 ***	−0.003
	(0.020)	(0.018)	(0.023)	(0.049)	(0.019)	(0.024)
PM	0.040	0.136	−0.079	−0.297 ***	0.157 **	−0.467 ***
	(0.059)	(0.084)	(0.071)	(0.088)	(0.069)	(0.099)
SG	−1.187	−4.724 **	2.484	6.195 ***	−3.756 *	3.125 **
	(1.609)	(1.961)	(2.076)	(2.246)	(2.065)	(1.409)
RD	0.056	0.124	−0.082	−0.418 ***	0.786 ***	−0.517 ***
	(0.072)	(0.110)	(0.082)	(0.135)	(0.248)	(0.122)
GDP	0.006 ***	0.006 ***	−0.002 **	−0.003 **	0.004 ***	−0.008 ***
	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)
m_TNG	0.187	−1.150 ***	−0.256 *	0.354	−1.098 ***	−0.281 **
	(0.118)	(0.112)	(0.135)	(0.297)	(0.115)	(0.110)
m_lnCFV	0.007	−0.043 *	0.128 ***	0.000	−0.061 **	0.077 ***
	(0.019)	(0.022)	(0.025)	(0.037)	(0.030)	(0.023)
m_lnTA	−0.126 ***	0.140 ***	−0.078 ***	0.087 *	0.150 ***	0.013
	(0.021)	(0.020)	(0.025)	(0.052)	(0.021)	(0.025)
m_PM	1.396 ***	−1.748 ***	−0.464 **	0.200	−3.203 ***	−1.005 ***
	(0.239)	(0.265)	(0.220)	(0.654)	(0.344)	(0.203)
m_SG	1.642 ***	−1.222 *	−0.118	−0.237	0.118	−0.614 *
	(0.225)	(0.710)	(0.162)	(0.727)	(0.414)	(0.354)
m_RD	1.684 ***	−1.844 ***	−0.507 **	−0.353	−5.327 ***	−1.137 ***
	(0.210)	(0.222)	(0.232)	(0.712)	(0.730)	(0.229)
resTBG	0.511 ***	−0.226 ***	−0.515 ***	0.098	−0.240 ***	−0.672 ***
	(0.056)	(0.053)	(0.062)	(0.076)	(0.054)	(0.047)
Constant	0.353 ***	−0.421 ***	−1.002 ***	−2.339 ***	−1.179 ***	−0.837 ***
	(0.067)	(0.074)	(0.086)	(0.131)	(0.078)	(0.072)
Merger effect	0.026	−0.011	−0.027	0.004	−0.011	−0.036
Observations	22,776	22,776	22,776	22,776	22,776	22,776
Number of firms	1752	1752	1752	1752	1752	1752
Wald chi2	1915.41	831.44	292.7	166.73	562.21	784.87
Prob >chi2	0.000	0.000	0.000	0.000	0.000	0.000
Pseudo R-squared	0.044	0.036	0.010	0.032	0.036	0.008
Notes: This table presents the impacts of bank mergers and changes in bank credit on working capital management. These models were estimated using the quasi-likelihood (nonlinear) model of Papke and Wooldridge (2008) with fixed effects. The standard errors are robust standard errors, adjusted for firm-level clustering. A variable prefixed with m_ indicates the average value of this variable. resTBG is the residual of all (i, t) pairs obtained by estimating a simplified form of TBGi,t. Definitions of other variables are provided in Appendix A. Standard errors are in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

Table 5

Merger effect on working capital management: Bank dependency.

Panel A
	(1)	(2)	(3)	(4)	(5)	(6)
	CASH	RCV	INV	OCA	TC	OCL
TBG	−1.287 ***	0.256 ***	0.724 ***	0.541 ***	0.510 ***	0.880 ***
	(0.075)	(0.073)	(0.078)	(0.140)	(0.075)	(0.063)
BDep×TBG	−0.040 ***	0.020	0.030 **	0.007	0.023	0.013
	(0.013)	(0.014)	(0.014)	(0.019)	(0.016)	(0.018)
TNG	−1.299 ***	−0.391 ***	−0.296 **	−1.558 ***	−0.108	0.358 ***
	(0.114)	(0.090)	(0.116)	(0.315)	(0.088)	(0.101)
lnCFV	−0.010 ***	−0.004 **	0.001	0.011 ***	0.003	0.018 ***
	(0.002)	(0.002)	(0.002)	(0.003)	(0.003)	(0.002)
PM	0.194 ***	0.074	−0.088	−0.423 ***	0.064	−0.512 ***
	(0.064)	(0.084)	(0.073)	(0.101)	(0.073)	(0.099)
SG	1.508	−5.993 ***	2.484	4.534 ***	−6.536 ***	2.782 *
	(2.760)	(2.301)	(1.958)	(1.565)	(2.241)	(1.553)
RD	0.212 ***	0.061	−0.092	−0.541 ***	0.679 ***	−0.563 ***
	(0.077)	(0.110)	(0.084)	(0.144)	(0.259)	(0.121)
GDP	0.003 ***	0.005 ***	−0.000	0.001	0.005 ***	−0.007 ***
	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)
m_TNG	0.047	−1.152 ***	−0.246 *	0.549 *	−1.026 ***	−0.275 **
	(0.130)	(0.109)	(0.135)	(0.320)	(0.114)	(0.111)
m_lnCFV	0.060 ***	−0.046 **	0.119 ***	−0.046	−0.089 ***	0.070 ***
	(0.017)	(0.022)	(0.024)	(0.031)	(0.033)	(0.021)
m_PM	1.054 ***	−1.709 ***	−0.422 *	0.491	−3.053 ***	−0.958 ***
	(0.263)	(0.262)	(0.222)	(0.620)	(0.346)	(0.205)
m_SG	1.535 ***	−1.253 *	−0.064	−0.253	0.126	−0.590 *
	(0.313)	(0.761)	(0.153)	(0.645)	(0.482)	(0.351)
m_RD	1.374 ***	−1.801 ***	−0.474 **	−0.033	−5.109 ***	−1.090 ***
	(0.239)	(0.216)	(0.236)	(0.660)	(0.755)	(0.224)
resTBG	1.289 ***	−0.268 ***	−0.719 ***	−0.527 ***	−0.539 ***	−0.850 ***
	(0.075)	(0.074)	(0.078)	(0.139)	(0.076)	(0.061)
Constant	−0.649 ***	−0.321 ***	−0.894 ***	−1.425 ***	−0.663 ***	−0.734 ***
	(0.025)	(0.031)	(0.032)	(0.049)	(0.035)	(0.024)
Observations	22,776	22,776	22,776	22,776	22,776	22,776
Number of firms	1752	1752	1752	1752	1752	1752
Wald chi2	1458.120	753.320	277.010	162.940	497.760	783.470
Prob > chi2	0.000	0.000	0.000	0.000	0.000	0.000
Pseudo R-squared	0.035	0.035	0.010	0.022	0.033	0.009
Notes: This table presents the impact of the change in bank credit on working capital management. These models were estimated using the quasi-likelihood (nonlinear) model of Papke and Wooldridge (2008) with fixed effects. The standard errors are robust standard errors, adjusted for firm-level clustering. The dummy variable BDep represents bank dependence and is equal to one if the firm is in the bottom three deciles of firm assets. A variable prefixed with m_ indicates the average value of this variable. resTBG is the residual of all (i, t) pairs obtained by estimating a simplified form of TBGi,t. Definitions of other variables are provided in Appendix A. Standard errors are in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.
Panel B
	(1)	(2)	(3)	(4)	(5)	(6)
	CASH	RCV	INV	OCA	TC	OCL
TBG	−1.184 ***	0.137 *	0.629 ***	0.676 ***	0.466 ***	0.771 ***
	(0.077)	(0.073)	(0.080)	(0.148)	(0.075)	(0.066)
BDep×TBG	−0.045 ***	0.021	0.033 **	0.007	0.025	0.016
	(0.013)	(0.014)	(0.014)	(0.019)	(0.016)	(0.019)
TNG	−1.292 ***	−0.397 ***	−0.302 ***	−1.557 ***	−0.111	0.349 ***
	(0.115)	(0.090)	(0.116)	(0.315)	(0.088)	(0.101)
lnCFV	−0.010 ***	−0.004 **	0.001	0.011 ***	0.003	0.018 ***
	(0.002)	(0.002)	(0.002)	(0.003)	(0.003)	(0.002)
PM	0.171 ***	0.097	−0.068	−0.450 ***	0.074	−0.488 ***
	(0.064)	(0.083)	(0.073)	(0.101)	(0.073)	(0.099)
SG	1.464	−5.775 ***	2.592	4.430 ***	−6.413 ***	2.904 **
	(2.635)	(2.186)	(1.909)	(1.543)	(2.235)	(1.481)
RD	0.189 **	0.085	−0.072	−0.571 ***	0.690 ***	−0.538 ***
	(0.077)	(0.110)	(0.084)	(0.145)	(0.261)	(0.122)
GDP	0.003 ***	0.005 ***	−0.001	0.001	0.005 ***	−0.008 ***
	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)
m_TNG	0.042	−1.148 ***	−0.240 *	0.550 *	−1.023 ***	−0.269 **
	(0.131)	(0.110)	(0.136)	(0.321)	(0.114)	(0.111)
m_lnCFV	0.061 ***	−0.047 **	0.118 ***	−0.046	−0.089 ***	0.069 ***
	(0.017)	(0.022)	(0.024)	(0.031)	(0.033)	(0.021)
m_PM	1.070 ***	−1.723 ***	−0.434 *	0.507	−3.059 ***	−0.971 ***
	(0.264)	(0.263)	(0.221)	(0.620)	(0.347)	(0.206)
m_SG	1.554 ***	−1.261	−0.078	−0.255	0.117	−0.611 *
	(0.321)	(0.767)	(0.155)	(0.646)	(0.486)	(0.341)
m_RD	1.387 ***	−1.816 ***	−0.484 **	−0.018	−5.116 ***	−1.102 ***
	(0.240)	(0.216)	(0.235)	(0.657)	(0.758)	(0.226)
resTBG	1.185 ***	−0.148 **	−0.623 ***	−0.663 ***	−0.493 ***	−0.740 ***
	(0.077)	(0.074)	(0.081)	(0.148)	(0.076)	(0.064)
Constant	−0.646 ***	−0.323 ***	−0.896 ***	−1.423 ***	−0.664 ***	−0.735 ***
	(0.025)	(0.031)	(0.032)	(0.049)	(0.036)	(0.024)
Observations	22,776	22,776	22,776	22,776	22,776	22,776
Number of firms	1752	1752	1752	1752	1752	1752
Wald chi2	1399.940	760.320	252.160	156.650	481.210	713.310
Prob > chi2	0.000	0.000	0.000	0.000	0.000	0.000
Pseudo R-squared	0.035	0.035	0.010	0.022	0.033	0.008
Notes: This table presents the impact of the change in bank credit on working capital management. These models were estimated using the quasi-likelihood (nonlinear) model of Papke and Wooldridge (2008) with fixed effects. The standard errors are robust standard errors, adjusted for firm-level clustering. The dummy variable BDep represents bank dependence and is equal to one if the firm is in the bottom three deciles of firm assets. A variable prefixed with m_ indicates the average value of this variable. resTBG is the residual of all (i, t) pairs obtained by estimating a simplified form of TBGi,t. Definitions of other variables are provided in Appendix A. Standard errors are in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

Table 6

PSM-DID results.

	CASH		RCV		INV		OCA		TC		OCL
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
ATT	0.255	0.270	−0.768 ***	−0.813 ***	−0.335 *	−0.345 *	−0.035	−0.034	−0.538 **	−0.576 ***	−0.153	−0.150
t-value	0.909	0.965	−3.031	−3.229	−1.855	−1.922	−0.224	−0.219	0.219	0.218	−0.429	−0.423
Standard error	0.280	0.280	0.253	0.252	0.180	0.180	0.155	0.154	−2.460	−2.645	0.357	0.356
Dependent Variables	TMGR	TMGR	TMGR	TMGR	TMGR	TMGR	TMGR	TMGR	TMGR	TMGR	TMGR	TMGR
BCR	0.045 ***	0.040 ***	0.045 ***	0.040 ***	0.045 ***	0.040 ***	0.045 ***	0.040 ***	0.045 ***	0.040 ***	0.045 ***	0.040 ***
	(0.014)	(0.014)	(0.014)	(0.014)	(0.014)	(0.014)	(0.014)	(0.014)	(0.014)	(0.014)	(0.014)	(0.014)
BNPL	0.252 ***	0.217 **	0.252 ***	0.217 **	0.252 ***	0.217 **	0.252 ***	0.217 **	0.252 ***	0.217 **	0.252 ***	0.217 **
	(0.086)	(0.087)	(0.086)	(0.087)	(0.086)	(0.087)	(0.086)	(0.087)	(0.086)	(0.087)	(0.086)	(0.087)
BDR	−2.190 ***	−2.334 ***	−2.190 ***	−2.334 ***	−2.190 ***	−2.334 ***	−2.190 ***	−2.334 ***	−2.190 ***	−2.334 ***	−2.190 ***	−2.334 ***
	(0.305)	(0.306)	(0.305)	(0.306)	(0.305)	(0.306)	(0.305)	(0.306)	(0.305)	(0.306)	(0.305)	(0.306)
BOC	−1.991 ***	−2.130 ***	−1.991 ***	−2.130 ***	−1.991 ***	−2.130 ***	−1.991 ***	−2.130 ***	−1.991 ***	−2.130 ***	−1.991 ***	−2.130 ***
	(0.351)	(0.342)	(0.351)	(0.342)	(0.351)	(0.342)	(0.351)	(0.342)	(0.351)	(0.342)	(0.351)	(0.342)
Disposal	−2.548 ***		−2.548 ***		−2.548 ***		−2.548 ***		−2.548 ***		−2.548 ***
	(0.930)		(0.930)		(0.930)		(0.930)		(0.930)		(0.930)
Charge-off	2.310		2.310		2.310		2.310		2.310		2.310
	(1.578)		(1.578)		(1.578)		(1.578)		(1.578)		(1.578)
TNG	0.005	−0.000	0.005	−0.000	0.005	−0.000	0.005	−0.000	0.005	−0.000	0.005	−0.000
	(0.190)	(0.189)	(0.190)	(0.189)	(0.190)	(0.189)	(0.190)	(0.189)	(0.190)	(0.189)	(0.190)	(0.189)
lnCFV	0.012	0.017	0.012	0.017	0.012	0.017	0.012	0.017	0.012	0.017	0.012	0.017
	(0.024)	(0.024)	(0.024)	(0.024)	(0.024)	(0.024)	(0.024)	(0.024)	(0.024)	(0.024)	(0.024)	(0.024)
lnTA	0.038	0.038	0.038	0.038	0.038	0.038	0.038	0.038	0.038	0.038	0.038	0.038
	(0.025)	(0.025)	(0.025)	(0.025)	(0.025)	(0.025)	(0.025)	(0.025)	(0.025)	(0.025)	(0.025)	(0.025)
PM	−1.045 *	−1.033 *	−1.045 *	−1.033 *	−1.045 *	−1.033 *	−1.045 *	−1.033 *	−1.045 *	−1.033 *	−1.045 *	−1.033 *
	(0.564)	(0.567)	(0.564)	(0.567)	(0.564)	(0.567)	(0.564)	(0.567)	(0.564)	(0.567)	(0.564)	(0.567)
SG	5.803	5.525	5.803	5.525	5.803	5.525	5.803	5.525	5.803	5.525	5.803	5.525
	(7.086)	(7.173)	(7.086)	(7.173)	(7.086)	(7.173)	(7.086)	(7.173)	(7.086)	(7.173)	(7.086)	(7.173)
RD	0.308	0.341	0.308	0.341	0.308	0.341	0.308	0.341	0.308	0.341	0.308	0.341
	(1.457)	(1.448)	(1.457)	(1.448)	(1.457)	(1.448)	(1.457)	(1.448)	(1.457)	(1.448)	(1.457)	(1.448)
GDP	−0.183 ***	−0.212 ***	−0.183 ***	−0.212 ***	−0.183 ***	−0.212 ***	−0.183 ***	−0.212 ***	−0.183 ***	−0.212 ***	−0.183 ***	−0.212 ***
	(0.026)	(0.025)	(0.026)	(0.025)	(0.026)	(0.025)	(0.026)	(0.025)	(0.026)	(0.025)	(0.026)	(0.025)
Constant	−1.194 **	−1.033 **	−1.194 **	−1.033 **	−1.194 **	−1.033 **	−1.194 **	−1.033 **	−1.194 **	−1.033 **	−1.194 **	−1.033 **
	(0.466)	(0.464)	(0.466)	(0.464)	(0.466)	(0.464)	(0.466)	(0.464)	(0.466)	(0.464)	(0.466)	(0.464)
Observations	6282	6282	6282	6282	6282	6282	6282	6282	6282	6282	6282	6282
Pseudo R-squared	0.135	0.129	0.135	0.129	0.135	0.129	0.135	0.129	0.135	0.129	0.135	0.129

Notes: This table presents the results of the PSM-DID analysis of Equations (8) and (9). The average treatment effect (ATT), its standard errors, and t-values are reported. Additionally, this table reports the estimation results of the first-step probit estimation. The definitions of the variables are provided in Appendix A. Standard errors are in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

Appendix A. Definition of Variables

Variables	Definition
Working capital variables
CASH	The ratio of cash and securities to total assets.
RCV	The ratio of receivables to total assets.
INV	The ratio of inventories to total assets.
OCA	The ratio of other current assets to total assets.
TC	The ratio of trade credit used (accounts payable) to total assets.
OCL	The ratio of other current liabilities to total assets.
Firm performance variables
TNG	The ratio of tangible assets to total assets.
lnCFV	The natural log of cash flow volatilities estimated.
lnTA	The natural log of total assets.
PM	Profit margin, the ratio of operating income to sales.
SG	Sales growth, the change rate of sales from the previous year.
RD	R&D intensity, the ratio of R&D expenses to sales.
GDP	The growth rate of real GDP.
Bank variables
TB	The ratio of total borrowings from banks to total assets.
TBG	The percentage change of total borrowings from banks.
MGR	Dummy variable that takes one if a firm experiences a bank merger, and zero otherwise.
TMGR	Dummy variable that takes one if a firm experiences top lender merger, and zero otherwise.
BDep	Dummy variable that takes one if a firm is in the bottom three deciles of firm size.
Disposal	The ratio of disposal of non-performing loans to total assets.
Charge-off	The ratio of charge-off amount of loans to total assets.
BCR	The bank capital ratio.
BNPL	The bank nonperforming loans ratio.
BDR	The bank deposit rate.
BOC	The bank operating cost.

Appendix B. Summary of Japanese Bank Merger Activities, 2007–2019

Year	Before Merger		After Merge
2007	Momiji Bank, Ltd.	Momiji Holdings, Inc	Momiji Bank, Ltd.
2007	The Shokusan Bank, Ltd.	The Yamagata Shiawase Bank, Ltd.	The Kirayaka Bank, Ltd.
2008	The Kirayaka Bank, Ltd.	Kirayaka Holdings, Inc.	The Kirayaka Bank, Ltd.
2008	North Pacific Bank, Ltd.	The Sapporo Bank, Ltd.	North Pacific Bank, Ltd.
2009	Resona Bank, Limited	Resona Trust & Banking Co., Ltd.	Resona Bank, Limited
2010	The Kanto Tsukuba Bank, Limited.	The Ibaraki Bank, Ltd.	Tsukuba Bank, Ltd.
2010	Kansai Urban Banking Corporation	The Biwako Bank, Ltd.	Kansai Urban Banking Corporation
2010	The Bank of IKEDA, Ltd.	The Senshu Bank, Ltd.	The Senshu Ikeda Bank, Ltd.
2012	The Juroku Bank, Ltd.	The Gifu Bank, Ltd.	The Juroku Bank, Ltd.
2012	North Pacific Bank, Ltd.	Sapporo Hokuyo Holdings, Inc.	North Pacific Bank, Ltd.
2013	Mizuho Bank, Ltd.	Mizuho Corporate Bank, Limited	Mizuho Bank, Ltd.
2013	The Kiyo Bank, Ltd.	Kiyo Holdings, Inc.	The Kiyo Bank, Ltd.
2018	The Yachiyo Bank, Ltd., The Tokyo Tomin Bank, Limited, ShinGinko Tokyo, Limited		Kiraboshi Bank, Ltd.
2019	The Kinki Osaka Bank, Ltd.	Kansai Urban Banking Corporation	Kansai Mirai Bank, Limited

Notes: The merges in the table are all completed.

References

Abadie, Alberto. Semiparametric difference-in-differences estimators. Review of Economic Studies; 2005; 72, pp. 1-19. [DOI: https://dx.doi.org/10.1111/0034-6527.00321]

Aktas, Nihat; Croci, Ettore; Petmezas, Dimitris. Is working capital management value-enhancing? Evidence from firm performance and investments. Journal of Corporate Finance; 2015; 30, pp. 98-113. [DOI: https://dx.doi.org/10.1016/j.jcorpfin.2014.12.008]

Amel, Dean; Barnes, Colleen; Panetta, Fabio; Salleo, Carmelo. Consolidation and efficiency in the financial sector: A review of the international evidence. Journal of Banking & Finance; 2004; 28, pp. 2493-519.

Asimakopoulos, Ioannis; Athanasoglou, Panayiotis P. Revisiting the merger and acquisition performance of European banks. International Review of Financial Analysis; 2013; 29, pp. 237-49. [DOI: https://dx.doi.org/10.1016/j.irfa.2012.08.010]

Bates, Thomas W.; Kahle, Kathleen M.; Stulz, René M. Why do US firms hold so much more cash than they used to?. Journal of Finance; 2009; 64, pp. 1985-2021. [DOI: https://dx.doi.org/10.1111/j.1540-6261.2009.01492.x]

Berger, Allen N.; Kashyap, Anil K.; Scalise, Joseph M.; Gertler, Mark; Friedman, Benjamin M. The transformation of the U.S. banking industry: What a long, strange trip it’s been. Brookings Papers on Economic Activity; 1995; 2, pp. 55-218. [DOI: https://dx.doi.org/10.2307/2534612]

Berger, Allen N.; Rosen, Richard J.; Udell, Gregory F. Does market size structure affect competition? The case of small business lending. Journal of Banking & Finance; 2007; 31, pp. 11-33.

Bonaccorsi di Patti, Emilia; Gobbi, Giorgio. The changing structure of local credit markets: Are small businesses special?. Journal of Banking & Finance; 2001; 25, pp. 2209-37.

Bonaccorsi di Patti, Emilia; Gobbi, Giorgio. Winners or losers? The effects of banking consolidation on corporate borrowers. Journal of Finance; 2007; 62, pp. 669-95. [DOI: https://dx.doi.org/10.1111/j.1540-6261.2007.01220.x]

Brennan, Michael J.; Maksimovics, Vojislav; Zechner, Josef. Vendor financing. Journal of Finance; 1988; 43, pp. 1127-41. [DOI: https://dx.doi.org/10.1111/j.1540-6261.1988.tb03960.x]

Burkart, Mike; Ellingsen, Tore. In-kind finance: A theory of trade credit. American Economic Review; 2004; 94, pp. 569-90. [DOI: https://dx.doi.org/10.1257/0002828041464579]

Chen, Chongyang; Kieschnick, Robert. Bank credit and corporate working capital management. Journal of Corporate Finance; 2018; 48, pp. 579-96. [DOI: https://dx.doi.org/10.1016/j.jcorpfin.2017.12.013]

Chortareas, Georgios E.; Garza-García, Jesús G.; Girardone, Claudia. Competition, efficiency and interest rate margins in Latin American banking. International Review of Financial Analysis; 2012; 24, pp. 93-103. [DOI: https://dx.doi.org/10.1016/j.irfa.2012.08.006]

De Veirman, Emmanuel; Levin, Andrew. Cyclical changes in firm volatility. Journal of Money, Credit and Banking; 2018; 50, pp. 317-49. [DOI: https://dx.doi.org/10.1111/jmcb.12462]

Degryse, Hans; Masschelein, Nancy; Mitchell, Janet. Staying, dropping, or switching: The impacts of bank mergers on small firms. Review of Financial Studies; 2011; 24, pp. 1102-40. [DOI: https://dx.doi.org/10.1093/rfs/hhp126]

Dzhagityan, Eduard. Perspectives of Macrofinance Regulation of Banking M&A in the Context of the International Reform of Banking Regulation. SSRN Electronic Journal.; 2018; Available online: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3195823 (accessed on 7 May 2024).

Emery, Gary W. An optimal financial response to variable demand. Journal of Financial and Quantitative Analysis; 1987; 22, pp. 209-25. [DOI: https://dx.doi.org/10.2307/2330713]

Fazzari, Steven M.; Petersen, Bruce C. Working capital and fixed investment: New evidence on financing constraints. RAND Journal of Economics; 1993; 24, pp. 328-42. [DOI: https://dx.doi.org/10.2307/2555961]

Focarelli, Dario; Panetta, Fabio; Salleo, Carmelo. Why do banks merge?. Journal of Money, Credit and Banking; 2002; 34, pp. 1047-66. [DOI: https://dx.doi.org/10.1353/mcb.2002.0054]

Fraisse, Henri; Hombert, Johan; Lé, Mathias. The competitive effect of a bank megamerger on credit supply. Journal of Banking & Finance; 2018; 93, pp. 151-61.

Francis, Bill; Hasan, Iftekhar; Wang, Haizhi. Banking deregulation, consolidation, and corporate cash holdings: US evidence. Journal of Banking & Finance; 2014; 41, pp. 45-56.

Gachigo, Justin; Ondigo, Herick; Aduda, Josiah; Onsomu, Zipporah. The Effect of Mergers and Acquisitions Strategies on Financial Performance of Commercial Banks in Kenya. European Scientific Journal ESJ; 2022; 18, 110. [DOI: https://dx.doi.org/10.19044/esj.2022.v18n25p110]

Harford, Jarrad; Klasa, Sandy; Maxwell, William F. Refinancing Risk and Cash Holdings. Journal of Finance; 2014; 69, pp. 975-1012. [DOI: https://dx.doi.org/10.1111/jofi.12133]

Hill, Matthew D.; Kelly, G. Wayne; Highfield, Michael J. Net operating working capital behavior: A first look. Financial Management; 2010; 39, pp. 783-805. [DOI: https://dx.doi.org/10.1111/j.1755-053X.2010.01092.x]

Jayaratne, Jith; Wolken, John. How important are small banks to small business lending?: New evidence from a survey of small firms. Journal of Banking & Finance; 1999; 23, pp. 427-58.

Karceski, Jason; Ongena, Steven; Smith, David C. The Impact of bank consolidation on commercial borrower welfare. Journal of Finance; 2005; 60, pp. 2043-82. [DOI: https://dx.doi.org/10.1111/j.1540-6261.2005.00787.x]

Kieschnick, Robert; Rotenberg, Wendy. Working capital management, the credit crisis, and hedging strategies: Canadian evidence. Journal of International Financial Management & Accounting; 2016; 27, pp. 208-32.

Kieschnick, Robert; Laplante, Mark; Moussawi, Rabih. Working capital management and shareholders’ wealth. Review of Finance; 2013; 17, pp. 1827-52. [DOI: https://dx.doi.org/10.1093/rof/rfs043]

Kling, Gerhard; Paul, Salima Y.; Gonis, Eleimon. Cash holding, trade credit and access to short-term bank finance. International Review of Financial Analysis; 2014; 32, pp. 123-31. [DOI: https://dx.doi.org/10.1016/j.irfa.2014.01.013]

Lin, Tsung-Te; Chou, Jian-Hsin. Trade credit and bank loan: Evidence from Chinese firms. International Review of Economics & Finance; 2015; 36, pp. 17-29.

Long, Michael S.; Malitz, Ileen B.; Ravid, S. Abraham. Trade credit, quality guarantees, and product marketability. Financial Management; 1993; 22, pp. 117-27. [DOI: https://dx.doi.org/10.2307/3665582]

Love, Inessa; Preve, Lorenzo A.; Sarria-Allende, Virginia. Trade credit and bank credit: Evidence from recent financial crises. Journal of Financial Economics; 2007; 83, pp. 453-69. [DOI: https://dx.doi.org/10.1016/j.jfineco.2005.11.002]

Ly, Kim Cuong; Shimizu, Katsutoshi. Funding liquidity risk and internal markets in multi-bank holding companies: Diversification or internalization?. International Review of Financial Analysis; 2018; 57, pp. 77-89. [DOI: https://dx.doi.org/10.1016/j.irfa.2017.12.011]

Mateut, Simona. Reverse trade credit or default risk? Explaining the use of prepayments by firms. Journal of Corporate Finance; 2014; 29, pp. 303-26. [DOI: https://dx.doi.org/10.1016/j.jcorpfin.2014.09.009]

Mkhaiber, Achraf; Werner, Richard A. The relationship between bank size and the propensity to lend to small firms: New empirical evidence from a large sample. Journal of International Money and Finance; 2021; 110, 102281. [DOI: https://dx.doi.org/10.1016/j.jimonfin.2020.102281]

Molina, Carlos A.; Preve, Lorenzo A. Trade receivables policy of distressed firms and its effect on the costs of financial distress. Financial Management; 2009; 38, pp. 663-86. [DOI: https://dx.doi.org/10.1111/j.1755-053X.2009.01051.x]

Montgomery, Heather; Takahashi, Yuki. Effect of bank mergers on client firms: Evidence from the credit supply channel. Japanese Economic Review; 2018; 69, pp. 438-49. [DOI: https://dx.doi.org/10.1111/jere.12157]

Montgomery, Heather; Harimaya, Kozo; Takahashi, Yuki. Too big to succeed? Banking sector consolidation and efficiency. Journal of International Financial Markets, Institutions and Money; 2014; 32, pp. 86-106. [DOI: https://dx.doi.org/10.1016/j.intfin.2014.05.005]

Ogada, Agnes; Achoki, George; Njuguna, Amos. Effect of mergers and acquisitions strategies on financial performance of financial services sector. International Journal of Finance; 2017; 1, pp. 18-37. [DOI: https://dx.doi.org/10.47941/ijf.36]

Papke, Leslie E.; Wooldridge, Jeffrey M. Panel data methods for fractional response variables with an application to test pass rates. Journal of Econometrics; 2008; 145, pp. 121-33. [DOI: https://dx.doi.org/10.1016/j.jeconom.2008.05.009]

Pasiouras, Fotios; Tanna, Sailesh; Zopounidis, Constantin. The identification of acquisition targets in the EU banking industry: An application of multicriteria approaches. International Review of Financial Analysis; 2007; 16, pp. 262-81. [DOI: https://dx.doi.org/10.1016/j.irfa.2006.09.001]

Petersen, Mitchell A.; Rajan, Raghuram G. The benefits of lending relationships: Evidence from small business data. Journal of Finance; 1994; 49, pp. 3-37. [DOI: https://dx.doi.org/10.1111/j.1540-6261.1994.tb04418.x]

Petersen, Mitchell A.; Rajan, Raghuram G. Trade credit: Theories and evidence. Review of Financial Studies; 1997; 10, pp. 661-91. [DOI: https://dx.doi.org/10.1093/rfs/10.3.661]

Rajan, Raghuram G.; Zingales, Luigi. What do we know about capital structure? Some evidence from international data. Journal of Finance; 1995; 50, pp. 1421-60. [DOI: https://dx.doi.org/10.1111/j.1540-6261.1995.tb05184.x]

Rhoades, Stephen A. The efficiency effects of bank mergers: An overview of case studies of nine mergers. Journal of Banking and Finance; 1998; 22, pp. 273-91. [DOI: https://dx.doi.org/10.1016/S0378-4266(97)00053-8]

Rosenbaum, Paul R.; Rubin, Donald B. The central role of the propensity score in observational studies for causal effects. Biometrika; 1983; 70, pp. 41-55. [DOI: https://dx.doi.org/10.1093/biomet/70.1.41]

Sapienza, Paola. The effects of banking mergers on loan contracts. Journal of Finance; 2002; 57, pp. 329-67. [DOI: https://dx.doi.org/10.1111/1540-6261.00424]

Singh, Anupama; Singh, Aarti; Sharma, Haresh Kumar; Majumder, Saibal. Criteria selection of housing loan based on dominance-based rough set theory: An Indian case. Journal of Risk and Financial Management; 2023; 16, 309. [DOI: https://dx.doi.org/10.3390/jrfm16070309]

Stein, Jeremy C. Information production and capital allocation: Decentralized versus hierarchical firms. Journal of Finance; 2002; 57, pp. 1891-921. [DOI: https://dx.doi.org/10.1111/0022-1082.00483]

Summers, Barbara; Wilson, Nicholas. Trade credit management and the decision to use factoring: An empirical study. Journal of Business Finance & Accounting; 2000; 27, pp. 37-68.

Uesugi, Iichiro; Yamashiro, Guy. The relationship between trade credit and loans: Evidence from small businesses in Japan. International Journal of Business; 2008; 13, pp. 141-63.

Went, Peter. A quantitative analysis of qualitative arguments in a bank merger. International Review of Financial Analysis; 2003; 12, pp. 379-403. [DOI: https://dx.doi.org/10.1016/S1057-5219(03)00031-0]

Yang, Xiaolou. The role of trade credit in the recent subprime financial crisis. Journal of Economics and Business; 2011; 63, pp. 517-29. [DOI: https://dx.doi.org/10.1016/j.jeconbus.2011.05.001]