Abstract
We examine whether and how interest rate liberalization affects firm leverage in China. We find that interest rate liberalization exerts a negative effect on the leverage of firms. Specifically, firms experience a reduction in total leverage during the liberalization period, and firms’ short-term leverage declines more relative to long-term leverage. Mechanism analysis shows that firms with high information asymmetry enjoy more decline in leverage relative to firms with low information asymmetry, and further, liberalization policy enables the reduction in credit transaction costs, which indicates that the behavior of banks actively collecting corporate information is an important channel that interest rate liberalization impacts firm leverage. Finally, in additional tests, we find that the impact is more salient when firms are non-state-owned, and loss making. Compare to operating liabilities, firms experience more reduction in financial liability.
1. Introduction
The questions of how and the extent to which interest rate liberalization affects firms’ financial decisions have been of important issue in the economics and finance literature, particular in the transition period in China. This period was characterized by a rapid raise in debt scale and a series of structure contradictions. Regarding the reasons for the excessive debt of Chinese enterprises, Li et al. [1] show that a large number of nonmarket-oriented factors have led to the problem of high corporate leverage, financial institutions do not fully follow the principle of profit maximization when they allocate funds, soft budget constraints in some low-efficiency companies is one of the determinants of high debt ratio. Interest rate control is an important factor that causes soft budget constraints and misallocation of credit resources for some inefficient companies [2].
Interest rate liberalization lifts restrictions on the price of interest rates and the amount of bank credit, permitting banks to offer deposits and lending rates according to market supply and demand. Although the prior literature has tested the consequences of interest rate liberalization in economic growth [3], corporate investment [4], banking crises [5], and capital misallocations [6], it is unclear how does a firm’s leverage evolves during the liberalization period, especially in an emerging market. Does interest rate liberalization affect corporate leverage? If so, what are the mechanisms? Are all corporates affected equally? Replys to these concerns, hitherto ambiguous according to prior theoretical and empirical evidence, has important practical significance. First, they are of important implication to policy makers in clearing about the value of interest rate liberalization, so that they can take steps to make timely adjustments if interest rate liberalization does not achieve the desired effect. Second, our research complements the evidence that associates with the role of policy in explaining the behavior of firm financing. Combining with the literature related to corporate debt, it may form a complete picture of the evolution of firms’ capital structures.
We perform our investigation in a unique setting of China, where interest rate liberalization reform is characterized by the non-synchronicity of the deposit and lending markets, that is, it follows the reform idea of lending before deposit. The segmentation of deposit and lending markets will lead to the difference of generation mechanism between deposit rates and lending rates [7]. Relative to the process of deposit rate liberalization, taking the lead in liberalizing lending rates may result in a higher degree of lending rate liberalization, which may cause a decrease in the tolerance of banks for credit default risks of micro enterprises. There is an added excellent background for our study as Chinese firms pay more attention to guanxi (relationship). Relationship lending is an important factor in the operation of Chinese firms. One of the purposes of interest rate liberalization is to optimize the allocation of credit resources and reduce relational loans. Thus, this relationship-focusing culture provides a good opportunity for us to investigate the mechanism of interest rate liberalization. Furthermore, Chinese firms have a high debt ratio, and they generally have low long-term debt ratios, with more than 50 percent having been unable to obtain any long-term debt [8]. This characteristic enables us to focus on the maturity of debt in China. Overall, the unique institutional background in China makes it particularly important to better understand both the benefits and costs of liberalization, as well as provide valuable experiences for other emerging countries.
Drawing from standard theory, we examine the questions posted above and, especially, make a comparison between different types of firms and debts relying on microlevel data in China. Our study attends to the behavior of total leverage, long-term leverage, and short-term leverage, after controlling standard corporate characteristics that have been selected in the informed research as important determinants of corporate leverage [9, 10]. To be specific, we study how the total leverage, long-term leverage and short-term leverage of firms behave with the promotion of interest rate liberalization and its economic aftermath. We argue that firms, particularly risky firms, as banks become less tolerant to default risks, are more willing to reduce their using of debt. Given the generally short maturity structure of Chinese companies’ debt, we expect interest rate liberalization to have a greater impact on short-term debt. We find that, along with the liberalization of interest rates, corporates experience a significant decline in total leverage; relative to long-term leverage, the reduction in short-term leverage is particularly significant.
Next, we identify the economic mechanisms that transmit interest rate liberalization to firm leverage. This investigation is of practical value. As information asymmetry and credit transaction cost are important factors that affect enterprises’ access to bank loans, the reduction in information asymmetry and transaction cost of credit caused by interest rate liberalization that will influence firm leverage. We find interest rate liberalization plays a greater role in reducing firm leverage under the circumstances of serious information asymmetry, and it reduces the transaction costs of corporate loans, indicating that the liberalization policy encourages banks to actively collect corporate information. This behavior reduces the information asymmetry between banks and enterprises, and thus reduces corporate leverage.
Finally, we consider our additional tests from three aspects. First, we focus on the comparison between firms of different natures, such as state-owned enterprises (SOEs) and non-state-owned enterprises (Non-SOEs), for which we expect to experience different changes. Second, in view of the different sources of financial liabilities and operating liabilities, we differentiate the former from the latter. We also distinguish between enterprises that are profit-making and loss-generating, because the latter tend to have higher leverage. We find that along with interest rate liberalization, Non-SOEs, and loss-making firms experience a greater decreases in leverage relative to SOEs, and profit-making firms, including both long-term leverage and short-term leverage; and the deleveraging in financial liabilities is more pronounced than operating liabilities.
We make the following contributions. First, we add to the study on interest rate liberalization. The previous literature has examined implications of interest rate liberalization for corporate behavior, but does not involve corporate leverage in emerging markets. We leverage off a unique micro-data in China where there is no active financial market before the start of interest rate liberalization. We document the immediate aftermath of interest rate liberalization, and show that interest rate liberalization reduces firm leverage.
Second, we provide empirical evidence of two new mechanisms through which interest rate liberalization influences corporate leverage. On the one hand, interest rate liberalization decreases the information asymmetry between firms and banks, making banks identify corporate risks more easily, and charge a higher interest rates for loans to risky firms. In turn, this raise in interest rate reduces the financing needs of high-risk firms, and this decreased financing demand leads to a reduction in firm leverage. On the other hand, interest rate liberalization reduces the incentive for banks to establish close ties with firms, making firms’ credit transaction cost lessened, which caused a reduction in banks’ relationship with lending and then firm leverage.
It is relevant to differentiate our analysis from V. M. González and F. González [11]. They provide an international evidence of banking liberalization increasing firms’ debt. We differ from them in three ways. First, we use a microscopic perspective by focusing on the liberalization policy to affect firm leverage. Second, the basic issue of whether and how interest rate liberalization affects firm leverage is especially important in emerging countries such as China, where bank loans are the main source of financing for enterprises and short-term debts dominate corporate borrowing. Third, we adopt the running of Loan Prime Rate (LPR) centralized quotation and release mechanism in 2013 as an external shock to identify the causality of interest rate liberalization on corporate leverage. However, the issue in their research is less obvious due to the endogenous problems.
2. Institutional Background and Hypothesis Development
2.1. Interest Rate Liberalization in China
China’s financial market has been characterized with bank-oriented financial system. As a comparison, the financial system in developed economies such as U.S. is dominated by capital markets. Therefore, interest rate liberalization will have a greater impact on firms in China.
Interest rate liberalization is an important reform of the financial system in China. In 1996, The People’s Bank of China (PBOC), the central bank of China, relaxed control over interbank lending rates, which marked the beginning of interest rate liberalization. In 2004, the PBOC relaxed the limitation on the ceiling of lending rates that banks can offer. In 2013, the PBOC completely widened the range of interest rates of bank loan floor and acceptance rate, symbolizing that interest rate controls on loans were completely lifted. In 2015, the PBOC further abolished the setting of deposit rate ceiling, which marked the full liberalization of deposit rate control.
To guide banks in reasonable pricing and deepen the reform of interest rate liberalization, the PBOC has taken comprehensive measures such as continuing to publish benchmark interest rates for deposits and loans, and guiding the pricing mechanism of market interest rates, and improving macro-prudential management. Specifically, on September 24, 2013, a self-regulatory mechanism for market interest rate pricing was established. On October 25, 2013, the centralized quotation and release mechanism of LPR was ready to run. LPR is a reference interest rate that calculated by the weighted average method of lending rates the quotation banks offer for their optimal borrowers. The first batch of quotation banks are Industrial and Commercial Bank of China, Agricultural Bank of China, Bank of China, Construction Bank of China, Bank of Communications, China Citic Bank, Shanghai Pudong Development Bank, Industrial Bank, and China Merchants Bank.
As an important part of the self-discipline mechanism of market interest rate pricing, the centralized quotation and release mechanism of LPR is a key link in the process of the in-depth reform of interest rate liberalization. With the liberalization of controls over interest rates, the independent pricing of financial institutions needs a pricing fundamental. Therefore, we need to ensure the centralized quotation and release mechanism of LPR is in order.
2.2. Hypothesis
Theory suggests that interest rate liberalization may impact the corporate leverage through different channels. Interest rate liberalization brings about an exogenous shock to firms’ over-investment incentives, particularly for risky companies. During the period of control over interest rates, the lending rates and deposit rates offered by banks are tightly controlled by the Chinese government, and may thus lead to real interest rates that are lower than the market equilibrium rate. At that time, financial markets were in short supply and many companies overinvested. In contrast, during the liberalization of interest rates, as real interest rates rise and the expected returns reduce, borrowers become reluctant to engage in overinvestment. The weakening of enterprises’ willingness to invest will inevitably cause the reduction of their financing needs and scale of debt [12].
From the perspective of banks, given the rise in management risk and competition, the loan marketing for them becomes more difficult after the range of lending rates widens [13]. On the one hand, the companies with good economic benefits, strong profitability, and low risk will choose their lenders according to the lending rates given by the banks. The bargaining power of low-risk companies in bank credit enhanced significantly, and banks often need to cut in interest rates to contest for low-risk firms, this fierce competition may eventually lead to bank failure [14]. On the other hand, during the period of control over interest rates, lending rates and the amount of credit are limited to a narrow range. Some inefficient and risky enterprises have easier access to bank loans with lower interest rates, due to the existence of government endorsements or guarantees. During the liberalization of lending rates, as risk increase and profitability become more uncertain, banks are reluctant to lend to risky firms due to the need for risk control. Especially, banks that are state-owned may also be more unwilling to overextend their credit business by providing funds to risky companies, because they are likely to subject to tight controls and supervision.
In a competitive market, it is difficult for banks to maintain a long-term lending relationship with enterprises, which may weaken the motivation of banks to obtain firms’ soft information by building close ties with enterprises [15]. The impact of financial liberalization on the access to loans depends on the correlation of information asymmetry. In a perfect financial market, greater market competition would lead to the increase of fund availability and the decline of loan costs. However, in the market with information asymmetry, as the competition in the financial market intensifies, the earnings of banks for keeping close ties to firms will reduce [16]. Hence, the availability of loans for companies, particularly for risky companies, may decline. We propose our hypothesis below: H1: The interest rate liberalization policy can reduce corporate leverage. The reform of interest rate liberalization in China is featured as a segmentation of deposit and loan markets, the reform framework is lending rate first and deposit rate second. This characteristic leads to a consequence that relative to the process of deposit rate liberalization, the lending rate liberalization will be even higher. Hence, the impact of deposit rates on loan pricing may be weakened, and thus may lead to a decline in the tolerance of banks to the credit default risk of firms. In the process of interest rate liberalization, the interest rate spreads of banks become narrow, which makes them more sensitive and cautious to default probabilities. As the uncontrollability in the liability field increases, banks are likely to pay more attention to risk control and the optimization of credit resource allocation structure. The maturity composition of loans is an important tool used for bank governance, the liberalization policy increases the banks’ incentive to reduce short-term lending over risky firms. As the maturity of long-term loans is longer, it is difficult for banks to withdraw them in the current period, except in the case of substantial breach of debt contracts. Different from long-term loans, the maturity of short-term loans is shorter, banks can suspend or rollover the loans according to firms’ profitability and debt performance. In the research of Fan et al. [17], generally, relative to developed economies, the maturity of debts of companies in transition economies are shorter, the short-term debts of listed companies in the United States are about 20%, while the short-term debts of listed companies in China are as high as 90%. They also show that it’s because banks in transition economies prefer to offer short-term lending, which are conductive to control risks more effectively. By reducing the problem of asset substitution, short-term debt can lessen the overinvestment of enterprises [18], and this effect will be strengthened under the interest rate liberalization. Banks may choose to suspend matured debt for firms with high default risk, thus will reduce the short-term leverage of enterprises. The liberalization policy leads to serious competition among banks. In the fierce market competition, banks are more motivated to collect firms’ information of their own accord. The more information the banks captures about the firms, the more accurate they can make the decision of whether to rollover the loan. In the process of interest rate liberalization, companies with a higher proportion of short-term debt and a higher risk of default are faced with a higher probability of loan suspension. Hence, the short-term debt of these firms reduces even more. From the perspective of firms, given the rise of short-term debt, the liquidity risks they face are higher. The more short-term debt they have, the terms of loans will stricter when there is a demand for rollover. Therefore, firms may choose to reduce investment rather than rollover short-term loans to avoid greater constraints and closer supervision. It can be seen that interest rate liberalization is helpful to adjust the maturity structure of enterprise debt and reduce its short-term debt ratio. We propose our hypothesis below: H2: Relative to long-term leverage, the interest rate liberalization policy has a more significantly impact on the reduction of the short-term leverage of enterprises.
3. Sample, Variables, Model and Identification
3.1. Sample Formation
Our firm-level data covers the period 2007–2018 and comes from the China Stock Market and Accounting Research (CSMAR) database. We begin with a sample of 23546 observations for 3608 A-share listed companies in China. We adopt the following filters on the initial sample. First, we exclude 723 observations from the financial industry. Second, we drop 691 observations that are specially treated. Third, we exclude 4176 observations with missing values. Our final sample contains 17, 956 observations representing 2,976 firms.
3.2. Variable Construction
We leverage two alternative measures of interest rate liberalization. The first is a composite index of deposit and lending rates, money market rate, bond yields, and financial product yield, which is calculated by Wang and Peng [19]. Our second measure is also a composite index of money market rate, bond yields, foreign currency interest rate, deposit, and lending rates, which is constructed by Liu and He [20]. We extend their data sample interval to 2018 on the basis of their research, and label the two measures Irlib1 and Irlib2, respectively.
Our three dependent variables are the ratio of total liabilities to total assets (Lev), the ratio of short-term liabilities to total assets (Slev), and the ratio of long-term liabilities to total assets (Llev). The three variables capture the extent to which companies’ assets are financed by total debt, short-term debt, and long-term debt separately.
We include control variables at the firm level, regional level, and macroeconomic level. Company size (Size) is the natural logarithm of total assets; profitability (ROA) is pretax profits divided by total assets; market-to-book ratio (MB) is the ratio of market capitalization to the book value of equity; Non-Debt Tax Shields (NDTS) is the natural logarithm of the sum of depreciation and amortization; fixed assets (PPE) is the ratio of fixed assets to total assets; cash flow (Cflow) is operating cash flow divided by total assets; capital expenditure (Capex) is the total amount of cash paid to construct property, plant, and equipment, intangible assets and other long-term assets divided by total assets; sales growth rate (Grow) is the increase rate of operating revenue; macro-economic climate index (Meci) is the natural logarithm of the average monthly leading index. We utilize fixed effects regression to test our hypothesis. This estimation can help us determine whether the results are affected by unobservable factors. The t-statistics in our results are based on robust standard errors [21].
3.3. Model Building
To investigate the impact of interest rate liberalization on corporate leverage, we use the following regression specification:Where the dependent variable Yit is, alternatively, the total leverage ratio (Lev), the short-term leverage ratio (Slev), or the long-term leverage ratio (Llev) for corporate i during year t. The independent variables, Irlib, including Irlib1 and Irlib2, which are the process of interest rate liberalization measured by two methods. According to our hypothesis, we expect that the coefficient of Irlib1 and Irlib2 should be significantly negative. Controlsit is our set of firm level, regional level, and macroeconomic level control variables aforementioned in the above section.
3.4. Identification Strategy
The centralized quotation and release mechanism of LPR was starting to run formally On October 25, 2013. This policy has the merit of creating treated and control groups based on whether the loans of firms come from the first batch of quotation banks. Firms, whose loans were all came from the first batch of quotation banks, were affected to a greater extent. We conduct the following difference-in-differences regression specification:Where Yit is the same as above, it is alternatively, the total leverage ratio (Lev), the short-term leverage ratio (Slev), or the long-term leverage ratio (Llev) for firm i during year t. Variable LPR is an indicator for firms’ borrowing sources. It equals 1 if the loans of firms are all from the first batch of quotation banks, and 0 otherwise. Post is a dummy variable for the post-LPR running period. It equals 1 if the year is any of 2013 onwards, and 0 otherwise. The μ and ρ are individual fixed effect and time fixed effect, respectively. As banks with pricing power can determine their borrowers, lending rates, loan limit, and maturity according to firms’ financial situation, default risk, and development prospect after the running of centralized quotation and release mechanism of LPR. Thus firms, particularly risky firms, whose borrowings are all from the first batch of quotation banks, will fall into higher borrowing rates and tougher terms of credit. We expect the coefficient of interaction term LPR × Post is significantly negative.
4. Empirical Analyses
4.1. Descriptive Statistics
Table 1 presents descriptive statistics for our key variables. The average firm has a total leverage ratio, short-term leverage ratio, and long-term leverage ratio of 0.464, 0.377, and 0.088, respectively. However, there is great variation in the three variables, since the standard deviations of them are 0.218, 0.194, and 0.108, respectively. Moreover, companies have more short-term debt relative to the long-term debt, and many companies do not have any long-term debt. The average process of interest rate liberalization is 0.733 (Irlib1) or 0.747(Irlib2). Market-to-book ratio averages at 2.431. An average firm has return-on-assets (ROA) of 0.057, Non-Debt Tax Shields (NDTS) of 0.025, and capital expenditure (Capex) of 0.057.
Table 2 presents the Pearson correlations among key variables. Not surprisingly, the three measures of leverage ratio (Lev, Slev, and Llev) and the two measures of interest rate liberalization (Irlib1 and Irlib2) are highly correlated. The negative correlations between interest rate liberalization measures and Lev (Slev) indicate that firms tend to experience a reduction in total leverage (short-term leverage) during the period of interest rate liberalization. Leverage ratios are also negatively correlated with profitability (ROA), and cash flow (Cflow), and positively correlated with firm size (Size), Non-Debt Tax Shields (NTDS), and macro-economic climate index (Meci). Moreover, interest rate liberalization is also negatively correlated with profitability (ROA), Non-Debt Tax Shields (NTDS), fixed assets (PPE), cash flow (Cflow), capital expenditure (Capex), sales growth rate (Grow), macro-economic climate index (Meci), and positively correlated with firm size (Size). Finally, correlations among our control variables indicate that there is no severe concern on multicollinearity.
4.2. Univariate Analysis
Table 3 shows results from estimating the differences in leverage ratio between the treated group and control group around the running of LPR centralized quotation and release mechanism. The leverage ratios (Lev, Slev, and Llev) of both the treated group and control group are reduced after the running of LPR centralized quotation and release mechanism. It means that inaccurate results may be obtained if the temporal variation difference is not eliminated when analyzing the impact of interest rate liberalization. In column (7), where the influence of temporal factors is avoided, we find that the difference-in-differences of total leverage ratio and the short-term leverage ratio are significantly negative, while the difference-in-differences of long-term leverage ratio is not pronounced. It preliminarily shows that interest rate liberalization can reduce the total leverage and short-term leverage of enterprises, but has no pronounced impact on long-term leverage.
4.3. Baseline Results on Firm Leverage
Our hypothesis indicates that as the advance of interest rate liberalization, companies are more likely to experience a decline in leverage ratios, particularly short-term leverage ratio. Panel A in Table 4 presents results from estimating the corporate impact on the leverage ratios (Lev, Slev, and Llev) of the interest rate liberalization. We see the three leverage ratios (Lev, Slev, Llev) all decline on average during the period of interest rate liberalization. The effect of interest rate liberalization on leverage ratios is economically large: the total leverage declines by 43.97 or 21.97 percent points, short-term leverage falls by 35.49 or 17.73 percent points, and long-term leverage reduces by 7.06 or 3.53 percent points throughout the sample period of 2007–2018.
Coefficients on control variables in Table 4 look as expected. Larger firms and firms with more nondebt tax shields have higher leverage ratios. An increase in fixed assets tends to be positive with leverage ratio, while an improvement in profitability is likely to be negative with leverage ratio. Growth companies, maybe due to their financing convenience, have high leverage ratios. Firms with more capital expenditures exhibit lower total leverage ratio and short-term leverage ratio, and higher short-term leverage ratio.
To compare the impact of interest rate liberalization on short-term leverage and long-term leverage, we follow Li et al. [22] and Cohen et al. [23], and use the statistical method offered by them to draw a comparison between a certain explanatory variable’s affection on two different explained variables within the same sample. First, we calculate the estimated value of Llev, and label it as Llehat. Second, we generate a new dependent variable (Dlev) by subtracting Llehat from Slev (i.e., Slev-Llehat). Finally, we employ our regression by taking Dlev as a dependent variable, and including the original set of independent variables (Irlib1 and Irlib2) and control variables. The coefficients of Irlib1 or Irlib2 and its statistical significance, represent the extent and significance of difference in the effect of interest rate liberalization on Slev and Llev.
Panel B in Table 4 shows the results of this difference analysis. We find the effect of Irlib1 on Slev and Llev are significantly differentiable (p ≤ 0.001), and similarly, Irlib2 also has a stronger effect on Slev, relative to Llev (p ≤ 0.001). Such results confirm that interest rate liberalization plays an important role in reducing the leverage ratios of firms, and there is a greater reduction in short-term leverage of firms relative to long-term leverage. That means a large portion of the reduction in total leverage is driven by the decline in the use of short-term debt.
4.4. Robustness Checks
4.4.1. Difference-in-Differences Estimation
Table 5, panel A reports difference-in-differences regression results of interest rate liberalization on firm leverage ratios. In column 1, we see negative and pronounced coefficients on LPR × Post (−0.0395, t = −4.55 for Lev), suggesting a greater decline in the total leverage ratio of firms that obtain all their loans from the first batch of quotation banks post-reform, relative to firms that do not obtain all their loans from first batch of quotation banks. Examining the coefficients in column 2 and column 3, we find that the centralized quotation and release mechanism of LPR has a different impact on debt maturity. Specifically, the coefficient on LPR × Post is negative and significant in column 2 (−0.0339, t = −4.12 for Slev), it’s also negative but not pronounced in column 3 (−0.0038, t = −0.93 for Llev). Such results indicate that firms with all their loans from the first batch of quotation banks experience a decline in total leverage, which is mainly driven by the decline of short-term leverage.
As the characteristics of the treated group and control group can be radically different, the leverage ratios may have divergent patterns around the liberalization reform. To alleviate this focus, we try to pair treated group with similar control group and carry out matching analysis by following procedures. First, we use firm characteristics to predict treatment probability. We select variables such as firm size (Size), profitability (ROA), ratio of fixed assets (PPE), capital expenditure (Capex), operational cash flow (Cflow), sales increase (Grow), etc, to estimate treatment prediction utilizing a probit model. Next, we attempt to build the matched samples based on the results of probit model. We pair each treated firm with a control firm based on the closest propensity score between them. We claim nonreplacement when choosing control firms and perform 1 : 1 matching via the nearest neighbor matching method. Table 6 shows that none of the control variables in the probit model are significant after matching.
Finally, we perform the difference-in-differences regression again by using our matched sample over the effect of liberalization policy on firm leverage ratios. Table 5, panel B also presents negative and pronounced coefficients on LPR × Post in columns 4 and 5 (−0.0345, t = −2.72 for Lev; −0.0257, t = −2.15 for Slev), and nonsignificant coefficient on LPR × Post in column 6 (0.0063, t = 1.04 for Llev).
We use the same method as shown in panel B of Table 4 to compare the effect of LPR × Post on Slev and Llev. The p-values reported in Table 5 suggest that the centralized quotation and release mechanism of LPR has a larger impact on short-term leverage ratio, relative to long-term leverage ratio. These results reconfirm that, relative to control firms, the centralized quotation and release mechanism of LPR reduces the total leverage for treated firms, and the magnitude of the reduction in short-term leverage is larger than long-term leverage.
4.4.2. Revalidation Based on Dynamic Panel Data
In view of the dynamic of financing behavior of firms, we take the 2period lag leverage ratio as control variable added into (1), and use System GMM regression to reestimate our main model. Panel A in Table 7 reports the results. In column 1 and column 2, we also see negative and pronounced coefficients on Irlib1 (−0.0792, t = −2.24) and Irlib2 (−0.0490, t = −2.23). Similarly, as shown in columns 3 and 5, the use of short-term debt also reduces significantly in firms; while the employ of long-term debt makes no change in columns 4 and 6.
Panel B of Table 7 shows the comparison results for the impact of interest rate liberalization on Slev and Llev. We find that both p values are less than 0.01, which suggests that the coefficient of Irlib1 (or Irlib2) on Slev is significantly smaller than that of Irlib1 (or Irlib2) on Llev. As the coefficient of Irlib1 (or Irlib2) is negative, we can draw the same conclusion as above that interest rate liberalization has a stronger affection on short-term leverage than on long-term leverage.
To summarize, the findings that interest rate liberalization can reduce firms’ total leverage and has a greater impact on short-term leverage relative to long-term leverage are confirmed again.
4.4.3. Alternative Measures of Variables
The documented negative relation between interest rate liberalization and firm leverage may be driven by the measures of variables. To address this concern, we run several variants to confirm the robustness of our results as above.
First, we utilize the following alternative measures of leverage ratio: bank borrowing to assets ratio (Loan), short-term borrowing to assets ratio (Sloan), and long-term borrowing to assets ratio (Lloan). These three measures express total leverage, short-term leverage, and long-term leverage as before, respectively.
The results are presented in Table 8 utilizing the same fixed effects regression as before. In columns 1 and 2 of Panel A, both measures of interest rate liberalization are significant in promoting the reduction of firms’ borrowing, which suggests that banks view a focus on offering firms with diverse lending rates during the period of interest rate liberalization. Turning to the maturity, we see the coefficients of Irlib1 (or Irlib2) remain negative and significant in columns 3 and 4 (columns 5 and 6) of panel A. The results of difference checks between the groups of Sloan and Lloan are reported in Panel B of Table 8. Both differences indicate that the short-term borrowing of firms exhibits more steeply falling during the liberalization period, relative to long-term borrowing.
The results reverify the findings from Table 4 that the total leverage of firms drops significantly, and short-term leverage of firms drop more than long-term leverage during interest rate liberalization.
Second, we construct another proxy for interest rate liberalization. As interest rate liberalization shows an obvious turning point in 2013, we transform the variable of interest rate liberalization into an indicator libdum that equals 1 after the running of LPR centralized quotation and release mechanism, and 0 otherwise. We reemploy the main estimations and Table 9 reports the results. We find that the coefficients on libdum are all negative and significant, and the difference of the coefficients on libdum between Slev and Llev is also pronounced. Thus, similarly, the results also confirm the findings in Table 4 that firms experience a reduction in total leverage, and the decline in firms’ short-term leverage is larger than that in long-term leverage after interest rate liberalization reform.
In summary, changes in the proxy for dependent variables and independent variables do not change the results presented on how interest rate liberalization affects firms’ leverage and maturities.
4.5. Mechanism Analysis
As above, we argue that interest rate liberalization has intensified competition across banks and stimulates banks to collect information of firms passionately, which can help alleviate the information asymmetry between banks and firms. As banks know more about firms, banks can apply differentiated lending rates to companies with different risks, thus reducing firms’ leverage ratios. To provide relevant evidences for the impact mechanism of interest rate liberalization on affecting corporate leverage ratios, we perform our mechanism analysis via two stages. First, we examine whether interest rate liberalization reduces information asymmetry and thus impacts corporates’ leverage. Next, we investigate whether the easing of information asymmetry is caused by banks’ active information mining behavior. For the former, we test whether the reduction effect of interest rate liberalization on corporates’ leverage is more pronounced when the degree of information asymmetry is higher. As for the latter, since it is difficult to directly verify the active behavior of banks, we provide evidence for it by examining whether interest rate liberalization reduces the transaction costs of enterprises in obtaining loans.
4.5.1. Information Asymmetry
By distinguishing different degrees of information asymmetry, we examine whether interest rate liberalization has a greater effect on the reduction of corporate leverage in the case of more serious information asymmetry, to provide evidence for the logic that interest rate liberalization reduces corporate leverage by alleviating information asymmetry. Following the prior studies [24, 25], we employ stock price synchronicity (SYN) to proxy for information asymmetry. To measure stock price synchronicity, we consider the following regression for each calendar year to decompose total return variations into common factors and firm-specific factors:Where Ret is the weekly return on A-shares traded on either the Shanghai or Shenzhen exchange for firm i and week t. MR is the value weighted A-share market return, and INR is the value weighted A-share industry return. We estimate the coefficient of determination () from the regression of the above equation for firm i. SYN is the logistic transformation of , which is as follows:
A higher stock price synchronicity, therefore, indicates a higher level of information asymmetry. According to the stock price synchronicity, we divide firms into three groups and choose the largest synchronicity group and the least synchronicity group to test the impact of interest rate liberalization on corporate leverage and maturity, respectively. This approach permits us to evaluate whether the impact of interest rate liberalization on corporate leverage is more pronounced at very high or low levels of information asymmetry.
The results are shown in Table 10. The dependent variables in columns 1 and 2 are Lev, while in columns 3 and 4, they are Slev, and in columns 3 and 4 are Llev. In columns 1and 2 of panel A, we see the coefficients on Irlib1 are negative for both high asymmetric firms and low asymmetric firms, while it is not significant for firms with low asymmetry. From the results of coefficient difference test, we find the difference between firms with the highest and lowest information asymmetry is significant. The results show that interest rate liberalization has a larger effect on the total leverage of firms with high level information asymmetry, relative to firms with low level information asymmetry.
Turning to maturity, columns 3 and 4 of panel A show that, relative to firms with low level information asymmetry, firms with high-level information asymmetry are more likely to experience a decline in short-term leverage during interest rate liberalization. The difference between them is economically significant. However, the results presented in columns 5 and 6 of panel A confirm that interest rate liberalization is negatively associated with long-term leverage of high asymmetric firms, while it has no pronounced effect on long-term leverage of low asymmetric firms. The difference checks for the coefficient on Irlib1 between the two groups are also significant.
In panel B of Table 10, we reestimate the baseline models based on information asymmetry grouping, but instead of including another proxy for interest rate liberalization (Irlib2). The results again show that relative to firms with low level information asymmetry, interest rate liberalization has a larger impact on all leverage ratios, including total leverage, short-term leverage, and long-term leverage, relative to firms with high level information asymmetry.
Overall, the results reported in Table 10 indicate that interest rate liberalization plays a bigger role in firms with high information asymmetry than in firms with low information asymmetry. Thus, it can be seen that the analysis based on asymmetric information grouping effectively supports our research logic, that is, interest rate liberalization decreases the three leverage ratios of enterprises by reducing their information asymmetry.
4.5.2. Credit Transaction Cost
Firms often establish close relations with banks through various activities, including formal and informal, to borrow more from banks. For example, enterprises can invite heads of banks to conduct field research, participate in entertainment activities and travel to enhance communication, and build special relationships with banks, and then reach borrowing agreements [26]. The costs generated by these formal and informal activities constitute the credit transaction costs of enterprises. In the period of interest rate control, competition across banks is not serious and relational loans are prevalent. Enterprises need a certain amount of relational capital to obtain bank loans. With the promotion of interest rate liberalization, the competition among banks is intensified, which leads to the reduction of the right of banks to obtain rent from enterprises [27].
If interest rate liberalization does promote banks to actively dig out more information related to enterprises and reduce the motivation of banks to establish close ties with enterprises, it is bound to reduce the opportunities for enterprises communicating with banks, thus reducing the credit transaction costs of enterprises. From the perspective of whether interest rate liberalization can reduce credit transaction costs, we examine whether interest rate liberalization promotes banks to collect more information about firms, to provide further evidence for the mechanism of interest rate liberalization on reducing corporate leverage ratios.
Since the relevant data of enterprise credit transaction costs cannot be directly obtained, we use Trcost to proxy for credit transaction costs. Trcost is the ratio of the ratio of other cash flows paid related to operating activities to total assets. We implement our estimation by including firm size (Size), cash holding(Cash), net cash flow from operating activities (Cflow), profitability (ROA), growth rate of sales revenue (Grow), executive compensation (Compen), shareholding ratio of the top five shareholders (Top5), board size (Bsize), proportion of independent directors (Indratio), property right (SOE) as control variables. We also control for industry and city dummies because some cities and industries may have been differentially impacted by interest rate liberalization.
Table 11 reports the regression with Trcost as the dependent variable. We find that the coefficients on Irlib1 and Irlib2 are both negative and significant, suggesting that firms are more likely to experience a reduction in credit transaction costs during the period of interest rate liberalization. These results provide further evidence that banks will actively collect more corporate information under the pressure of interest rate liberalization.
4.6. Additional Analysis
4.6.1. Does It Matter If the Nature of Firms Are Different?
Our evidence so far confirms that interest rate liberalization negatively affects firms’ total leverage and short-term leverage. A priori, it seems meaningful to differentiate enterprises that are state-owned and non-state-owned, because the former are likely to obtain bank loans easier than the latter. Implicit government guarantees of being a state-owned enterprise can mitigate the negative bank credit supply. On the contrary, non-state-owned enterprises’ access to bank credit is more likely to rely on their relationships with banks, and, as a result, probably more severely affected by the reduction of external finance due to credit discrimination.
To determine whether the negative relation is the same in different types of firms, we divide the samples into two groups by firm nature, that is, SOEs and Non-SOEs. If the ultimate owners of firms are governments or other state-owned enterprises, we classify them as SOEs, and Non-SOEs otherwise. Since SOEs have easier access to bank loans, we expect interest rate liberalization to have less impact on SOEs.
In Table 12 we study the differences in the impact of interest rate liberalization on leverage ratios (Lev, Slev, Llev) between SOEs and Non-SOEs. Our results show that declines both occur in SOEs and Non-SOEs. Examining the coefficients between SOEs and Non-SOEs, we see that the absolute values of the coefficients for total leverage ratio and short-term leverage ratio in Non-SOEs relative to SOEs are much larger, but the difference of coefficients for long-term leverage between SOEs and Non-SOEs is insignificant. Such results suggest that for Non-SOEs, interest rate liberalization plays a greater role, leading to a more significant decline in total leverage and short-term leverage.
For Irlib1, the reduction in total leverage ratio (Lev) reaches 25.09 percent points on average in SOEs (column 1 of panel A) and 60.37 percent points on average in Non-SOEs (column 2 of panel A). From the results of the coefficient difference test in columns 1 and 2, we see the difference between SOEs and Non-SOEs is significant. To get a sense of interest rate liberalization associated with maturity, we again estimate the baseline model with Slev and Llev as the dependent variables. Columns 3 and 4 show that the decline in short-term leverage ratio (Slev) reaches 23.89 percent points on average in SOEs and 57.98 percent points on average in Non-SOEs, the difference between them is also pronounced. As banks in developing countries are more willing to issue short-term loans, in columns 5 and 6, we see that interest rate liberalization has no pronounced effect on long-term leverage (Llev) in both SOEs and Non-SOEs (−0.0185, t = −0.42 for SOEs; −0.0407, t = −1.11 for Non-SOEs), and the difference between them is not significant.
The results in Table 12 indicate that Non-SOEs exhibit a sharper decline in total leverage and short-term leverage during the interest rate liberalization period, consistent with our expectation that the effects of interest rate liberalization on the leverage ratios between SOEs and Non-SOEs are different.
4.6.2. Does Type of Debt Matter?
As enterprises have different motives to hold financial liability and operating liability, the impact of interest rate liberalization on the two types of liabilities may also be different. Financial liability is the funds borrowed from financial markets to maintain daily operations or expand the business scale of firms, including corporate bonds, long-term and short-term bank loans, etc. Operating liability is the debts incurred by the daily business activities of firms with employees, customers, and suppliers, including wages payable, accounts payable, notes payable, and taxes payable, etc. Interest rate liberalization enables financial institutions to set differentiated lending rates based on different project risks, which may drive inefficient enterprises to reduce financing from financial markets.
To address this concern, we divide the debts of firms into financial liabilities and operating liabilities. We employ Filia and Opelia to proxy for financial liability and operating liability, respectively. Filia is the ratio of interest-bearing liabilities to total assets, and Opelia is the ratio of interest-free liabilities to total assets.
Table 13 reports the regression results that distinguish financial liability ratio and operating liability ratio. From Panel A, we see the coefficient on Irlib1 or Irlib2 for financial liability ratio is significantly negative at the 1% level, and the coefficient for operating liability ratio is negative, but not significant. The difference checks in Panel B show that interest rate liberalization has a greater impact on the financial liability of firms, relative to operating liability.
These results suggest that interest rate liberalization will reduce the willingness of enterprises to carry out interest-bearing debt financing, which leads to a more significant reduction of the financial debt ratio of enterprises, thus resulting in a decline in the total leverage ratio of enterprises.
4.6.3. Does It Matter If Firms Are Profitable or Loss Making?
In general, compared with profitable enterprises, loss-making enterprises have more incentive to take on debt. Sometimes banks do not fully follow the profit maximization principle when providing credit to enterprises, they often lend to inefficient enterprises. The ultimate goal of liberalization reform is to give full play to the leading role of market, and improve the mismatch of credit resources. Therefore, we expect that interest rate liberalization will have far more impact in reducing the leverage ratio of loss-making enterprises. To check the differentiated impact of interest rate liberalization on the leverage between Table 13 profitable and loss making firms, we construct an indicator Loss that equals 1 when the net income of a firm is less than 0, and equals 0 otherwise. In addition to the original variables in the baseline model, we add Loss and the interaction term Loss × Irlib1 (or Loss × Irlib2) into the model, and then regression it. Furthermore, we examine whether the impact of interest rate liberalization on the leverage of inefficient SOEs is weaker than that of Non-SOEs.
Table 14 reports the regression results. In columns 1 and 2, we show that the coefficients on the interaction term Loss × Irlib1 and Loss × Irlib2 are both negative and significant, suggesting that loss making firms experience more reduction in total leverage relative to profitable firms during the liberalization period. Columns 3 and 4 use the splitting sample and show that the coefficients on Loss × Irlib1 are both negative and pronounced, and the difference check of Loss × Irlib1 between SOEs and Non-SOEs is also significant, suggesting that Non-SOEs are more likely to experience a decline in leverage. Analogously, columns 5 and 6 reveal the same conclusion as columns 3 and 4. Turning to the coefficients on Loss, we see loss making firms tend to lend more in markets.
These results indicate that interest rate liberalization can improve the capital allocation among enterprises to some extent, but it is not perfect and there are still some deviations.
5. Conclusion
We employ the estimations of both the fixed effects model and difference-in-differences to explore whether the interest rate liberalization is imposed on the firm leverage ratio and its maturity, as well as the corresponding channel. Our investigation focuses on micro enterprises in China, which are characterized as having high debt ratio, particularly high short-term debt ratio.
We document several findings. First, we find that firms experience deleveraging during the period of interest rate liberalization. Compare to long-term leverage, firms enjoy more reduction in short-term leverage. Second, in the case of more serious information asymmetry, interest rate liberalization has a greater effect on reducing the leverage ratios of corporates, and interest rate liberalization can reduce the transaction cost of corporate loans, which further support the mechanism of rate liberalization reducing corporate leverage ratios. Finally, our additional checks show that deleveraging is larger in Non-SOEs, and loss making companies, relative to SOEs, and profitable firms. Moreover, deleveraging of firms is associated with a decline in the use of financial liability rather than operating liability.
Our findings offer some policy implications. As China enters the era of new financial services and eliminating financial discrimination (This is a new trend offered by the Chinese government, See, e.g. the issue on official website of the Chinese government: http://www.gov.cn/zhengce/2019-02/14/content_5365818.htm), our analysis provides experiences and lessons to excellently grasping over the ambiguous welfare effect of liberalization policy. Moreover, we find that interest rate liberalization has not completely eliminated the unequal financing treatment enjoyed by Non-SOEs. Thus, Chinese government should implement other structure reforms together with interest rate liberalization, which may mitigate financial distortions on Non-SOEs.
Data Availability
The data used to support the findings of the study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
This research was supported by Education Department of Hubei Province in China (Grant no. 21Q133, and D20201605).