Credit Expansion, Corporate Finance and Overinvestment: Recent Evidence from China

    Credit Expansion, Corporate Finance and Overinvestment: Recent Evidence from China Jianfu Shen, Michael Firth, Winnie P.H. Poon PII: DOI: Reference:

S0927-538X(16)30049-X doi: 10.1016/j.pacfin.2016.05.004 PACFIN 840

To appear in:

Pacific-Basin Finance Journal

Received date: Revised date: Accepted date:

14 August 2015 16 May 2016 21 May 2016

Please cite this article as: Shen, Jianfu, Firth, Michael, Poon, Winnie P.H., Credit Expansion, Corporate Finance and Overinvestment: Recent Evidence from China, Pacific-Basin Finance Journal (2016), doi: 10.1016/j.pacfin.2016.05.004

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Credit Expansion, Corporate Finance and Overinvestment: Recent Evidence from China Jianfu Shen

RI P

Hang Seng Management College

T

Department of Economics and Finance & Research Institute for Business Hang Shin Link, Siu Lek Yuen, Shatin Hong Kong

[email protected]

NU

Email:

852-3963-5082

SC

Phone:

Michael Firth*

MA

Department of Finance and Insurance Lingnan University 8 Castle Peak Road, Tuen Mun

ED

Hong Kong

Phone: Fax:

852-2616-8950 852-2462-1073

AC

CE

PT

E-mail: [email protected] Winnie P.H. Poon

Department of Finance and Insurance Lingnan University 8 Castle Peak Road, Tuen Mun Hong Kong Phone: Fax: E-mail:

852-2616-8179 852-2616-5326

[email protected]

___________________________________________________________________________ *Corresponding author. The authors are very grateful to the Editor, Jun-Koo Kang and an anonymous referee for their useful comments and valuable suggestions to improve the quality of this paper. The authors would also like to thank Dorla A. Evans, Zhihong Chen and the participants at the 2015 Journal of Law, Finance, and Accounting (JLFA) International Conference for their valuable comments and helpful suggestions. This paper was previously titled “Bank loan supply and corporate capital structure: Recent evidence from China”.

1

ACCEPTED MANUSCRIPT Credit Expansion, Corporate Finance and Overinvestment:

T

Recent Evidence from China

RI P

Abstract

This paper examines the impacts of a recent credit expansion event on corporate policies in China. During the credit boom in 2009 and 2010, the large and state-owned firms

SC

increased leverage ratios by 2.89% and 1.68% (on a quarterly basis) more than their matched firms. State-owned firms had higher increases in loan financing and corporate investment

ED

MA

NU

than their matched firms due to government intervention and better access to the credit market. Small and non-state-owned firms had no significant change in loan financing but undertook less net equity issuance than did the matched firms during this stimulated boom. These findings shed significant light on the effects of bank lending segmentation on capital structure and corporate investment decisions in response to macroeconomic shocks in China.

JEL Classifications: G21; G32; G38

PT

Keywords:

AC

issuance

CE

Bank loan supply; corporate investment; corporate leverage; loan financing; net equity

2

ACCEPTED MANUSCRIPT Credit Expansion, Corporate Finance and Overinvestment: Recent Evidence from China

T

1. Introduction

RI P

This paper examines the impacts of a recent credit expansion event on corporate policies in China. Monetary policy and bank loan supply are frequently used to stimulate economic

SC

growth in emerging countries. A recent phenomenon is the significant credit growth since 2008 in large emerging markets like India, China, Turkey, and Brazil (Onaran, 2013). In

NU

China, the supply of bank loans substantially increased in 2009 and 2010 following the

MA

adoption of an expansionary monetary policy. The consequences of bank loan supply shock on corporate financing policies and investment across different groups of firms in China have

ED

not been emphasized in the literature.

We discuss three aspects of the impacts related to the credit expansion event in China:

PT

the changes of leverage and loan financing, the substitutions between loan financing and equity financing, and the changes in the level of corporate investment. The finance literature

CE

argues that market frictions and the external credit supply influence a firm’s capital structure

AC

and corporate investment decisions. Firms with access to the public bond market, bank loan ratings, and traded Credit Default Swaps (CDS) are found to be associated with higher leverage than the firms without such access, loan ratings and CDS trading (Faulkender and Petersen, 2006; Sufi, 2009; Saretto and Tookes, 2013). The exogenous shocks on the supply of capital have varying impacts on capital structure and investment, which depend on firms’ abilities to raise capital (Leary, 2009; Almeida et al., 2011; Erel et al., 2012) and the substitution among different forms of external financing (Kashyap, Stein, and Wilcox, 1993; Becker and Ivashina, 2014). We explore the three issues in the context of China, where the capital market is dominated by the banking system1. 1

The banking system in China is often the sole and most important external financing source for companies, as 3

ACCEPTED MANUSCRIPT In China, firms that have a greater ability to access the credit market are large firms and state-owned enterprises/firms (SOEs), as they receive preferential treatment from the banks

T

(the major banks are state-owned)2. These firms have the advantage of easier access to bank

RI P

loans. Small and non-state-owned (non-SOE) firms have a reduced or impaired access to bank loans. We expect that positive bank lending shocks will generate more impacts on firms

SC

with the ability to obtain loans (large and state-owned firms in this study) than firms with impaired access to loans (small firms and non-SOEs).

NU

To investigate the impacts of credit expansion in 2009 and 2010, we compare the

MA

changes of corporate financing and investment before and during the credit boom in large firms, state-owned firms, small firms and non-state-owned firms (four treatment groups) with

ED

the changes of their matched firms (control groups) respectively, following the methods used by Almeida et al. (2011) and Kahle and Stulz (2013). If the differences in changes between

PT

the treatment groups of firms and their matching firms are significant, we can argue that the credit expansion significantly influenced the corporate behaviors of these treatment groups.

CE

The findings in this paper can be summarized as follows. First, we find that large firms

AC

and SOEs increased leverage ratios by 2.68% and 1.89% (on a quarterly basis) more than the matching firms during the boom period. Leverage ratios decreased more in small firms and non-SOEs than in the matching firms. State-owned firms are found to receive more bank loans (by approximately 2.28% on a quarterly basis) than matching firms. The changes of loan financing are not significantly different from control groups in large firms, small firms

non-bank financings are relatively rare due to the immature capital market. Equity issues (as well as bond issues) are subject to strict quotas set by the regulator, the China Securities Regulatory Commission (the CSRC), and in some years the quota is zero thus closing down the Initial Public Offering (IPO) market and even the market for secondary offerings. 2

The banking system is controlled by the government and is used as a policy tool for addressing national and

social priorities. Furthermore, access to credit may be determined by political considerations and connections rather than determined on a commercial basis (Firth, Lin and Wong, 2008). 4

ACCEPTED MANUSCRIPT and non-state-owned firms. Second, there is no evidence indicating that there are substitutions between loan financing and equity financing in China. The equity financing

T

decreases in all four treatment groups (large, state-owned, small and non-state-owned firms)

RI P

and the net equity issuance of small firms and non-state-owned firms decreased more than the matching firms during the boom period. State-owned firms did not use less equity financing

SC

than the matching firms even though they received more bank loans than matching firms. Third, the growth rate of corporate investment in net fixed asset increased significantly

NU

during the credit boom in state-owned firms. The SOEs with high state ownership increased

MA

corporate investment growth by 3.98% (on a quarterly basis) more than their matching firms. This finding suggests that SOEs overinvest during credit expansion as the government wants

ED

to stimulate the economy by capital spending and therefore order SOEs to invest more. This paper is the first to explore how credit expansion impacts corporate financing and

PT

investment in the world’s largest emerging market. We find that a bank lending supply shock plays an important role in the corporate policies of firms in China. Corporate financing and

CE

investment of SOEs are very sensitive to the changes in bank loans as they rely on this source

AC

of capital. Our paper also has important policy implications for China’s monetary policy, banking reform and government intervention in SOEs’ investment. First, our empirical findings support the claims highlighted in the recent report on China’s banking system from the U.S. Congressional Research Service that China’s banks give preferential treatment in lending to selected companies, usually large, state-owned and historically-served firms (Martin, 2012). China’s stimulus program from late 2008 to the end of 2010 lent more support to these companies than to small- and medium-sized firms, despite the government’s avowed intent to help small- and medium-sized firms. Second, SOEs receive more loans from state-controlled banks and invest more to help achieve the objective of the Chinese government, e.g., boosting the GDP growth. We find that the increase of investment only 5

ACCEPTED MANUSCRIPT occurs in SOEs rather than in large firms, which is consistent with overinvestment due to the government intervention (Chen et al., 2013; Deng et al., 2015).

T

The paper is structured as follows. The next section presents a brief review of the

RI P

relevant literature and provides a background on corporate financing and loan supply in China. The third section presents the data, variables, and empirical strategy. Empirical results

SC

and associated discussions are presented in section 4. The final section concludes the paper.

NU

2. Background on Corporate Financing in China and a Brief Literature Review

MA

2.1 Credit Expansion and Bank Loan Supply in China To mitigate the shocks from the global financial crisis, China adopted both an

ED

expansionary fiscal policy and monetary policy to stimulate economic growth at the end of 2008. The central government shifted its monetary policy to a moderately loose level,

PT

followed by several instruments to boost bank loan supply3 between the end of 2008 and 2010. The growth of bank loans and money supply can be easily observed from

CE

macroeconomic data. Figure 1 shows that the money supply (M2) and bank credit increased

AC

suddenly in 2008Q4. The net increases in loans in 2009 and 2010 were 9.63 trillion Yuan and 7.95 trillion Yuan, respectively. The quarterly increase in loans reached its historical peak in the first quarter of 2009, with a figure of 4.62 trillion Yuan. [Insert Figure 1 Here] The credit expansion event may have had an effect on the corporate policies of firms in China; however, the specific impacts are not immediately obvious. Some reports have shown that only 10 percent of the massive increased bank lending flowed to smaller firms (Leow, 2009; Ramzy, 2009). Yet Lardy (2012) observes that at the aggregate bank-lending level, household businesses, small-sized, and medium-sized firms obtained more loans than large 3

The details of instruments can be found in China Monetary Report Quarter Four, 2008. 6

ACCEPTED MANUSCRIPT firms in 2009 and 2010; and the growth of loans to small- and medium-sized firms was much higher than the growth in lending to large firms (See Figure 1 in Lardy, 2012). Our

T

study of China’s companies provides new evidence on the impacts of the credit expansion on

RI P

corporate financing and investment, and identifies which types of firms may benefit in a credit boom.

SC

2.2 Literature Review of Bank Lending and Corporate Policies

Some studies have explored the impacts of external bank lending on corporate policies

NU

in the US. Leary (2009) uses two events, the 1961 emergence of the market for certificates of

MA

deposit and the 1966 Credit Crunch, to investigate the impacts of bank loan shocks on capital structure. The firm’s characteristics (small or large; bank-dependent or non-bank-dependent)

ED

that are related to informational asymmetry and transaction costs in financing, along with the supply of bank loans, determine the amounts of debt that a firm may use in a specific period.

PT

The increases (decreases) of leverage in small and bank-dependent firms are more significant for positive (negative) loan supply shocks than for large firms with access to the public

CE

capital markets. He also finds that bank-dependent firms can more easily substitute equity

AC

financing and internal funds with external loan financing. Macroeconomic conditions have been found to affect a firm’s capital-raising decisions.

Erel et al. (2012) find that

lower-rated, non-investment-grade firms have a lower probability of raising capital when the overall market turns worse, but the poor macroeconomic conditions do not affect the financing ability of higher-rated firms. The availability of external financing can also influence corporate real outcomes including capital investment. Almeida et al. (2011) find that firms with large proportions of long-term debt after the 2007 financial crisis had to reduce their investment. Lemmon and Roberts (2010) examine three events that cause a reduction in capital availability to non-investment grade firms, and test the impacts of capital availability on corporate financing 7

ACCEPTED MANUSCRIPT patterns. Net security issuance and corporate investment are found to decrease in below-investment grade firms after these events. Overall, they confirm that exogenous shocks

T

in the supply of capital have significant impacts on corporate behavior, including financial

RI P

and investment decisions.

Duchin, Ozbas and Sensoy (2010) investigate the impact of the 2007 financial crisis on

SC

corporate investment. They find that corporate investment declined following the financial crisis and the negative shocks on the external capital supply, and the decrease was more

NU

substantial if the firms were financially constrained. In contrast, Kahle and Stulz (2013) find

MA

that bank-dependent firms did not have more reductions in corporate investment and net debt issuance than matched firms during the 2007 financial crisis. They did not find a substitution

ED

effect between debt issuance and equity issuance.

The literature has discussed the role of banks and bank loans in corporate investment.

PT

Bank loans can facilitate more efficient corporate investment, as banks can monitor the firms and mitigate the information problem (Diamond, 1984). However, a bank can also extract the

CE

surplus of a firm if it can intervene in the firm’s decisions (Rajan, 1992). This is the cost or

AC

dark side of bank loans. This cost may be very severe in a country where the banking system dominates the capital market. Kang and Stulz (2000) find that in Japanese firms that rely on bank loans and cannot use alternative debt financing, banking shocks have more adverse impacts on stock performance and corporate investment than those without bank loans. Wu and Yao (2012) argue that the bank rent extraction can also cause overinvestment in bank-controlled firms in Japan. Wu, Sercu and Yao (2009) indicate that public debt issuance can be used as tools to mitigate the bank rent extraction by low-growth firms whereas new equity issuance can be used by high-growth firms. Similar to the situation in Japan, the banking system plays a major role in China’s capital market. However, the major banks in China are fully controlled by the government, although 8

ACCEPTED MANUSCRIPT there have been significant reforms since 1978. A recent reform introduced in 20034 was to transform previously state-owned banks into listed, joint-stock and competitive commercial

T

entities (Hao, Shi, and Yang, 2014). As a result of this reform, five major commercial banks

RI P

in China have been listed in Hong Kong and the mainland China exchanges starting from 20055. There is an ongoing debate on whether the banks should continue to be used to

purely on commercial principles (Martin, 2012).

SC

implement policies set by the central government or whether they should operate based

NU

Many studies in China indicate that large and state-owned firms still have priority in

MA

using bank credit (Song, 2005; Ferri and Liu, 2010; Yeh, Shu and Chiu, 2013). Cull and Xu (2000; 2003) find that the state-owned firms heavily relied on bank loans from state-owned

ED

banks in the 1980s and 1990s. The state-owned banks are used as policy tools to finance investments for state-owned firms. Podpiera (2006) finds that the state-owned banks do not

PT

consider enterprise profitability in lending decisions even after the banking reforms. Small firms and non-SOEs also have severe information asymmetry as state-owned banks do not

CE

have credit information and credit records for these firms (Firth et al., 2009). Some

AC

interviews show that banks would make lending decisions to known clients rather than riskier non-state-owned firms (Martin, 2012). Bank lending can also affect corporate investment behaviors in China. Similar to bank rent extraction in Japan, China’s government can directly intervene in the investment decisions through the state-controlled banks. Firth, Lin, and Wong (2008) show that state-owned banks lend support to the capital spending of state-owned firms even though these firms have poor performance, which may cause the problem of overinvestment.

4

In 2003, the China Banking Regulatory Commission (CBRC) was established to oversee China’s banks and promote the reforms of the banking system. 5 These banks are The Agricultural Bank of China, Bank of China, China Construction Bank, Industrial and Commercial Bank of China, and the Bank of Communications. They were listed on the exchanges in 2010, 2006, 2005, 2006, and 2005, respectively. 9

ACCEPTED MANUSCRIPT Government intervention affects corporate investment decisions in China and the impacts are more significant in state-owned firms than non-state-owned firms (Chen et al, 2011; Chen et

T

al, 2013). Government-controlled firms have to increase capital expenditure even when the

RI P

investment opportunities are not profitable, so as to achieve multiple social objectives (Firth et al., 2012). In a credit boom period, the government lends more to state-owned firms and

MA

3.1 Research Design

NU

3. Research Design, Data and Variables

SC

encourages (or orders) them to invest more in real estate (Deng et al., 2015).

The event of credit expansion from 2008Q4 to 2010Q4 provides a good opportunity to

ED

investigate the impacts of bank loan supply on corporate financing and investment in China. As mentioned above, we discuss three main questions related to these impacts. First, which

PT

firms can obtain more bank loans in the event of credit expansion? This question is related to the debate whether the state-controlled banks still provide favorable treatment to state-owned

CE

firms and large firms, and whether non-state-owned firms and small firms can access the

AC

credit market since the banking reform in 2003. Second, is there a substitution effect between bank loans and equity financing in China? If some firms receive more loans during credit expansions, they may choose to issue less equity (Leary, 2009). Third, does bank loan supply affect corporate investment? As the major banks are controlled by the state in China, the government can intervene in corporate policies by determining which firms can obtain the loans and ordering these firms to make investments. The situation is similar to the holdup problem in Japan’s main banking system. Following previous studies (Leary 2009; Kahle and Stulz, 2013; Chen et al., 2012), we investigate the impacts of credit expansion on corporate financing and investment in the groups of state-owned firms vs. non-state-owned firms, and large firms vs. small firms. We 10

ACCEPTED MANUSCRIPT use a matching approach to explore the impacts of the credit boom in different groups of firms6, by comparing the changes of corporate financing and investment in these treatment

T

firms before and after credit expansion with the changes of their matching firms (Abadie and

RI P

Imbens, 2006; Almeida et al., 2011; Kahle and Stulz, 2013). We determine the groups of treatment firms by firm size and state ownership based on the sample of listed firms in

SC

China’s stock markets (the data sample will be introduced in the next section). The groups of large firms and small firms are the firms in the top and bottom quintile of total assets in our

NU

data sample at 2008Q3, following the approach used by Leary (2009). Similarly, we choose

MA

the firms in the top quintile of state ownership and the firms with zero state ownership as of 2008Q3 as the state-owned group and the non-state-owned group.

ED

Control firms with similar covariates are chosen from the non-treated firms to match the firms in each treatment group as of 2008Q3 (Abadie et al., 2004). Following the studies of

PT

Almeida et al. (2011) and Kahle and Stulz (2013), the variables used for matching firms are the market-to-book ratio, cash flow, cash holdings, size, leverage ratio, state ownership, and

CE

industrial classification code. The treatment and control firms have identical distributions in

AC

these dimensions, and the difference in corporate financing and investment can be attributed to “the treatment” (access or limited access to bank loans). The changes in corporate financing and investment in the treatment groups due to the credit boom are estimated and compared with the changes in the control groups. The Abadie-Imbens’ average effect of the treatment on the treated (ATT) is based on the difference-in-difference estimations as well as by the estimators from the traditional difference-in-difference (DID) method. Table 1 presents the conditions used to establish treatment groups and the data values of treatment groups. Panel A shows the value of total assets and state ownership at the mean,

6

The matching approach is a difference-in-difference method. We also use a multivariate regression method

similar to Duchin, Ozbas and Sensoy (2010). The results still hold. 11

ACCEPTED MANUSCRIPT minimum, 20th percentile, 40th percentile, 60th percentile, 80th percentile, and maximum. These values are used to create four treatment groups: large firms, state-owned firms, small

T

firms and non-state-owned firms. Panel B gives the conditions used to form the four groups.

RI P

There are 283 firms in the groups of both large firms and small firms; 284 in the group of state-owned firms; and 497 in the group of non-state-owned firms. The average values of

SC

total assets and state ownership in the four groups are also reported in Panel B. [Insert Table 1 Here]

NU

3.2 Data and Variables

MA

We test the impacts of loan supply caused by the credit boom event occurring from 2008Q4 to 2010Q4 in China. Our sample consists of all listed companies on the Shanghai and

ED

Shenzhen Stock Exchanges with quarterly data, excluding the firms in the finance and public utility industries. The firm-level data are obtained from the China Stock Market and

PT

Accounting Research Database (CSMAR). The macroeconomic data are from National Bureau of Statistics of China and the People’s Bank of China. The total time period covers

CE

from 2006Q3 to 2010Q47. Similar to previous studies (Almeida et al., 2011; Kahle and Stulz,

AC

2013), we drop observations with negative values in total assets, sales or cash holdings. If the asset growth or sales growth is larger than 100% in a quarter, the observations are also excluded. Observations are also deleted if total liabilities are larger than total assets or market value of total assets. Firm variables are further winsorized at the 99% level (top and bottom 0.5%) to reduce the outlier effects. We construct variables to measure corporate financing and investment. The variable LEVERAGE is the leverage ratio, which is total liabilities divided by total assets. Bank loan

7

There are 9 quarters in the period of credit expansion. We take 9 quarters before credit boom as the pre-boom

period and use them as comparison to calculate the changes in corporate investment and financing due to changes in loan supply. 12

ACCEPTED MANUSCRIPT financing is measured by the variable LOAN, calculated as proceeds from banks or other financial institutions divided by total assets. Net equity issuance (EQUITY) is constructed by

T

proceeds from equity issuance net of dividend payment, scaled by total assets. The corporate

RI P

investment (INV) is defined as change in net fixed assets plus depreciation, divided by the beginning net fixed assets8, following Kang and Stulz (2000) and Firth et al. (2012). Several

SC

variables are included as control variables9: market-to-book value (MTBV), cash flow (CF), cash holdings (CASH), state ownership (STATESHARE), and total assets (LNTA). The

NU

definitions of variables are given in Appendix 1.

MA

[Insert Appendix 1 Here] Panel A of Table 2 reports the descriptive statistics of variables for the sample of

ED

firm-quarter observations. The mean and median of corporate investment ratio are 8.78% and 3.75%, respectively. The average leverage ratio is 50.76% of total assets. Loan financing is Cash flow (a measure of

PT

very large in China, averaging 16.61% of total assets.

internally-generated funds) is 3.91% of total assets. The funds raised from net equity issuance

CE

is 2% of total assets, slightly lower than internally-generated funds. The holding of cash and

AC

cash equivalents (CASH) is high (16.21% of total assets) in China’s firms. On average, 19.20% of shares are owned by state. [Insert Table 2 Here]

The mean values of corporate investment, leverage ratios, bank loan financing and net equity issuance in the treatment groups are presented in Panel B of Table 2. Large firms have higher corporate investment, leverage ratios, and bank loan financing than small firms; 8

We use several alternative measures of corporate investment, such as change in net fixed assets plus

depreciation, divided by total assets (Chen et al., 2013) or cash payments for fixed assets, intangible assets and other long term assets net of cash receipts from these assets, divided by beginning total assets (Chen et al., 2011). The results remain similar. 9

These control variables are the same variables as those used in Almeida et al. (2011) and Kahle and Stulz

(2013). 13

ACCEPTED MANUSCRIPT however, the small firms use more equity financing. In state-owned firms and non-state-owned firms, the two groups have similar average values in corporate investment

T

and bank loan financing. State-owned firms have higher leverage ratios and less net equity

RI P

financing than non-state-owned firms, similar to the comparisons between large firms and small firms.

SC

Panel C of Table 2 provides nonparametric results of corporate financing and investment before and during credit boom periods. It shows that corporate investment and loan financing

NU

increased significantly during the credit boom. The increase in loan financing can be

MA

attributed to increased external credit supply after the credit expansion. However, leverage and net equity issuance decreased in the period. The market–to-book value, a proxy for

ED

investment opportunities, increased significantly following the onset of credit expansion.

PT

Total assets also increased in the periods of credit boom.

4. Discussion of Results

CE

4.1 Credit Expansion and Capital Structure

AC

In this section, we present the results of how leverage ratios of the four treatment groups changed during the credit boom from 2008Q4 to 2010Q4 in China. The large firms and state-owned firms may have had greater increases in leverage ratios during the lending boom than other firms because they had easier access to the bank loan market. Table 3 reports the estimators for quarterly leverage ratios before and during the credit boom (2008Q4 – 2010Q4)10. The data in Panel A of Table 3, column 1, show that the average quarterly leverage ratios

10

We also examine the changes in leverage ratios in the treatment groups one year before the boom, and the

first year of the credit boom (2007Q4-2008Q3 vs. 2008Q4-2010Q4), and the first quarter of 2007 and the first quarter of 2008 (2007Q1 vs. 2008Q1). The results are similar to the findings for the whole credit boom period. 14

ACCEPTED MANUSCRIPT slightly increased during the credit boom. For the treatment groups, we find that the leverage of large firms and SOEs increased by 2.97% and 2.38%, respectively, in the credit boom

T

period compared with leverage in the pre-boom period. The increases in the leverage of the

RI P

two groups are significant at the 1% level. However, the leverage ratios in small firms and non-SOEs dropped significantly in the whole credit boom period. The decreases in leverage

SC

ratios were 1.21% and 1.22% for the two groups during the credit boom period. [Insert Table 3 Here]

NU

The key variables, DID and ATT, in the credit boom period are reported in Panel B. We

MA

find that the increases in leverage ratios in the control groups were significantly less than the increases in treatment groups of large and state-owned firms in the credit boom period.

ED

According to the ATT estimators, during the credit boom the large and state-owned firms had 2.89% and 1.68% greater increases in leverage than their matched firms. However, leverage

PT

ratios of small firms and non-SOEs decreased in comparison with matched firms during the credit boom. In particular, the leverage of non-state-owned firms significantly decreased by

CE

1.85% and 1.55% according to the DID and ATT estimations.

AC

We plot the average quarterly leverage ratios of the four treatment groups before and during the credit boom period in Figure 2. The results are similar to the results by DID and ATT estimations. The average leverage ratios in large and state-owned firms increased during the credit boom period. However, we do not observe the same increasing pattern in leverage ratios of small and non-state-owned firms. In fact, the leverage ratios of small and non-state-owned firms decreased in 2008 (before the credit boom), possibly due to the shock of global financial crisis. Non-state-owned firms did not increase leverage in the credit boom periods. The leverage ratios of small firms slightly increased during the credit boom but their leverage ratios did not increase to the level in 2007. [Insert Figure 2 Here] 15

ACCEPTED MANUSCRIPT In sum, we find that the leverage ratios increased in the groups of large firms and state-owned firms, but decreased in the groups of small firms and non-state-owned firms in

T

comparison with their matching firms. We explain this pattern as the result of the external

RI P

bank lending shock created by the credit expansion policy and the access to bank loans. Large firms and state-owned firms can easily access the bank loan market in the state-controlled

SC

banking system, which increases their debt financing in the periods of credit expansion. Another plausible explanation is that large firms and state-owned firms may have shrunk

NU

their equities in 2009 and 2010 due to the shock of the global financial crisis, which also

MA

results in a lower debt ratio11. However, we find that the total equities in large firms and state-owned firms (and also small firms and non-state-owned firms) slightly increased rather

ED

than decreased during the credit boom (in unreported results). Thus the increase of leverage ratios should more likely be attributed to the increased debt financing.

PT

4.2 Credit Expansion and Loan Financing We conduct the same tests for the loan financing ratio before and during the credit boom

CE

periods. Panel A of Table 4 shows that during the credit boom periods, the loan financing

AC

ratios significantly increased by 1.04%, 1.93%, and 2.01% of total assets for the firms in the full sample, the large-firm group, and the state-owned group, respectively. The loan financing ratios also increased slightly in small firms and non-SOEs (1.07% and 0.28%, respectively) during the credit boom; however, the increase in non-SOEs is not significant. The results indicate that credit expansion does have significant impacts on loan financing among listed firms but the impacts are larger in large firms and state-owned firms than small firms and non-state-owned firms. [Insert Table 4 Here] Panel B shows DID and ATT estimators for the treatment groups. State-owned firms 11

We thank the referee for raising this issue. 16

ACCEPTED MANUSCRIPT have significant increases in loan financing using both the DID and ATT estimations in the credit boom period. Within the credit boom period, the ratios of loan financing over total

T

assets in state-owned firms are 1.89% and 2.28% higher than the ratios of the matched firms

RI P

according to DID and ATT estimations. The estimations are significant at the 1% and 5% levels, respectively. The estimators for large firms are not significant. The DID and ATT

impacts of the credit boom are trivial or negative.

SC

estimators for small firms and non-SOEs are negative or not significant, indicating that the

NU

Overall, we find that the positive shock in credit supply played a significant and positive

MA

role in the financial policies of large firms and SOEs because these firms were able to access bank loans in China’s banking system. The small firms and non-SOEs, however, were not Increased loan financing helps explain the findings in Table 3

ED

affected by the credit boom.

that leverage ratios of large firms and state-owned firms significantly increased during the

PT

credit boom. It also presents strong evidence that state-controlled banks still provide preferential treatment to state-owned firms.

CE

4.3 Credit Expansion and Net Equity Issuance

AC

Some studies investigate the substitution between debt financing and equity financing in bank-dependent firms in the U.S. (Leary, 2009; Kahle and Stulz, 2013). In China, a seasoned equity offering (SEO) is highly regulated by the government, which requires that in three years before the application of a SEO firms should have a profit and an average return on equity greater than 10% (Poon, Chan and Firth, 2013). It is unclear whether large and state-owned firms have priority to use equity financing in the capital market and whether they would reduce equity financing when debt financing is increased. Table 5 reports the results for quarterly net equity issuance before and during credit expansion. Net equity issuance is defined as the proceeds from equity issuance net of dividend payments, scaled by total assets. Panel A shows that small firms and 17

ACCEPTED MANUSCRIPT non-state-owned firms used more equity financing before and during credit boom than large firms and state-owned firms. These two groups have restricted access to the credit market and

T

thus have to rely more on equity financing than firms with access. The net equity issuance

RI P

reduced significantly in all four treatment groups. The decreases in net equity issuance were 1.17%, 1.69%, 6.76%, and 5.04% of total assets in the groups of large firms, state-owned

SC

firms, small firms and non-state-owned firms, respectively. Interestingly, there were larger decreases in small firms and non-state-owned firms than large firms and state-owned firms.

NU

[Insert Table 5 Here]

MA

Panel B provides the estimators from the matching approach. The results indicate that there are no significant differences in changes of net equity issuance in large firms and

ED

state-owned firms, compared with their matching firms. The small firms and non-state-owned firms have significantly greater reductions of 3.74% and 2.32% in net equity issuance in

PT

comparison with matched firms according to ATT estimations. Both DID and ATT estimators are significant at the 1% level.

CE

Combining the results in Tables 4 and 5, we find that SOEs had greater increases in loan

AC

financing than their matched firms but no significant differences in net equity issuance. Small firms and non-SOEs had larger decreases in net equity issuance than their matched firms but did not receive more bank lending. The findings are not consistent with the argument of a substitution effect that firms using more (less) debt financing issue less (more) equity (Leary, 2009). We can only confirm that small firms and non-SOEs, with restricted access to credit market, have to rely on equity issuance to raise capital. The decreases of equity financing in small firms and non-SOEs during the credit boom period may have been caused by the declining stock market and suspension of IPOs in China after 2008 (Bo, Huang and Wang, 2011). It is interesting to find that small and non-state-owned firms used more equity financing 18

ACCEPTED MANUSCRIPT than large and state-owned firms in the whole sample period (before and during the boom period). In unreported results, we find that the equity financings in small and

T

non-state-owned firms were consistently larger than equity financings in large and

RI P

state-owned firms between 2003 and 201012. Fama and French (2005) find that small and growing firms can still issue equity although they have severe asymmetric information

SC

problem. Small firms and non-SOEs in China have more growth opportunities and fewer tangible assets than large firms and SOEs 13 . Asymmetric information about growth

NU

opportunities arises in small firms and non-SOEs while large firms and SOEs have more

MA

asymmetric information about assets-in-place. The different types of asymmetric information problems may explain why small firms and non-SOEs issue more equity that their

ED

counterparties (Wu and Au Yeung, 2012). Small firms and non-SOEs have good projects but have restricted access to external debt financing. Hence, they have to rely on equity financing.

PT

The equity issuance in these firms may not necessarily be taken as the signal of bad projects or stock overvaluations (Wu and Wang, 2005), so the cost is not as high as shown in the

CE

classic pecking order theory.

AC

4.4 Credit Expansion and Corporate Investment We examine the impacts of bank lending on corporate investment in this section. Panel A of Table 6 shows that only state-owned firms had a significant increase in corporate investment during the credit boom. The changes of corporate investment were not significant in large firms, small firms or non-state-owned firms. Panel B indicates that state-owned firms had significantly greater increases in corporate investment than their matching firms during

12

The average quarterly net issuances are 0.0209, 0.0257, 0.0694 and 0.0540 in large firms, SOEs, small firms

and non-SOEs, respectively, between 2003 and 2010. 13

The average quarterly market-to-book ratios are 1.6516, 2.2021, 3.4266 and 2.4979 and the average quarterly

tangibility ratios are 0.3129, 0.2944, 0.2542 and 0.2544 for large firms, SOEs, small firms and non-SOEs, respectively, between 2003 and 2010. 19

ACCEPTED MANUSCRIPT credit boom. According to the DID and ATT estimations, the corporate investment ratios increased by 4.36% and 3.98% (quarterly growth of net fixed assets) more than the matched

T

firms. The coefficients are significant at the 1% level. In contrast, corporate investment

RI P

decreased significantly more in non-state-owned firms than their matching firms. The changes of corporate investment are not significantly different from matching firms in the

SC

groups of large firms or small firms.

[Insert Table 6 Here]

NU

The finding that only state-owned firms invested more during the credit boom is quite

MA

interesting. One explanation for the increase of corporate investment in SOEs is based on the demand shock caused by SOEs’ taking government-led projects during credit expansion. In

ED

this period, China implemented expansionary fiscal policy and launched a 4 trillion Yuan stimulus program (Xinhua, 2008), including projects in affordable housing, rural

PT

infrastructure, transport infrastructure, environmental projects, and earthquake restructure (Naughton, 2009). These investment projects are normally allocated to state-owned firms by It is not surprising that state-owned firms increased investment if they

CE

the government14.

were in the industries related to the stimulus program. However, our treatment group of

AC

state-owned firms has excluded firms in public utilities15. In addition, there are 47 different industries represented in the 283 state-owned firms, most of which are not relevant to the projects in the stimulus program. The increase of corporate investment may not be the result 14

After the central government announced the stimulus program, local governments in each province proposed

a list of projects. The National Development and Reform Commission (NDRC) established the spending guidelines for sectors and projects and checked the proposed projects. The investment projects are mostly related to infrastructure. The discussions of the stimulus program can be found in Naughton (2009). 15

The sectors in the public utility industry include airport transportation, highway transport, railway

transportation, water transportation, communication service, information service, public facilities service, forestry, support service for transportation, water generation and supply, electric power, steam and hot water generation and supply, and other public service. These industries are closely related to the fiscal stimulus program. 20

ACCEPTED MANUSCRIPT of demand shock directly from stimulus program. The literature of bank rent extraction shows that banks can extract most of firms’ profits

T

from the positive NPV projects if the bank-firm relation is close (Rajan, 1992).

RI P

Bank-dependent firms in Japan contract investment when bank financing was reduced in the early 1990s (Kang and Stulz, 2000). Our results reveal that SOEs, with close relations with

SC

state-controlled banks, increased investment more when the bank loan supply was increased in China. In 2009 and 2010, the Chinese government used bank lending to increase capital

NU

spending and boost economic growth 16 ; and the government gave more lending to

MA

state-owned firms and encouraged them to invest more (Deng et al., 2015). The finding is also consistent with previous studies (Chen et al., 2011; Chen et al., 2013) indicating that

ED

government intervention can affect corporate investment decisions in China and the impacts

PT

are more significant in state-owned firms than non-state-owned firms.

5. Conclusions

CE

In this paper, we explore the question of whether a bank lending shock affects corporate

AC

financing and investment in China’s listed companies during the credit boom occurring from 2008Q4 to 2010Q4. We find that bank loan shocks significantly affected corporate leverage for large or state-owned firms, which are given privileged access to the banking system. Firms with the ability to obtain bank loans had larger increases in leverage ratios than firms with limited access to banking facilities when the bank loan supplies increased. SOEs received more bank loans during the credit boom than their matching firms, while loan financing did not change significantly in small firms or non-state-owned firms. We also show that the net equity issuance decreased significantly all in large firms,

16

According to the monetary policy report (PBC, 2008), the purposes of loose monetary policy are to “boost

domestic demand”, provide “financial support for economic development” and maintain economic growth. 21

ACCEPTED MANUSCRIPT state-owned firms, small firms and non-state-owned firms during credit boom.

However,

net equity issuance did not decrease more for large firms or state-owned firms than their

T

matching groups, although they received more loans than matching firms. In contrast, the

RI P

small firms and non-state-owned firms decreased net equity issuance more than their matching firms. There was no substitution effect between loan financing and equity financing

SC

in China during the credit boom.

Lastly, we find that only SOEs had significant increases in corporate investment during

NU

credit boom. The corporate investment increased more in SOEs than their matching firms.

MA

The increase of corporate investment was driven by the government’s intervention to invest more in SOEs to achieve the goal of economic growth (Deng et al., 2015). Using

ED

investment-to-GDP ratios, Lee, Syed and Liu (2012) find that China has overinvested after 2000 and the overinvestment was more substantial when China adopted fiscal and monetary

PT

policies to tackle the global financial crisis. The investment-to-GDP rose from 41% in 2007 to 48% in 2010 (Gros, 2015). Our evidence shows that the overinvestment occurred in the

CE

SOE sector and was financed by bank loans during the credit boom period.

AC

Overall, these results suggest that bank lending plays a significant role in determining capital structure and corporate investment in China. These findings also have important policy implications for the reforms of the banking system in China. The monetary policy in China may have little impact on helping small firms and non-SOEs as long as access to banking facilities is restricted and in the absence of a liquid corporate bond market. In contrast, the credit expansion is associated with more loan financing and corporate investment of SOEs. The capital structure and corporate investment of SOEs are sensitive to bank loan supply due to the government’s intervention and the link between SOEs and state-controlled banks.

22

ACCEPTED MANUSCRIPT References: Abadie, A. & Imbens, G.W. 2006, "Large sample properties of matching estimators for average

T

treatment effects", Econometrica, 74(1), 235-267.

RI P

Abadie, A., Drukker, D., Herr, J.L. & Imbens, G.W. 2004, "Implementing matching estimators for average treatment effects in Stata", Stata Journal, 4(3), 290-311.

SC

Almeida, H., Campello, M., Laranjeira, B. & Weisbenner, S. 2011, “Corporate debt maturity and the

NU

real effects of the 2007 credit crisis”, Critical Finance Review, 1(1), 3-58. Becker, B. & Ivashina, V. 2014, "Cyclicality of credit supply: Firm level evidence", Journal of

MA

Monetary Economics, 62, 76-93.

Bo, H., Huang, Z., & Wang, C. 2011, “Understanding seasoned equity offerings of Chinese

ED

firms”, Journal of Banking & Finance, 35(5), 1143-1157. Chen, D., Khan, S., Yu, X., & Zhang, Z. 2013, “Government intervention and investment

PT

comovement: Chinese evidence”, Journal of Business Finance & Accounting, 40(3-4), 564-587. Chen, S., Sun, Z., Tang, S., & Wu, D. 2011, “Government intervention and investment efficiency:

CE

Evidence from China”, Journal of Corporate Finance, 17(2), 259-271. Cull, R. & Xu, L.C. 2003, "Who gets credit? The behavior of bureaucrats and state banks in allocating

AC

credit to Chinese state-owned enterprises", Journal of Development Economics, 71(2), 533-559. Cull, R. & Xu, L.C. 2000, "Bureaucrats, state banks, and the efficiency of credit allocation: The experience of Chinese state-owned enterprises", Journal of Comparative Economics, 28(1), 1-31. Deng, Y., Morck, R., Wu, J., & Yeung, B. 2015, “China’s pseudo-monetary policy”, Review of Finance, 19(1), 55-93. Diamond, D. W. 1984, “Financial intermediation and delegated monitoring”, The Review of Economic Studies, 51(3), 393-414. Duchin, R., Ozbas, O. & Sensoy, B.A. 2010, "Costly external finance, corporate investment, and the subprime mortgage credit crisis", Journal of Financial Economics, 97(3), 418-435. 23

ACCEPTED MANUSCRIPT Erel, I., Julio, B., Kim, W. & Weisbach, M.S. 2012, "Macroeconomic conditions and capital raising", Review of Financial Studies, 25(2), 341-376. Fama, E. F., & French, K. R. 2005, “Financing decisions: Who issues stock?”, Journal of Financial

RI P

T

Economics, 76(3), 549-582.

Faulkender, M. & Petersen, M.A. 2006, "Does the source of capital affect capital structure?", Review

SC

of Financial Studies,19(1), 45-79.

Ferri, G. & Liu, L. 2010, "Honor thy creditors before thy shareholders: are the profits of Chinese

NU

state-owned enterprises real?", Asian Economic Papers,9(3), 50-71. Firth, M., Lin, C., Liu, P. & Wong, S.M. 2009, "Inside the black box: bank credit allocation in China’s

MA

private sector", Journal of Banking & Finance, 33(6), 1144-1155. Firth, M., Lin, C., & Wong, S. M. 2008, “Leverage and investment under a state-owned bank lending

ED

environment: Evidence from China”, Journal of Corporate Finance, 14(5), 642-653. Firth, M., Malatesta, P. H., Xin, Q., & Xu, L. 2012, “Corporate investment, government control, and

PT

financing channels: Evidence from China's listed companies”, Journal of Corporate

CE

Finance, 18(3), 433-450.

Gros, D. 2015, “Monetary policy and the over-investment cycle: China as an extreme case”, CEPS

AC

commentary, 27 August 2015. Hao, X., Shi, J. & Yang, J. 2014, "The differential impact of the bank–firm relationship on IPO underpricing: Evidence from China", Pacific-Basin Finance Journal, 30, 207-232. Kahle, K.M. & Stulz, R.M. 2013, "Access to capital, investment, and the financial crisis", Journal of Financial Economics, 110(2), 280-299. Kang, J., & Stulz, R. M. 2000, “Do banking shocks affect borrowing firm performance? an analysis of the Japanese experience”, The Journal of Business, 73(1), 1-23. Kashyap, A.K., Stein, J.C. & Wilcox, D.W. 1993, "Monetary policy and credit conditions: evidence from the composition of external finance", The American Economic Review, 83(1), 78-98. Lardy, N.R. 2012, “Sustaining China's economic growth after the global financial crisis”, Peterson 24

ACCEPTED MANUSCRIPT Institute for International Economics, Washington, D.C. Leary, M.T. 2009, "Bank loan supply, lender choice, and corporate capital structure", The Journal of

T

Finance, 64(3), 1143-1185.

RI P

Lee, M. I. H., Syed, M. M. H., & Liu, X. 2012, “Is China over-investing and does it matter?”, IMF working paper, WP/12/277.

SC

Lemmon, M. & Roberts, M.R. 2010, "The response of corporate financing and investment to changes

NU

in the supply of credit", Journal of Financial and Quantitative Analysis, 45(3), 555-587. Leow, J. 2009, “China loans hard to get—smaller enterprises left dry as bulk of lending goes to big

MA

projects”, Wall Street Journal, May 14, 2009.

Martin, M.F. 2012, "China's banking system: issues for Congress", CRS Report for Congress, Feb 20,

ED

2012.

Naughton, B. 2009, “Understanding the Chinese Stimulus Package”, China Leadership Monitor, no.

PT

28.

Podpiera, R. 2006, "Progress in China's Banking Sector Reform: Has Bank Behavior Changed?", IMF

CE

Working Paper, WP/06/71.

Onaran, Y. 2013, “Bad loans could spark an emerging-markets crisis”, Bloomberg Businessweek, Nov

AC

21, 2013.

People’s Bank of China (PBC) 2008, “China monetary policy report quarter four, 2008”, available in the website: http://www.pbc.gov.cn/publish/english/982/index.html. Poon, W. P., Chan, K. C., & Firth, M. A. 2013, “Does having a credit rating leave less money on the table when raising capital? A study of credit ratings and seasoned equity offerings in China”, Pacific-Basin Finance Journal, 22, 88-106. Rajan, R. G. 1992, “Insiders and outsiders: The choice between informed and arm's‐length debt”, The Journal of Finance, 47(4), 1367-1400. Ramzy, A. 2009, “In China's lending boom, small businesses go begging”, TIME, Jul 15, 2009. 25

ACCEPTED MANUSCRIPT Saretto, A. & Tookes, H.E. 2013, "Corporate leverage, debt maturity, and credit supply: the role of credit default swaps", Review of Financial Studies, 26(5), 1190-1247. Song, L. 2005, "Interest rate liberalization in China and the implications for non-state banking", in

RI P

Cambridge, MA: Harvard University Asia Center, 111-130.

T

Yasheng Huang, Tony Saich and Edward Steinfeld, eds, Financial Sector Reform in China,

Sufi, A. 2009, "The real effects of debt certification: Evidence from the introduction of bank loan

SC

ratings", Review of Financial Studies, 22(4), 1659-1691.

Banking & Finance, 36(12), 3427-3443.

NU

Wu, X., & Au Yeung, C. K. 2012, “Firm growth type and capital structure persistence”, Journal of

MA

Wu, X., Sercu, P., & Yao, J. 2009, “Does competition from new equity mitigate bank rent extraction? Insights from Japanese data”, Journal of Banking & Finance, 33(10), 1884-1897.

ED

Wu, X., & Wang, Z. 2005, “Equity financing in a Myers–Majluf framework with private benefits of control”, Journal of Corporate Finance, 11(5), 915-945.

PT

Wu, X., & Yao, J. 2012, “Understanding the rise and decline of the Japanese main bank system: The

CE

changing effects of bank rent extraction”, Journal of Banking & Finance, 36(1), 36-50. Xinhua News Agency, 2008, “China plans 10 major steps to spark growth as fiscal, monetary policies

AC

ease”, Xinhua News Agency Beijing, Nov 9, 2008. Yeh, Y., Shu, P. & Chiu, S. 2013, "Political connections, corporate governance and preferential bank loans", Pacific-Basin Finance Journal, 21(1), 1079-1101.

26

ACCEPTED MANUSCRIPT Table 1: Total assets, state ownership and treatment groups This table shows how the treatment groups are constructed. Panel A presents the values of total assets (million Yuan) and state ownership at the mean, minimum, 20th percentile, 40th percentile, 60th percentile, 80th percentile

T

and maximum on 2008Q3. Panel B indicates the conditions and the number of firms for four treatment groups.

40th

60th

80th

Min

percentile

percentile

percentile

percentile

Max

Total assets

6794.48

63.53

839.33

1462.77

2532.56

5328.89

1122650

State ownership

0.2218

0

0

0.0483

0.2798

0.4486

0.7331

SC

Mean

NU

Panel A

20th

RI P

The average total assets (million Yuan) and state ownership for the treatment groups are also given in Panel B.

Total assets

State ownership

N

(mean)

(mean)

Condition

Large firms

In top 20% of total assets

283

26777.48

0.3012

State-owned firms

In top 20% of state ownership

284

17057.78

0.5529

Small firms

In bottom 20% of total assets

283

552.96

0.1527

Non-state-owned firms

With state ownership = 0

497

3264.62

0

AC

CE

PT

ED

MA

Panel B

27

ACCEPTED MANUSCRIPT Table 2: Descriptive statistics and nonparametric tests Panel A presents the descriptive statistics for the full sample (2006Q3 – 2010Q4). INV is the corporate investment ratio. LEVERAGE is the book leverage ratio. LOAN is funds from borrowings to total assets. EQUITY measures

T

the net equity issuance. CASH is the cash holdings scaled by total assets. MTBV is market–to-book value. CF is

RI P

the cash flow divided by total assets. STATESHARE is state ownership. LNTA is the natural logarithm of total assets. The categorical variables LARGE, STATE, SMALL NONSTATE are dummy variables to measure large firms, state-owned firms, small firms and non-state-owned firms. The detailed definitions of other control

SC

variables can be found in Appendix 1.

Panel B reports the means of investment, leverage, loan financing and net equity issuance in treatment groups. Panel C presents the means of some key variables before credit boom (2006Q3 – 2008Q3) and during credit boom

NU

(2008Q4 – 2010Q4). It also gives the differences of the means and their t-statistics before and during the credit

Panel A: descriptive statistics in full sample

Variable

Mean

Obs

MA

boom.

Median

Std. Dev.

Min

Max

0.0878

0.0375

0.2469

-0.5306

2.2047

0.5076

0.5201

0.1863

0.0580

0.9995

0.1661

0.1234

0.1587

0

0.6830

Dependent variables and key variables 18999

LEVERAGE

19854

LOAN

19476

EQUITY

18187

0.0200

0

0.0735

-0.0114

0.6703

CASH

19854

0.1621

0.1321

0.1210

0.0023

0.7233

19389

2.5473

2.0055

1.8219

0.8837

16.2798

19450

0.0391

0.0303

0.0477

-0.1614

0.2452

19633

0.1920

0.0728

0.2215

0

0.7331

19854

21.5134

21.4040

1.1601

18.3440

24.9918

LARGE

19854

0.1999

0

0.3999

0

1

STATE

19854

0.1997

0

0.3998

0

1

SMALL

19854

0.1894

0

0.3919

0

1

NONSTATE

19854

0.3411

0

0.4741

0

1

PT

CF

AC

STATESHARE

CE

MTBV

LNTA

ED

INV

Categorical variables

Panel B: Mean values of investment, leverage, loan financing and net equity issuance in treatment groups

Treatment group

INV

LEVERAGE

LOAN

EQUITY

Large firm

0.1043

0.6012

0.2022

0.0157

Small firm

0.0859

0.4120

0.1157

0.0328

State-owned firm

0.0930

0.5308

0.1615

0.0166

Non-state-owned firm

0.0951

0.4660

0.1643

0.0302

28

ACCEPTED MANUSCRIPT

Panel C: nonparametric tests (before and during credit boom)

Pre-boom

Diff.

INV

0.0835

0.0918

LEVERAGE

0.5100

0.5053

LOAN

0.1623

0.1696

EQUITY

0.0280

0.0135

MTBV

2.5173

LNTA

21.3906

(Boom - Pre-boom)

T

(2006Q3-2008Q3) (2008Q4-2010Q4)

0.0083**

RI P

Variable

Boom

t-statistic 2.3220 1.7630

0.0073***

3.1921

-0.0145***

13.3138

2.5757

0.0584**

2.2294

21.6321

0.2415***

14.7407

AC

CE

PT

ED

MA

NU

SC

-0.0047*

29

ACCEPTED MANUSCRIPT Table 3: Quarterly leverage ratio Panel A gives the quarterly leverage ratio for the whole sample and four firm treatment groups. It reports the average leverage ratio in these samples before and during the credit boom. The groups of large firms and small firms are the firms in the top and bottom quintile of total assets in our data sample on 2008Q3. The state-owned

T

group and non-state-owned group are the firms in the top quintile of state ownership and the firms with zero

RI P

state ownership at the end of the third quarter of 2008. We compare the differences of leverage ratios before and during the credit boom in the whole sample and subsamples using a t-test.

SC

Panel B reports difference-in-difference results for the treatment groups and their control groups. The approach of Abadie et al. (2004) is employed to match the firms in these four treatment groups with non-treatment firms in the data sample on 2008Q3. The variables used in the matching process are the market-to-book ratio, cash

NU

flow, cash holdings, size, state ownership, and industrial classification code, following the methods used in Almeida et al. (2011) and Kahle and Stulz (2013). DID is the traditional difference-in-difference estimator. ATT

MA

is the Abadie-Imbens bias-corrected average treated effect on the treatment group. ***, **, and * are 1%, 5%, and 10% significance levels using heteroskedasticity-consistent standard errors, respectively.

Panel A: Quarterly leverage ratio before and during credit boom Large

State-owned

Small

Non-state-owned

(1)

(2)

(3)

(4)

(5)

0.5065

0.5889

0.5218

0.4180

0.4728

Boom period (2008Q4-2010Q4)

0.5100

0.6186

0.5456

0.4059

0.4607

Difference (Boom - Pre-boom)

0.0035*

0.0297***

0.0238***

-0.0121**

-0.0122***

1413

283

283

282

495

Period averages

AC

Number of firms

CE

PT

Pre-boom (2006Q3-2008Q3)

ED

Whole

Panel B: The average changes of leverage ratio before and during credit boom Large

State-owned

Small

Non-state-owned

(1)

(2)

(3)

(4)

Pre-boom (2006Q3-2008Q3) versus boom period (2008Q4-2010Q4) Treatment: boom - pre-boom

0.0027

0.0057

-0.0121

0.0064

Control: boom - pre-boom

0.0297

0.0238

0.0010

-0.0122

DID

0.0269***

0.0182**

-0.0132*

-0.0185***

ATT

0.0289***

0.0168**

-0.0048

-0.0155**

Number of treatment firms

283

283

282

495

Number of observations in matching

1413

1413

1413

1413

30

ACCEPTED MANUSCRIPT Table 4: Quarterly loan financing ratio Panel A gives the quarterly loan financing ratio for the whole sample and four firm treatment groups. It reports the average loan financing ratio in these samples before and during the credit boom. The groups of large firms and small firms are the firms in the top and bottom quintile of total assets in our data sample on 2008Q3.

T

State-owned group and non-state-owned group are the firms in the top quintile of state ownership and the firms

RI P

with zero state ownership at the end of the third quarter of 2008. We compare the differences in loan financing ratios before and during the credit boom in the whole sample and subsamples using a t-test.

SC

Panel B reports difference-in-difference results for the treatment groups and their control groups. The approach of Abadie et al. (2004) is employed to match the firms in these four treatment groups with non-treatment firms

NU

in data sample on 2008Q3. The variables used in the matching process are the market-to-book ratio, cash flow, cash holdings, size, leverage ratio, state ownership, and industrial classification code, following the methods used in Almeida et al. (2011) and Kahle and Stulz (2013). DID is the traditional difference-in-difference

MA

estimator. ATT is the Abadie-Imbens bias-corrected average treated effect on the treatment group. ***, **, and * are 1%, 5%, and 10% significance levels using heteroskedasticity-consistent standard errors, respectively.

Panel A: Quarterly loan financing ratio before and during credit boom Large

State-owned

Small

Non-state-owned

(1)

(2)

(3)

(4)

(5)

0.1587

0.1914

0.1488

0.1101

0.1612

Boom period (2008Q4-2010Q4)

0.1691

0.2107

0.1690

0.1208

0.1640

Difference (Boom - Pre-boom)

0.0104***

0.0193***

0.0201***

0.0107**

0.0028

1413

283

283

282

495

Period averages

AC

Number of firms

CE

PT

Pre-boom (2006Q3-2008Q3)

ED

Whole

Panel B: The average changes of quarterly loan financing ratio before and during credit boom Large

State-owned

Small

Non-state-owned

(1)

(2)

(3)

(4)

Pre-boom (2006Q3-2008Q3) versus boom period (2008Q4-2010Q4) Treatment: boom - pre-boom

0.0193

0.0201

0.0107

0.0028

Control: boom - pre-boom

0.0204

0.0012

0.0107

0.0092

DID

-0.0011

0.0189***

0.0000

-0.0064

ATT

-0.0029

0.0228**

0.0116

-0.0086

Number of treatment firms

283

283

282

495

Number of observations in matching

1413

1413

1413

1413

31

ACCEPTED MANUSCRIPT Table 5: Quarterly net equity issuance Panel A gives the quarterly net equity issuance for the whole sample and four firm treatment groups. It reports the average net equity issuance ratio in these samples before and during the credit boom. The groups of large firms and small firms are the firms in the top and bottom quintile of total assets in our data sample on 2008Q3.

T

State-owned group and non-state-owned group are the firms in the top quintile of state ownership and the firms

RI P

with zero state ownership at the end of the third quarter of 2008. We compare the differences of net equity issuance ratios before and during the credit boom in the whole sample and subsamples using a t-test.

SC

Panel B reports difference-in-difference results for the treatment groups and their control groups. The approach of Abadie et al. (2004) is employed to match the firms in these four treatment groups with non-treatment firms

NU

in the data sample on 2008Q3. The variables used in the matching process are the market-to-book ratio, cash flow, cash holdings, size, leverage ratio, state ownership, and industrial classification code, following the methods used in Almeida et al. (2011) and Kahle and Stulz (2013). DID is the traditional

MA

difference-in-difference estimator. ATT is the Abadie-Imbens bias-corrected average treated effect on the treatment group. ***, **, and * are 1%, 5%, and 10% significance levels using heteroskedasticity-consistent standard errors, respectively.

ED

Panel A: Quarterly net equity issuance before and during credit boom Whole

Large

State-owned

Small

Non-state-owned

(1)

(2)

(3)

(4)

(5)

Pre-boom (2006Q3-2008Q3)

0.0402

0.0231

0.0292

0.0862

0.0689

Boom period (2008Q4-2010Q4)

0.0137

0.0114

0.0123

0.0186

0.0184

-0.0169***

-0.0676***

-0.0504***

283

282

495

CE

PT

Period averages

Difference (Boom - Pre-boom)

AC

Number of firms

-0.0265*** -0.0117*** 1413

283

Panel B: The average changes of quarterly net equity issuance before and during credit boom Large

State-owned

Small

Non-state-owned

(1)

(2)

(3)

(4)

Pre-boom (2006Q3-2008Q3) versus boom period (2008Q4-2010Q4) Treatment: boom - pre-boom

-0.0117

-0.0169

-0.0676

-0.0504

Control: boom - pre-boom

-0.0122

-0.0194

-0.0183

-0.0168

DID

0.0006

0.0025

-0.0492***

-0.0336***

ATT

-0.0034

0.0033

-0.0374***

-0.0232***

Number of treatment firms

283

283

282

495

Number of observations in matching

1413

1413

1413

1413

32

ACCEPTED MANUSCRIPT Table 6: Quarterly corporate investment Panel A gives the quarterly corporate investment for the whole sample and four firm treatment groups. It reports the average corporate investment ratio in these samples before and during the credit boom. The groups of large firms and small firms are the firms in the top and bottom quintile of total assets in our data sample on 2008Q3.

T

State-owned group and non-state-owned group are the firms in the top quintile of state ownership and the firms

RI P

with zero state ownership at the end of the third quarter of 2008. We compare the differences of corporate investment ratios before and during the credit boom in the whole sample and subsamples using a t-test.

SC

Panel B reports difference-in-difference results for the treatment groups and their control groups. The approach of Abadie et al. (2004) is employed to match the firms in these four treatment groups with non-treatment firms

NU

in the data sample on 2008Q3. The variables used in the matching process are the market-to-book ratio, cash flow, cash holdings, size, leverage ratio, state ownership, and industrial classification code, following the methods used in Almeida et al. (2011) and Kahle and Stulz (2013). DID is the traditional

MA

difference-in-difference estimator. ATT is the Abadie-Imbens bias-corrected average treated effect on the treatment group. ***, **, and * are 1%, 5%, and 10% significance levels using heteroskedasticity-consistent standard errors, respectively.

ED

Panel A: Quarterly corporate investment before and during credit boom Whole

Large

State-owned

Small

Non-state-owned

(1)

(2)

(3)

(4)

(5)

Pre-boom (2006Q3-2008Q3)

0.0900

0.1025

0.0816

0.0875

0.1042

Boom period (2008Q4-2010Q4)

0.0920

0.1061

0.1053

0.0904

0.0992

0.0020

0.0036

0.0237***

0.0029

-0.0050

1389

279

279

271

483

CE

PT

Period averages

Difference (Boom - Pre-boom)

AC

Number of firms

Panel B: The average changes of quarterly corporate investment before and during credit boom Large

State-owned

Small

Non-state-owned

(1)

(2)

(3)

(4)

Pre-boom (2006Q3-2008Q3) versus boom period (2008Q4-2010Q4) Treatment: boom - pre-boom

0.0036

0.0237

0.0029

-0.0050

Control: boom - pre-boom

0.0102

-0.0197

-0.0048

0.0157

DID

-0.0066

0.0436***

0.0077

-0.0207**

ATT

-0.0063

0.0398***

0.0063

-0.0217**

Number of treatment firms

279

279

271

483

Number of observations in matching

1389

1389

1389

1389

33

ACCEPTED MANUSCRIPT Figure 1 Quarterly increase in M2 and bank loan

2000q1

2002q3

MA

0

NU

SC

100 Millions Yuan 20000 40000

RI P

T

60000

Quarterly Increase in Loan and M2

2005q1

2007q3

2012q3

Quarterly Bank Loan

ED

Quarterly M2

2010q1

PT

Figure 2 Book value of leverage ratios in treatment groups

AC

.4

.45

.5

.55

.6

CE

.65

Leverage Ratio in Treatment Groups

2006q2

2007q3

2008q4

Large Firm State-owned Firm

2010q1 Small Firm Non-state-owned Firm

34

2011q2

ACCEPTED MANUSCRIPT Appendix 1: Definitions of variables Variable code

Variable name and definition Corporate investment = (change in net fixed assets + depreciation) /

INV

T

beginning net fixed assets.

Book value of leverage ratio = total liability / total assets.

LOAN

Loan financing = proceeds from borrowings / total assets.

RI P

LEVERAGE

Net equity issuance = (proceeds from equity issuance - dividend payments) /

EQUITY

SC

total assets.

Market-to-book value = (market value of tradable shares + book value of

MTBV

NU

non-tradable shares + book value of liabilities) / book-value of total assets.

CF

Cash flow = (net income + depreciation) / lagged total assets.

CASH

Cash holdings = (cash and cash equivalents) / lagged total assets.

MA

STATESHARE State ownership = shares owned by state / total outstanding shares LNTA

Firm size = natural log value of book value of total assets. Large firm is an indicator variable equal to 1 if the firm size is in top two

LARGE

ED

deciles of all firm observations in 2008Q3, and 0 otherwise. State-owned firm is an indicator variable equal to 1 if the state ownership is

STATE

in top two deciles of all firm observations in 2008Q3, and 0 otherwise.

PT

Small firm is an indicator variable equal to 1 if the firm size is in bottom two

SMALL

deciles of all firm observations in 2008Q3, and 0 otherwise.

CE

Non-state-owned firm is an indicator variable equal to 1 if the state ownership is zero in 2008Q3, and 0 otherwise.

AC

NONSTATE

35