Institutional development and financing decisions: evidence from a cross-regional study on Chinese listed firms

Abstract

In this paper, we empirically investigate how differences in the development of legal and financial institutions across Chinese provinces and municipalities affect the financing decisions of Chinese listed firms. Our results indicate that a stronger regional enforcement of property rights reduces firms’ reliance on bank loans. Conversely, in regions with a larger government expropriation risk, firms raise more and shorter-term bank debt. Active regional bank lending positively impacts the debt ratio and the fraction of bank loans, but shortens loan maturities. The size of the local banking sector, the market capitalization as well as the liquidity of local stocks bear no relation with the capital structure. Overall, these relations do not depend upon the identity of the firm's controlling shareholder. Nonetheless, our results do suggest that state-controlled firms benefit from easier stock market access.

Keywords:

• capital structure
• bank loans
• debt maturity
• legal and financial institutions
• ownership

JEL Classification:

• G32
• G38
• K42

Acknowledgements

The authors thank Christian Andres, James Ang, Hong Bo, Katrien Craninckx, Joseph P.H. Fan, Tom Franck, Ann Gaeremynck, Alessandra Guariglia, Mathieu Luypaert, K.C. John Wei, Gunther Wuyts, Xinzhong Xu, Shu Yan, Ning Zhu, and participants in the Benelux Corporate Finance Day (Rotterdam), the China International Conference in Finance (Guangzhou), the Chinese Economic Association Conference (Dublin), the Journal of Corporate Finance Special Conference on Emerging Markets (Beijing), the European Journal of Finance Special Conference on the Chinese Capital Market (Durham) for their suggestions and comments on an earlier draft of this paper. Lihong Wang would like to thank the Fundamental Research Funds for the Central Universities in China (0155-ZK1005) for financial support.

Notes

1. Demirgüç-Kunt and Maksimovic (1999) construct a sample of firms from 30 countries, including both developing and developed ones. They show that the stage of development of legal and financial systems creates cross-country differences in financing choices, measured by firm reliance on long-term finance. These results have been confirmed by Giannetti (2003), De Jong, Kabir, and Nguyen (2008), and Fan, Titman, and Twite (2012), although it is clear from this research that institutional factors do not unconditionally exert the same influence on firms’ financing decisions. Moreover, De Jong, Kabir, and Nguyen (2008) also point out that a few firm-level variables are not consistently significant with the right sign in each of the 42 countries in their sample.

2. In contrast, Ayyagari, Demirgüç-Kunt, and Maksimovic (2010), using survey data on 2400 Chinese firms during 1999–2002, conclude that the security of property rights and the degree of perceived banking corruption are not related to a firm's reliance on bank loans. Like Li, Yue, and Zhao (2009), the sample studied by Ayyagari et al. is dominated by non-listed firms (97.1%).

3. The survey by Fan, Wang, and Zhu (2006), covering the period 2001–2005, is an extension of that developed by Fan and Wang (2004). The data collected by Fan and Wang (2004) cover the period 1998–1999 and 2001–2002; they have been used recently by Li, Yue, and Zhao (2009), Ayyagari, Demirgüç-Kunt, and Maksimovic (2010), among others. Yet, it is difficult to combine both sources in one study, as the survey questions and the corresponding scale used to calculate the various institutional characteristics differ substantially across the two surveys.

4. Although abolished in 1999, this quota system continued to govern equity markets until the end of 2002, as many companies that had been approved for listing were placed in a queue and released to the market only slowly over time.

5. At the end of 2005, the four state-owned banks, commonly known as the ‘big four’, still accounted for 52.5% of total assets in the Chinese banking sector. In contrast, the twelve joint stock banks and the more than 100 city commercial banks represented 15.5% and 5.4%, respectively, of the total assets. The remainder of total assets is spread among other banks, including policy banks, rural commercial banks, rural cooperative banks, foreign banks, urban credit cooperatives, and rural credit cooperatives (Almanac of Finance and Banking in China, 2006).

6. Interestingly, Cull and Xu (2005) point out a large cross-regional variation in the amount of collateral that private-controlled firms are required to provide as security for their bank loans.

7. Red tape generally includes the completion of seemingly unneeded paperwork, the requirement of unnecessary licenses, the approval of decisions by multiple people or committees, and various low-level rules that make conducting one's business slower, more difficult, or both.

8. We did collect the data on the Chinese firms becoming cross-listed on an international stock exchange, to investigate the changes in their capital structure variables as a result of this event. After listing abroad, firms experience a significant decline in their overall debt ratio, in their bank debt ratio, and in their fraction of bank loans that mature in one year. Next, we discern a significant negative correlation between the size of informal (non-tax) payments to local officials and the index proxying for the nature of bank lending in a region, and the change in the overall debt ratio of local firms. However, we do not observe significant correlations between institutional characteristics and changes in the bank debt ratio and in the maturity structure of bank loans.

9. These five fields include (1) the relationship between the government and the market, (2) the development of the non-state sector in the economy, (3) the development of the product market, (4) the development of the factor market, and (5) the development of market intermediaries and the legal environment.

10. We find that our results are robust when using the market debt ratio, which is calculated as the book value of total liabilities relative to the market value of equity plus the book value of total liabilities. However, one difficulty with this definition of market leverage is that a large fraction of shares in Chinese listed firms cannot be traded freely and thus do not have a market price. No consensus exists about how to compute the market value of a firm with a non-trivial fraction of non-tradable shares. Yet, as tradable and non-tradable shares have identical voting rights and cash flow rights, we used the price of the tradable A shares as a proxy. For preferred stock, representing only 0.2% of total shares outstanding, we also used the price of these common A shares to calculate its market value. Because of these reasons, we do not utilize market leverage as the main variable in our analyses.

11. For 655 out of 785 sample firms, we can calculate MTB_efw as of the IPO year; the MTB_efw calculation starts as of the second listing year onwards for the other 130 firms. Overall, the average (median) number of years used per firm to calculate MTB_efw equals 6.35 (6.58).

12. The results of this robustness check are not reported in Table 4. The outcome of all robustness checks that are discussed but not reported in the paper can be obtained from the authors upon request.

13. We indeed find that the parameter estimates and significance levels of the institutional variables are robust. Nonetheless, column 3 in Table 5 also reveals that the legal enforcement of property rights no longer meets the 10% significance level (p-value of 0.139). We attribute this outcome to the fairly high correlation between the regional rule of law index and regional GDP per capita, as shown in Table 3. Yet, as the parameter estimate of legal enforcement is not affected and as regional macroeconomic variables are not significant, we conclude that the effect of the regional rule of law index is not driven by regional economic development.

14. In line with Giannetti (2003), we also tried to include asset maturity, i.e. the ratio of long-term assets to total assets, in the bank debt maturity regression model, but found a correlation too high with asset tangibility (ρ=0.845).