Efficiency of trade credit and bank finance: an ethnic minority area in China

Abstract

Using 1998–2008 firm-level microdata for industrial firms, we investigate the efficiency of financial intermediation through trade credit and bank loans in an ethnic minority area in China, the Xinjiang Uygur Autonomous Region. We find: (1) after receiving trade credit, ethnic minority firms tend not to repay it when they are financially distressed; (2) bank finance allocates more funds to more efficient ethnic minority firms while also allowing state-holding Han firms with worse performance levels to access bank loans; (3) efficient financial intermediation to ethnic minority firms is achieved through bank loans to relatively large firms.

Keywords:

• Trade credit
• bank finance
• ethnic minority firms
• efficiency of financial intermediation
• China

JEL CLASSIFICATION NUMBERS:

• G32
• O16
• O53
• P34

Notes

Acknowledgements

We are grateful for helpful comments from participants of all conferences and seminars in the process of writing this paper and anonymous referees. All views and errors remain our own.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The data of industrial firms in Tibet has too small a sample size because the industrial sector in this area is quite underdeveloped. Furthermore, there are almost no ethnic minority (Tibetan) firms in the data. This is why we do not take Tibet as our case.

2 We also obtain ethnic information for firms from the list of members of the People’s Congress as explained in Section 3.

3 Due to space limitations, the estimation results for μ_j, μ_jt and μ_c are not reported. μ_i cannot be estimated due to the nature of the system generalized method of moments estimation technique that we adopt.

4 As mentioned above, Firth et al. (2009) examine how banks allocate loans to private Chinese firms, which are a segment of non-state-owned firms, using cross-sectional firm-level microdata from 2002. It can be safely stated that our study tries to extend theirs in that we focus on trade credit in addition to bank finance in determining efficiency, use panel data rather than 1-year cross-sectional, and adopt a GMM estimator for panel data. The latter data and econometric coping strategy contribute to settling possible endogeneity problems, as explained in the text.

5 The remaining endogeneity problem of the independent variables is typically caused by reverse causality from dependent to independent variables and omitting an independent variable in the error term e_it.

6 To roll them up, the sum of the coefficients of Performance *D_{state-holding} and Performance *D_minority captures the responses of the dependent for the state-holding ethnic minority firms whereas that of Performance *D_non-state and Performance *D_minority express them for the non-state-holding ethnic minority firms. However, our sample includes very few state-holding ethnic minority firms.

7 See Roodman (2008) for the system GMM estimation.

8 We use two-step GMM instead of one-step GMM since the former is asymptotically more efficient. Thus, we apply the Windmeijer (2005) finite sample correction to the two-step covariance matrix to settle the potentially downward-biased two-step standard errors.

9 In our unbalanced panel, the number of time periods of the available data is only 2–3 (years) on average. Our data are typical small time series and large cross section panels. One can safely conclude that our panel data tend not to suffer the problem of instrument proliferation as a result of their nature (Roodman 2008).

Additional information

Funding

This work was funded by Scientific Research (C) No. 19K01633 and Scientific Research (C) No. 17K03683 from the Ministry of Education, Science, Sports, and Culture of Japan (MESSC), The Murata Science Foundation, and The Mitsubishi Foundation No. 30230.