External finance and trade credit extension in China: does political affiliation make a difference?

Abstract

Using a dataset of 65,706 Chinese firms over the period 2000–2007, we show that politically affiliated firms benefit from easier access to short-term external finance and extend more trade credit than their non-affiliated counterparts. Furthermore, we observe that the sensitivity of trade credit extension to short-term liabilities, which is largest for private firms producing differentiated goods, decreases with the degree of political affiliation. This suggests that gaining political affiliation contributes not only to alleviating individual firms’ financing constraints, but also to reducing the overall level of constraints in the economy through the additional trade credit being made available.

Keywords:

• trade credit extension
• political affiliation
• short-term liabilities
• Chinese firms

JEL Classification:

• G31
• K0
• P2

Notes

1. Hereafter, we will refer to accounts receivable, trade credit extended, and trade debit (TD) interchangeably.

2. According to our data (described in Section 4), over the period 2000–2007, the TD to assets ratios in China was 18.7%. This number compares favorably with the ratios ranging between 7% and 19%, reported by Bartholdy and Mateus (2008) across southern European countries. Similarly, according to our data, the TD to sales ratio is given by 17.2%, which compares favorably with the ratio of 17.1% reported by Bougheas, Mateut, and Mizen (2009) for the UK.

3. In addition, Wu, Rui, and Wu (2012) study the links between trade credit received and extended, and cash holdings, while Fabbri and Klapper (2011) find that firms that receive trade credit from their suppliers are also more likely to extend trade credit to their customers. Finally, Hale and Long (2011a, 2011b) show that private firms are able to manage their accounts receivable better than other firms.

4. See, among others, Li et al. (2008), Chen et al. (2011), and Xu, Xu, and Yuan (forthcoming), who studied the links between political connections and firm performance in the context of China.

5. Using UK, US, and Chinese data, respectively, Atanasova (2007), Molina and Preve (2009), and Fabbri and Klapper (2011) document that firms with better access to trade credit find it easier to finance receivables.

6. A vast literature originated by Fazzari, Hubbard, and Petersen (1988) has considered the sensitivity of fixed investment to cash flow as a proxy for the degree of financing constraints faced by firms (see Hubbard 1998; Bond and van Reenen 2007, for surveys). Yet, it has been shown that if investment opportunities are not properly accounted for in an investment regression, then a positive and significant investment-cash flow sensitivity could simply be observed because cash flow proxies for investment opportunities (Cummins, Oliner, and Hassett 2006). The measure of financing constraints that we propose does not suffer from these problems as investment opportunities are not as relevant to trade credit extension as to fixed investment.

7. Similarly, Cunat (2007) shows that when a firm uses a specialized product, the buyer and the seller enter a symbiotic relationship in which neither has the incentive to damage the trust that exists between the two.

8. Our data contain disaggregated information on bank loans, accounts payable, and other short-term liabilities only at the end of our sample period. For this reason, we are unable to differentiate the effects of these three components of total short-term liabilities on accounts receivables in our regressions. We do not believe this to be a serious shortcoming as our hypothesis H1 is focused on the links between accounts receivable and total short-term external finance. For those years in which detailed information is available, it is noteworthy that bank loans constitute on average 70% of total short-term liabilities.

9. A negative relationship between trade credit and inventories is also consistent with Daripa and Nilsen (2011), according to whom sellers subsidize the shift of inventories to buyers.

10. Our results were robust to replacing the time dummies interacted with industry dummies with time dummies and two variables describing the structure of the market in which the firm operates, i.e. the sales share of the eight largest firms in the firm's two-digit industry, and the ratio of the firm's sales relative to its two-digit industry total sales.

11. All our regressions are performed in Stata using the command xtabond2 developed by Roodman (2009).

12. In those cases in which we found evidence of higher-order serial correlation, we used four lags of the regressors as instruments, and reported the m4 test in the tables.

13. See the Appendix for details about the structure of our panel and for definitions of all variables used.

14. It should be noted that, according to our classification scheme, foreign-owned firms include firms owned by Hong-Kong, Macao, and Taiwan agents, as well as agents from other foreign countries. Privately owned firms include firms owned by legal entities and individuals. A similar classification was used in Guariglia, Liu, and Song (2011). Instead of using information on the fraction of capital paid in by different agents to define ownership, we could have used registration codes. However, registration codes are not entirely reliable as they are updated only with considerable delay (Dollar and Wei 2007). Moreover, firms might have an incentive to falsely register as foreign simply to take advantage of the tax benefits accorded to the latter. Defining ownership categories based on majority average ownership also has the advantage of minimizing the effects of measurement error in the ownership variables which can affect individual years.

15. See the Appendix for details. Other measures of political affiliation have been used in the Chinese context and related to firm performance. For instance, Chen et al. (2011) and Xu, Xu, and Yuan (forthcoming) classify firms as politically connected if their chairperson or CEO is a current or previous government official. Li et al. (2008) consider party membership as another form of political connection. As these measures are not available in our dataset, we were unable to use them in this particular paper.

16. Like Bougheas, Mateut, and Mizen (2009), Atanasova (2012) also uses data from the FAME database and reports an average accounts receivable to sales ratio of 15.1%. The different figures obtained in the two studies are probably due to differences in the sample period.

17. Cull, Xu, and Zhu (2009) group firms into five ownership categories: state-owned, collective, legal-person, domestic private, and foreign. We include legal entities into the private group as in Guariglia, Liu, and Song (2011). Also see the Appendix and footnote 14.

18. It should be noted that 96.6% of SOEs has some form of political affiliation. It is also interesting to observe that the accounts receivable to sales ratio for SOEs with high political affiliation is given by 25.5%.

19. Cull, Xu, and Zhu (2009) use the ratio of interest payments to sales as a proxy for bank finance, since just like us, they cannot identify the amount of bank borrowing from total short-term liabilities. They explain trade credit extension with firm ownership, profitability, their bank finance proxy, and possible interactions between these three variables. We prefer to use the short-term liabilities to sales ratio instead, as the interest payments to sales ratio may capture both interest rate and amount of loan variation across categories of firms and over time. However, to facilitate comparison, we experimented with the use of the interest payments to sales ratio instead of the short-term liabilities to sales ratio. The results, which are available upon request, were similar to our main findings.

20. One could question whether the differences in the coefficients on short-term liabilities for firms operating in differentiated and standardized sectors are statistically significant. In order to test for this, we included interactions of all variables in the model with a dummy equal to 1 for firms selling differentiated goods, and 0 otherwise, in a single regression estimated on the full sample of firms within each ownership group. We found that within each group, both the Stliabs term and its interaction with the dummy were positive and precisely determined, suggesting that the difference in the sensitivities of accounts receivable to short-term liabilities across firms producing standardized and differentiated goods is indeed statistically significant. These results are available upon request.

21. Because our model is estimated in first-differences, we are unable to verify whether the characteristics of the goods produced by the firm have a direct impact on trade credit extension. These characteristics, which are subsumed in the α_i term in Equation (1), are in fact time-invariant and disappear in the first-differencing process. The same argument applies to firms’ ownership and political affiliation.

22. Elasticities are calculated using the following formula: (coefficient on Stliabs*mean value of Stliabs)/mean value of AR.

23. It is also possible that the causality runs the other way, i.e. that firms which extend more trade credit end up having a lower liquidity, simply because they accumulate less cash.

24. The negative age coefficient for private firms producing differentiated goods could be explained by the lack of financial flexibility characterizing older firms (Ding, Guariglia, and Knight, forthcoming).

25. The insignificant coefficient for short-term liabilities in the case of state-owned firms without political affiliation in both differentiated and standardized industries may be explained by the small sample characterizing these two categories.

26. We tested whether the differences in the coefficients on short-term liabilities for affiliated and non-affiliated firms are statistically significant. To this end, we included interactions of the PA dummy with all variables in the model in a single regression estimated on the full sample of firms within each industrial sector and ownership group (with the exception of state-owned firms, as only a handful of them are not politically engaged). We found that, in the differentiated sector, the interaction term between Stliabs and the dummy was negative and statistically significant for private and foreign firms, but not for collective firms. These results are available upon request.

27. For robustness, we classify our firms into ownership categories according to the 100% rule. According to this rule, a firm is classified as, for instance, private if it was 100% privately owned throughout the sample period. Classifying firms on the basis of a 50% rule (which includes firms transiting from one ownership group to another) or a 100% rule (which excludes them) delivers very similar results. This suggests that transitions do not have a significant effect on our main findings.

28. We thank an anonymous referee for suggesting this exercise. Wu, Rui, and Wu (2012) use the ratio of total bank loans to GDP of the province to measure financial deepening.

29. Our way of classifying firms into ownership groups excludes from our sample firms with mixed ownership in which no group has a majority share. For instance, a firm characterized by 40% private ownership, 30% state ownership, and 30% foreign ownership would be excluded. Firms of this type of mixed ownership make up less than 4% of our sample.

30. Within this category, firms owned by individuals represent approximately 60% of the total. As firms owned by legal entities include firms owned by state legal entities, one could question their inclusion in the private category. One reason for including them is that while the state's primary interest is mainly political (i.e. aimed at maintaining employment levels or control over certain strategic industries), legal entities are profit-oriented (Wei, Xie, and Zhang 2005). Since our dataset does not allow us to discriminate between state and non-state legal entities, we were unable to exclude the former from our private category.