Does bank shareholding affect corporate debt restructuring? Evidence from financially distressed firms in China

ABSTRACT

This paper investigates the impact of bank shareholding on corporate debt restructuring. Using a sample of financially distressed firms in China from 2007 to 2016, we find that distressed firms with bank shareholders are more likely to restructure their debt than firms without bank shareholders. Moreover, the alleviation of renegotiation friction in both distressed state-owned enterprises (SOEs) and non-state-owned enterprises (non-SOEs) and reducing information asymmetry in non-SOEs facilitate the positive impacts of bank shareholding on debt restructuring. In addition, the impact is more evident in distressed non-SOEs characterized by higher profitability prior to restructuring. Further, distressed non-SOEs with bank shareholders are more likely to recover from distress than their peers, while the results are opposite for distressed SOEs. We argue that while bank shareholding facilitates restructuring in distressed non-SOEs, it aggravates the soft budget constraint in troubled SOEs. Our results are robust after accounting for the selection bias of bank shareholding.

KEYWORDS:

• Bank shareholding
• debt restructuring
• financial distress
• China

JEL CLASSIFICATION:

• G32
• G33
• G34

Disclosure statement

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

Notes

¹ See, e.g. Brunner and Krahnen (2008), Rauterkus (2009), Huang and Huang (2011), Demiroglu and James (2015), Micucci and Rossi (2017).

² The law even requires banks to dispose of shares acquired through a debt-for-equity swap, or from pledged shares, within one year after obtaining them. The Commercial Banking Law of 2003 changes the maximum holding period to two years.

³ Banks defined in this paper include state-owned commercial banks, joint-stock commercial banks, urban commercial banks, and rural credit cooperatives.

⁴ CSMAR and RESSET databases also provide data on corporate debt restructuring, but the number of restructuring is far less than the number collected from firms’ annual reports.

⁵ We also use a contemporary measure of bank shareholding, i.e. BSH_t, to replace BSH_t-2. Our conclusions sustain.

⁶ Instead of using interaction terms, we also divide the full sample (SOE sample and non-SOE sample) further into two subsamples based on the median value of these variables, and estimate all regression models separately in each subsample. Our results sustain.

⁷ In our sample, the median value of the number of banks involved in a default is two. Thus HdefaultN equals one when a firm has no less than three banks involved in a default.

⁸ Multiple creditors are common in our financial distress sample: a single creditor occurs in only 36.10% of the default events. The proportion of defaults with two (three) creditors is 19.87% (11.82%). And the proportion with three or more creditors reaches around 44.00%.

⁹ Although not reported, we find that firms in our financial distress sample have higher leverage, less cash holding, and lower profitability than total firms in the database.

¹⁰ Note that a restructuring proposal may include several methods, thus making the total number of restructuring methods in column (1) of Table 3 exceed the total number of firm-year observations with restructuring events occurring (i.e. 334).

¹¹ We also calculate the marginal effect at means. The conclusion is similar to the average marginal effect. In order to save space, we do not report it.