World financial crisis and the rise of Chinese commercial banks: an efficiency analysis using DEA

Abstract

The current financial crisis has hit the banking giants of the world really hard. It is striking to note that some of the large Chinese commercial banks have emerged to be the biggest winners as a result of the crisis, thanks to reforms over the past 10 years. The most significant reform before the crisis was ownership diversification, aiming to improve corporate governance and efficiency. Within 1 year from October 2005, three of the four biggest State-Owned Banks (SOBs) were listed on the stock exchanges. This article will study whether this reform has really improved bank efficiency. Adopting the Data Envelopment Analysis (DEA) approach, this article examines whether Initial Public Offering (IPO) is effective in enhancing bank performance. Using data of 14 listed banks during 1999 to 2008, the results show that on average, bank efficiency increased by almost 5% after listing. Despite the fact that Joint Equity Banks (JEBs) still perform better than SOBs, the latter manage to catch up and reduce the efficiency gap with the former during the past few years. This in part explains why the Chinese banking system has been less affected by the current world financial crisis than their western counterparts, leading to an important conclusion that SOB reforms in China over the past 10 years have produced remarkable results.

Acknowledgements

The research was supported by the Leverhulme Trust (F/00765/A) and the EU–China Financial Cooperation Project (EuropeAid 112901/C/SV/CN).

Notes

¹ See Appendix A for all the listed Chinese commercial banks until the end of 2007.

² For a detailed proof, see Cooper et al. (2006, Appendix A.4). The ‘Duality Theorem’ suggests that: ‘(i) In a primal-dual pair of linear programs, if either the primal or the dual has an optimal solution, then the other one does also, and the two optimal objective values are equal; (ii) If either the primal or the dual problem has an unbounded solution, then the other has no feasible solution; (iii) If either problem has no solution then the other problem either has no solution or its solution is unbounded.’

³ When running DEA, the analyst is often concerned with the nature of returns to scale that would best reflect the operations of the DMUs in the sample. CRS implies that outputs would increase proportionately to additional inputs. Conversely, Variable Returns to Scale (VRS) means a disproportionate rise or fall in outputs when inputs have increased. Normally, it has two types, Increase Returns to Scale (IRS) or Decrease Returns to Scale (DRS).

⁴ Data of Ningbo Bank in 1999 are unavailable.

⁵ Scale efficiency loss for BOC can be estimated as: TE = PTE × SE, under CRS assumption, before IPO, SE = TE/PTE = 0.8923/0.9739 = 0.9162; after IPO, SE′ = TE′/PTE′ = 0.9325/1.0415 = 0.8953. Therefore, scale efficiency has decreased by 0.9162 – 0.8953 = 0.021 or 2.1%.

⁶ Improvement of scale economy for Ningbo bank could also be proved as: TE = PTE × SE, under CRS assumption, before IPO, SE = TE/PTE = 0.9421/0.9841 = 0.9573; after IPO, SE′ = TE′/PTE′ = 0.9698/1.0018 = 0.9681. Therefore, scale efficiency has been improved by 0.9681–0.9573 = 0.011 or 1.1%.

⁷ For ownership, ‘0’ represents nonstate owned banks and ‘1’ represents SOBs and for IPO, ‘1’ means that bank has already listed on the stock exchanges while ‘0’ stands for the contrary.