China’s secondary privatization: new evidence on controlling shareholders tunnelling

ABSTRACT

This article utilizes the 2005 split-share structure reform (SSSR) in China as a natural experiment and conducts difference-in-differences (DID) tests to analyse corporate governance changes among Chinese SOEs compared to POEs. We show that tunnelling significantly reduced in both POEs and SOEs after the SSSR. More importantly, we find a significant and positive ‘privatization effect’ on SOEs’ tunnelling activities during the post-reform period suggesting the reductions of tunnelling were smaller among SOEs than POEs following the SSSR. In contrast, excess returns around the SSSR indicate that investors expected a negative ‘privatization effect’ on SOEs’ tunnelling. These findings suggest that the quality of corporate governance did not improve among SOEs as a result of the secondary privatization as the stock market expected without fundamental changes to firm ownership and control following the SSSR. The benefits of privatization accrue to the government controlling shareholders rather than minority investors.

KEYWORDS:

• Privatization
• tunnelling
• split-share structure reform
• corporate governance
• China

JEL CLASSIFICATION:

• G34
• G38
• G15

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ In contrast, Rousseau and Xiao (2008) and Huang and Wang (2011) document significant performance improvement after transfer of former SOEs’ control from the government to private investors in China.

² See Jiang, Lee and Yue (2010) and Li (2010) for regulations related to tunnelling through intercorporate loans and ‘other receivables’ as a vehicle for tunnelling in China.

³ According to Jiang et al. (2015), NOFO is an unambiguous measure of the expropriation of minority shareholders and has been widely used as a convenient way of tunnelling by the controlling shareholders in China. In 2006, the CSRC noticed the pervasiveness of this type of tunnelling practice and officially named it as a ‘Non-Operational Fund Occupancy’ problem.

⁴ See Liu and Tian (2012) and Liao, Liu and Wang (2014) for descriptions of the SSSR and transformations of share types.

⁵ Results using Logit regressions are very similar to the Probit regressions we report in the article.

⁶ The majority of firms completed SSSR prior to 2009. Only 3 firms reformed during 2009–2012.