Earnings management and listing regulations in China

Abstract

Purpose – The purpose of this paper is to document earnings management of Chinese firms. Design/methodology/approach – The paper takes advantage of the introduction of stringent delisting requirements around 2000 that non-cross-listed firms with consecutive earnings losses for more than two years would be delisted from the mainland Chinese exchanges. The paper examines whether listed firms in Chinese market manage earnings to avoid listings. The paper also examines whether mainland Chinese firms cross-listed in Hong Kong exchanges manage earnings the same way. The measure for earnings management is derived from a kernel density estimate for the return on equity distribution, following Bollen and Pool (2009). Findings – The paper finds that the new delisting threats induce rampant earnings management on mainland markets, and cross-listing in Hong Kong has a curbing effect on earnings management. The paper also finds that prices became less value relevant after the implementation of delisting regulations, and investors rationally discounted the reliability of earnings announcements in China. Such market responses were absent for cross-listed firms in Hong Kong. Originality/value – There is little conclusive evidence about whether cross-listing in a non-US market has a curbing effect on earnings management. The paper contributes to this literature by using this unique exogenous policy change in China and following a difference-in-difference approach in identifying the potential curbing effect. The particular measure adapted from Bollen and Pool (2009) utilizes information of the whole distribution of return on equity, thus extends earlier crude comparison of nearest two bars around zero and partially deals with the potential endogeneity problem.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 12 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.