中文摘要：

随着国际资本流动日趋自由化,越来越多的企业实行交叉上市,对交叉上市的研究显得尤为重要。捆绑假说是解释企业交叉上市动因的重要理论,该假说认为企业在外部监管较为严格的地区交叉上市,可以约束控股股东和管理层谋取私有收益,抑制公司盈余管理,从而提高公司盈余质量。然而,国内外学者以不同数据和方法对捆绑假说加以研究,并未得到一致结论。部分研究表明,交叉上市企业盈余管理水平低于非交叉上市企业,支持捆绑假说,而其他研究认为交叉上市企业盈余质量并未显著高于非交叉上市企业。而且,已有研究大多仅考虑应计盈余管理,忽略了真实盈余管理。真实盈余管理由于其隐蔽性强、会计弹性大等特点而被越来越多的企业采用。交叉上市企业在面临更为严格的外部环境时会减少应计盈余管理,增加真实盈余管理。基于捆绑假说视角,探讨交叉上市企业盈余管理方式的转变。考虑应计盈余管理和真实盈余管理两种方式,采用修正的Jones模型和Roychowdhury模型分别估算应计盈余管理和真实盈余管理水平,以2007年至2014年中国发行A＋H股、A＋B股、A＋H＋N股的交叉上市企业和纯A股上市企业数据为样本,通过控制年度和行业的OLS回归方法检验交叉上市企业盈余管理方式的转变,并进一步研究交叉上市企业的上市地对其盈余管理的影响。研究结果表明,交叉上市企业倾向于减少应计盈余管理,但其真实盈余管理水平显著高于纯发行A股的上市企业;发行A＋H股的上市企业应计盈余管理和真实盈余管理水平均与发行A＋B股的上市企业无显著差异;发行A＋H＋N股的上市企业其应计盈余管理水平与发行A＋H股、A＋B股的上市企业无显著差异,但其真实盈余管理水平显著低于这两种上市企业。研究结果有助于增强对交叉上市企业盈余质量的理解,也为投资者的投?

英文摘要：

With the increasing liberalization of international capital flows, a growing number of companies choose to cross-list abroad. Thus, the studies on cross-listing are particularly important. Bonding hypothesis, proposed by scholars, is an important theory to explain the cross-listing motivation. Bonding hypothesis suggests that cross-listing in a location with stringent external environment can constrain the controlling shareholders and management from seeking private gains, inhibit earnings management, and improve a company＇s earnings quality. However, domestic and international studies on bonding hypothesis using different data and methods did not reach a unanimous conclusion. For instance, there are two opposing views about the relationship between cross-listing and earnings management in China. Some studies show that earnings management in cross-listed companies is less than that in non-cross-listed companies. This supports the bonding hypothesis. Other studies show that the earnings quality of cross-listed companies is not significantly better than that of non-crosslisted companies. Furthermore, most of the studies mentioned above only considered accruals earnings management, ignoring real earnings management. Real earnings management is increasingly widespread because of its strong concealment, accounting flexibility, and so on. We argue that cross-listed companies facing a more stringent external environment prefer to reduce accruals earnings management and increase real earnings management. The study explores the earnings management of cross-listed companies from the perspective of bonding hypothesis. The study uses an OLS regression to empirically investigate earnings management among companies that are cross-listed in A ＋ H shares, A ＋ B shares, A ＋ H-N shares and those listed only in A shares. The study discusses the influence of cross-listing location on earnings management. Considering both accruals and real earnings management, we adopted the modified Jones and Roychowdhury models to estima