【作者】 蒋基路；

【导师】 王华；

【作者基本信息】 暨南大学 ， 会计学， 2009， 博士

【摘要】 本文以澄清资产公允计量的判断尺度,并明晰特定应计项的盈余管理度量开篇,在新制度经济学的系统理论框架下,从盈余管理的内部动因和外部诱因两条线索展开分析,考察坏账信息中的盈余管理证据。企业的委托-代理契约性质是盈余管理的内部动因,决定了盈余被操控的幅度；信息不对称的产品市场、借贷市场、股票市场、以及控制权市场中广泛存在的私人控制权利益,是盈余管理的外部诱因,决定了盈余被调节的方向。坏账信息这一特定应计项上的盈余管理证据,使本文得以在新制度经济学的框架之下,搭建起一个连通企业、市场与政府的盈余管理研究完整结构。追溯盈余管理计量研究的发展历程,研究者所考察应计项的演变,启示本文对关键变量的设计与筛选,以及检验程序的安排,奉行将偏误降至最低的主旨。本文参考McNichols和Wilson (1988)对坏账准备计提的研究(以下简称MW1988),提出一个刻画特定应计项操控性部分的代理变量(Resprov),该代理变量的无偏性藉由总体计量得以佐证,从而增加了动因检验和诱因检验的信心。拆解国内经验证据进行统计分析,本文初略描绘了近十年来本土盈余管理实证研究的全景,海外与本土经验证据的差异说明,不考虑治理环境差异而移植海外研究的代理变量表征主题,将造成检验结果的水土不服。此外,本土研究中的传统减值变量忽略了盈余的期间特性,因而在表征盈余管理方面可能存在偏误。实证结果显示：对中国上市公司样本期间(2002-2005年)坏账准备计提的综合评价为“少提减值”,或者说,“计提不足”。除了巨亏公司大幅超额计提进行清洗和高盈利公司小幅超额计提进行平滑以外,大部分公司倾向于少提或转回坏账以调高盈余。(1)在特定应计项的操控幅度方面,有可能发挥制约作用的公司内部治理机制包括：集中的股权结构,机构投资者参与公司治理,董事会中审计委员会的设立,以及地区治理环境中更高的市场化水平；而国有产权属性、股权制衡、以及独立董事制度,没有取得制约特定应计项操控幅度的证据,转型经济体的制度背景对此做出了最好的注解,在内部人转移或侵占上市公司资源的“掏空”行为,制衡股权引发的控制权利益激烈争夺,以及独立董事“独力”“懂事”的无奈选择背后,是产权界分根深蒂固的一股独大现状,政府的角色错位,内部人控制的同仇敌忾,法律监督和投资者保护机制的缺失,以及上市公司信息披露的不透明,这就注定了某些企业制度对于解决代理问题的无能为力。(2)在特定应计项的调节方向方面,股票市场牟利的检验结果比产品市场、借贷市场、和控制权市场都更为理想。综合股票市场牟利来看,在特定应计项上管理盈余的两类中国上市公司,其一是高盈利区间的绩优公司,其二是处于被ST边缘的微利公司和清洗公司,相反,注定难逃ST命运的公司,反而没有表现出明显的对特定应计项的操控。产品市场牟利证明了政治成本在解释特定应计项的调节方向上依然适用,而借贷市场中以会计信息为契约基础的牟利,也依然可以应用债务契约来解释,与其他三个市场相比,控制权市场并非基于经营业绩的薪酬契约与职位任免,导致牟利与会计信息的联系自行断裂,缺乏特定应计项盈余管理的证据,因而最可能在本土经验证据上遭遇滑铁卢的,就是分红计划。廓清新制度经济学的整体框架,是希望能够为盈余管理的治理提供理论指导,以确定企业制度与市场制度安排的偏向和优先次序。产权明晰的企业制度与信息充分披露的市场制度是治理的终极目标,二者都可以降低交易费用,然而盈余管理问题超出了企业制度的解决范围,无论最终控制人的性质是国有、民营还是外资,无论股权集中还是分散,无论股权制衡程度如何,无论董事会的作用如何,只要市场牟利的诱因存在,管理者都可能采取管理盈余、操纵盈余、甚至会计舞弊的机会主义行为,其后果都会影响资源配置的效率和误导投资者而造成社会财富损失,带来不确定的、可能是巨额的交易费用,这就是科斯(1960)在《社会成本问题》中所言的“当交易费用巨大时,即使明晰的产权也无济于事”,正如市场牟利诱因是盈余管理的根源,市场制度也是解决盈余管理问题的根本,为了减少内部人“说谎”的机会主义行为,政府的职责是辅助(而不是取代)市场建立传递或甄别信号的激励机制和约束机制,只有提高市场识别“谎言”的能力,才能使盈余管理的私人牟利动机失效,这就是法制、或者法制缺乏下政府监管的重心和方向。更多还原

【Abstract】 The paper started with the judgement yardstick to measure assets’fair value and earnings management on specific accruals without bias. Under New Institutional Economics theories’framework, the paper inspects the earnings management evidence on bad debts information along two clues, one is internal motivation and the other is external causality. The agent contract characteristic of modern corporations is the internal motivation, affecting the extent of discretionary reported earnings, while the private controlling benefits in markets of products, credits, stocks, and managers, are the attractive causality, determining the derection of adjustments towards target income. The evidence on bad debts helped to construct an integrated structure connecting corporations, markets and governments together and make explaination for the behavior of managers.Tracing back to the history of earnings management measuring, the developments of total accruals studied by researchers inspired us to pursue the substance of minimized bias in designing and selecting key variables, and arranging test process as well. Following McNichols & Wilson (1988) model on bad debts provision, we developed a proxy named Resprov for discretionary part of specific accruals, and verified its unbias through overall test by modified Jones Model (1995), thus increased our confidence in this proxy during the following motivation test and causality test. Breaking the test results in domestic top journals into different topics for statistics and analysis, the paper tentatively sketched a whole scene portrait of local earnings management studies for recent ten years. The differenc between oversea and local evidence illustrated that ignoring the eviorment disparity, and transplant west proxy to capture research subjects without considering institutions background, will lead to the failure of tests and unexplainable results. The paper also found that traditional proxy in local writedown studies neglected the period characteristcs of earnings and has the suspicion of bias. The paper shows, the synthetical appraisal for bad debts information provided by Chinese listed companies is insufficient provision. Except for large loss companies overfulfilled to take big bath, and large profit companies slightly exceeded to smooth, most companies have the tendency of providing less allowance for bad debts. (1) Under the clue of motivation, the coporate mechanisms which might restrict the extend of earnings management, are concentrated stock equity structure, institutional investors participation, audit committee establishment in board, and developed market process in region enviorment, while state-owned property, confrontationed equity structure, and outside directors failed to get evidence of restraining earnings manangement on specific accruals. The coporate governance enviorment in a transformed economic entity has made best annotation for this result. (2) Under the clue of causality, stock markets has get more idealed testing results than other markets. Using specific accruals to mange reported earings is more common in large profit companies to smoothing income into future periods, and companies in the edge of ST range, both above and beneath the scope, no matter in what period does their target income lies, the insufficient provision or exceeded provision reflected their desire to avoid the destiny of being ST by regulartors. In the opposite, large loss companies which doomed to be ST didn’t express apparent manipulation on specific accruals. Because accounting information’s role as contract foundamental basis is different in China markets, thus connection between private controlling benefits in certain markets and earnings management has voluntarily break up, such as managers’ compensation of Positive Accounting Hypothesis in controlers market.Clarifing the framework of New Instituional Economic theories as guidance in earnings management studies, the paper expects to provide direction in governing earnings management before it lead to great damage of society’s wheath. The regulators’ responsibility is to help (but not replace) the markets to build up information delivering and examining machanisms, thus increase markets’ability to distinguish the "lies".更多还原