【作者】 陈瑶；

【导师】 罗荣华；

【作者基本信息】 西南财经大学 ， 金融学， 2013， 硕士

【摘要】 随着企业之间横向、纵向的兼并活动越来越频繁,跨行业、跨区域经营逐渐成为企业未来发展的一种趋势。企业的扩大不仅伴随着人力资源的扩充和厂房设备的扩建,更重要的是融资需求的增加。然而现阶段我国资本市场的局限性限制了企业通过债券融资或其他直接融资渠道来获得经营所需的巨额资金,而传统的单个银行一对一的贷款模式越来越无法满足企业庞大的资金需求,在这种情况下,多家银行联合贷款模式以及银团贷款模式应运而生并且得到迅速发展。自2005年-2011年期间,在同一天同时向多家银行贷款的企业数目占到进行银行贷款总企业数目的17.9%,其贷款金额更是占到了市场总银行贷款金额的49.8%。在企业与银行的借贷关系中,银行与企业即相互依赖又相互促进。一方面,银行会通过放宽或紧缩信贷资金影响企业的经营发展：在资金链出现危机时,企业通过银行贷款渡过难关或者利用杠杆效应增加公司收益。另一方面,借款企业经营状况的好坏又会对银行的发展产生较大的影响：信贷资金能否收回以及资金收益的高低都取决于企业的经营绩效以及未来发展状况。所以,银行不仅在授信前会对借款公司进行充分的资信调查,授信中对其信贷条件严格审核,而且在授信后也对企业的经营活动以及资金的使用情况进行持续的监督,从而发挥银行作为“大贷款人”的监督作用。在传统的单一银行贷款模式下,银行与公司之间的信息不对称是制约银行监督效率的关键因素,而在联合贷款模式(多家银行同时向一家公司提供贷款)下,参与贷款的银行数目众多,银行与银行之间的信息不对称问题也严重影响着银行体系的监督效率,并且参与者之间的相互作用及产生后果均具有更大的不确定性,使得银行体系的监督效率更加难以确认与衡量。因此当外部投资者试图通过公司的借款行为获取公司经营、财务、治理等方面的相关信息时,联合贷款行为传递出来的无论是信息含量还是信息质量都会在很大程度上不同于单一银行贷款模式下公司的借贷行为所传递出的信息。公司发布银行贷款公告后会引起公司股价变化,这一现象很早就受到了国外学者们的关注与讨论,通过考察公告后的市场反应来判断银行的监督效率也是学者们在研究银行是否具有“大债权人”监督作用时普遍采用的一种研究方法。就国内的研究而言,胡奕明、谢诗蕾、邓莉和张宗益等人都表示我国银行对上市公司是存在一定的监督作用的。但联合授信模式下银行的监督效率是否不同于传统单一银行贷款模式下银行的监督效率,国内对此问题的研究尚属空白,更别说通过考察贷款模式对市场超额收益率的影响来判断银行的监督效率问题了。由上文的分析可知,不同的贷款模式下参与者之间的作用机制不同因而会向市场传递出不同的公司信息进而引起公司超额收益率的不同变化。究竟联合授信模式下银行之间的作用关系是合作还是竞争?银行之间的这种相互作用是否会影响银行的监督效率,以及影响程度如何?这是本文的研究重点。首先,本文以我国上市公司在2005年至2011年间发布的银行贷款公告为研究对象,遵循一定的原则筛选出313个有效样本,其中包括56个联合贷款样本和257个单一贷款样本。接着本文运用事件研究法对银行贷款公告所引起的市场反应进行实证研究,并利用参数以及非参数的统计方法对总体样本的市场超额收益率进行显著性检验。结果发现,市场对我国上市公司发布银行贷款公告这一事件产生了强烈的反应,具体表现为相应的事件窗内,上市公司的股价出现了显著异于0的平均超额收益率和累积超额收益率,并且市场的整体反应为负,说明我国的投资者在短期内把公司的银行贷款行为视为一个“利空”的事件,这与国外的相关实证研究结果有所出入。另外,在本文的研究中,银行发布贷款公告这一事件所产生的公司股价异常收益率基本上出现在公告日的前第7天左右,且通过了显著性检验,相关证据表明该贷款信息很可能已提前泄露,我国市场中存在“信息泄漏”问题。此外,为考察两种不同模式的贷款公告引起的公司超额收益率是否具有显著差异,本文将总体公司样本按照贷款模式分为了联合贷款子样本和单一贷款子样本,并比较分析了两个子样本在不同事件窗口下的平均超额收益和累计超额收益率,结果发现：首先在整个事件窗内,两个样本的累计超额收益都呈现明显的下降趋势,并且联合贷款子样本的累计超额收益的下降幅度要大于单一贷款累计超额收益的下降幅度,这在一定程度上说明相对于单一银行贷款行为市场投资者似乎更加不认可公司的联合银行贷款行为。此外,本文还对两个子样本的累计超额收益率在不同事件窗口内的差异性进行了显著性T检验和符号秩检验,统计结果显示,在部分事件窗内(尤其是公告前后[-1,2]时间段),两个子样本的累计超额收益率的差异性在5%的置信水平上通过了显著性检验,从而初步验证了不同模式的贷款公告引起的市场反应的确有所不同。其次,在得到上市公司银行贷款公告的超额收益率后,本文在控制了公司特质、贷款特征以及外部环境相关变量的基础上,重点研究了贷款模式对累计超额收益率的影响。通过将超额收益率对贷款模式及其他控制变量进行多元线性回归分析,意在找出不同贷款模式下银行具有不同监督效率的确切证据。此外,为进一步研究联合授信模式下影响银行监督效率的因素,本文在联合贷款公司子样本下,考察了联合贷款特征变量—牵头行的国有性质,牵头行承担的贷款金额占比对超额收益率的影响,从而进一步揭示了联合授信模式下各银行之间的作用关系及其对银行监督效率的影响。具体分析过程为：联合授信模式下的借贷关系本质上是一种共同代理关系,根据共同代理理论,银行之间的合作与共谋能发挥信息的规模优势,获得共同代理关系下的规模经济,使得银行更加有效的发挥其“大贷款人”的监督作用。但基于以下几个方面的原因,参与贷款的银行之间难以达成合作,一方面,银行之间的信息不对称程度不同以及不同的公司业绩评价标准限制了银行之间的合作,另一方面,无论是在授信前还是在授信后银行都会有强烈的“搭便车”动机。在授信前,银行自身实力的限制、地域的限制、公司的业务往来历史、其他银行的声誉等因素都会增加银行收集公司信息的成本,当银行对此成本无法负担或者银行放贷后的收益无法弥补此成本时,银行就会有强烈的动机“搭便车”。在授信后,银行监督活动的外部性也会导致银行的“搭便车”行为。最后,某些银行持有“大树底下好乘凉”的心态加上联合授信中的“羊群效应”会使得银行在对公司的放款上产生激烈竞争,银行之间的竞争性不仅削弱了公司向其提供相关信息的主动性,也会从心理上提升公司的谈判地位,强化公司进行信息操作、数据作假、谎报实情的动机。可见,在联合贷款模式下,银行之间的不合作不仅无助于改善银行与公司之间的信息不对称问题,还可能导致银行贷后监督活动的缺失、对公司的竞相放款以及公司的信息欺诈行为,产生额外的共同代理成本,降低银行的监督效率。因此,发布联合贷款公告后的公司其市场超额收益会更低。此外,在联合贷款模式下,国有银行因其雄厚的资金实力、专业的人才队伍、众多的营业网点以及完善的信息共享系统等原因,在对借款公司的贷前调查、信贷审核以及贷后监督方面比其他非国有银行更具优势,加之部分股份制银行进行贷前调查的能力有限或者成本过高,在决定是否放贷时具有强烈的动机“搭便车”,更有甚者,部分银行会看在国有银行的“面子”上,盲目对公司贷款而忽视了对公司真实风险的防范与控制。此外,牵头行承担较大的贷款份额也会导致其他银行的“搭便车”和竞争行为,并且公司的透明度与牵头行承担的贷款金额占比成反比,占比越高,公司透明度越低。因此,若牵头行是国有银行或承担了较大比例的贷款金额,联合授信模式下银行之间的“搭便车”以及竞争放款问题会更严重,产生的共同代理成本更高,其公告发布后公司的超额收益也相对更低。在上文的假设前提下,本文采用虚拟变量代表银行贷款模式并对事件窗内的累计超额收益进行回归分析,结果显示相比单一贷款,联合贷款公告发布后公司会获得较低的超额收益。为进一步找出联合贷款模式下影响超额收益率的因素,本文又将总样本公司分为了单一贷款子样本和联合贷款子样本,并在联合贷款子样本下,考察了牵头行的国有性质以及牵头行承担的贷款金额占比对公司超额收益的影响。结果显示,当牵头行的是国有银行或者承担了较大比例的贷款金额时,事件窗内的累计超额收益率会相对更低,相关证据均表明市场投资者认为在联合贷款模式下银行之间存在着“搭便车”以及竞争行为从而产生了共同代理成本,削弱了银行的监督效率。最后本文的创新之处在于：国内现有的研究联合贷款模式下银行监督效率问题的文献几乎没有,本文通过考察股票市场对不同贷款模式的不同反应反映联合授信模式的效率问题,并在此基础上进一步研究了联合授信模式下牵头行性质、承担的贷款金额占比对公司超额收益率的影响,试图找出影响银行监督效率的某些因素,据此判断联合授信模式下银行之间的作用关系。更多还原

【Abstract】 As acquisitions and mergers become more frequent, companies are expanded not only along with the expansion of human resources and equipment, but more important is a surge in demand of financing, which give rise to multiple banking and syndicated loan. According to statistics, since2005-2011period, the number of companies for multiple banking accounts on about17.9%of the total number of companies for bank loans, the amount is about49.8%. This shows that the multiple banking play a vital role in company financing. Due to this special function, mutualfy promote relationship always exists between banks and companies. On the one hand, banks will affect development of the companies through loosening or tightening credit capital; on the other hand, security of bank credit funds is closely connected with companies management condition. So banks will carry out rigorous initial screening and lasting ex-post monitoring to lend money, in this point of view, banks take more advantage of information than other investors, which is helpful to improve system of external supervision to the borrowing companies, and to a certain extent reduces adverse selection and borrowers’moral hazard and, hence, the problem of information asymmetry. In traditional single banking, the information asymmetry between the bank and company is a key factor to the efficiency of bank supervision, but in multiple banking, the information asymmetry between banks also affects the supervision efficiency to a large extent. The efficiency of supervision and the significant influence of bank loans will have a significant impact. So announcements of different mode of bank loan agreements by the listed company will lead to different market reaction.First of all, this thesis researches into the bank loan announcements of Chinese listed companies from2005to2011, and313valid samples are obtained from254listed firms guided by certain principle, including56samples of multiple banking and257single banking samples. Then traditional event-study method is applied according to Brownand Warner (1985) to investigate the different effect of different mode of loan announcements on the market value of borrowers. The empirical results show that CARs (cumulative abnormal returns) in our sample bank loan announcements are negative. This enables us to evaluate events’impact on stock prices, and thus, the market reaction to loan announcements. This illustrates that Chinese investors regard this behavior of the company as a "bad" event in the short term, which is quite different from the relevant foreign empirical results. In addition, according to this research, CARs in the window [-7,-3](with t=0referring to the announcement date) are negative and significantly different from0, which shows the leakage of bank loan information during the pre-event period may exist in Chinese market. Moreover, to reflect and comparatively analysis the different efficiency of bank supervision, this thesis divides the total samples into two subsamples according to the different modes of loan announcements. The empirical results show that CARs of multiple banking subsample declines than single banking subsample, CARs in the window [-1,2] of two subsamples are significantly different from each other, to some extent indicating that investors prefer single banking to multiple banking relatively.Secondly, after obtaining CARs of the listed companies’bank loans announcement, to understand whether multiple banking announcement play a more negative impact on stock prices, we focus on dummy variable that reflects different loan modes, and observe its impact to CARs, on the basis of controlling related variables of loan itself and borrowing companies, as well as the external environment. Through multiple linear regression analysis of CARs, the article attempts to find out some concrete evidence to prove that mutual effects between banks in multiple banking will lead to a lower efficiency of supervision.Concrete analysis process as follows:according to the common agent theory, the cooperation between banks can play a scale advantage of information, hence access to economies of scale and more effective supervision. But cooperation between banks can hardly be reached based on the following reasons:the different level of information asymmetry, different evaluation criteria for company performance, as well as the "free-rider" impulse prior to or after the credit. In addition,"flock effect" of multiple banking will lead to fierce competition between banks which not only weaken the company’s initiative to provide relevant information, but also boost company’s bargaining position psychologically, resulting much more grave asymmetric information problem between banks as well as company’s information fraud, then generate additional common agency costs, reduce the efficiency of bank supervision, therefore, CARs of multiple banking loans announcement will be much lower. As to multiple banking subsamples separately, the lead arranger’s state-owned or a bigger share of the loan can also give rise to much more fierce "free-rider" impulse and competition between banks, the higher the proportion, the lower corporate transparency, finally a higher agency cost and a lower efficiency of supervision.In the above assumptions, this paper firstly used a dummy variable to stand for different loan modes, then carried out regression analysis to the event window of the accumulation excess returns, and inspected excess returns’change in different loan modes conditions. The results showed that the dummy variable coefficient were the same with our assumptions basically. In order to further validate that multiple banking can lose bank supervision efficiency, another regression analysis is applied to research the relationship between CARs and state-owned as well as amount proportion of lead arranger. Test results show that excess return of the listed company caused by bank loans announcement with state-owned or a larger amount proportion of lead arranger were much lower which passed the significance test with confidence level at5%. Therefore, the conclusion perfectly confirmed our previous assumptions:noncooperation indeed exist between banks in multiple banking which result in efficiency loss of bank supervision.更多还原