股票首次公开发行（IPO）抑价是证券市场上一个长期存在，并且对市场上的各个参与者有着重要影响的问题。由于抑价现象违背了有效市场假说，理论界对此提出的种种假说与证券市场上发生的事实有较大出入，因此抑价被视为“IPO之谜”。我国证券市场的IPO上市首日表现出异常高的抑价水平，不但远远高于西方成熟市场，而且显著高于其他新兴市场。这种现象的存在严重阻碍了我国股票市场的健康发展。国外大量相关研究表明，承销商在股票首次公开发行中起着决定性作用，承销商的声誉对于其作用的发挥有着重要的影响，IPO抑价与承销商声誉之间存在着负相关关系。但由于我国的股票市场仍然属于新兴市场，有很多不规范和有待完善之处，很多在成熟市场适用的假说在我国并不一定能得到验证。那么，在我国这个新兴的股票市场上，IPO的高抑价率与承销商声誉之间到底是一种什么样的关系，高声誉承销商所承销的IPO新股的抑价率是否也相对较低呢？近几十年以来，对IPO抑价问题的研究重点逐渐集中于证券市场普遍存在的信息不对称问题。根据信息不对称所考察的主体不同，可以分为发行人与投资者之间的信息不对称、投资者之间的信息不对称以及承销商与发行人之间的信息不对称。要解决信息不对称问题就必须寻求第三方来生产和传递信息，承销商正是充当了这个第三方角色，承销商能够凭借其中介地位向投资者传递有关发行公司的信息，缓解发行人和投资者之间的信息不对称。根据“认证中介”理论，承销商中介职能的顺利发挥有赖于声誉的信号显示和传递机制，投资者根据承销商的声誉等级来甄别其所承销的IPO公司质量的好坏，承销商声誉机制的这种信号显示和传递功能对证券发行市场的有效性非常重要。据此，可以从理论上推出IPO抑价水平与承销商声誉的基本关系：如果承销商声誉的信号显示和传递功能起作用，则越是声誉好的证券承销商，越能够缓解发行公司和投资者之间的信息不对称程度，因此在其他条件相同的情况下，声誉越好的承销商所承销的IPO的抑价水平应该越低。本文选择核准制下发行的IPO新股为研究对象，这是由于核准制下，政府部门从股票发行推荐人的角色中逐渐淡出，相应地，作为上市公司推荐人的承销商获得了遴选和推荐股票发行的权力，一定程度上担负起股票发行的风险。伴随着监管体制的进一步放松，具有承销业务资格的承销商数量与日俱增，发行市场的竞争性进一步增强。声誉逐渐成为承销商的生存之本，其信号作用得以更好的发挥。本文采用经市场调整的IPO抑价率来衡量抑价水平，概览了核准制实施以后我国A股市场IPO抑价的总体情况：与审批制阶段相比较，核准制下我国股票发行市场中的IPO抑价的总体水平有所下降，但与成熟市场相比仍然偏高，且波动幅度也要比国外成熟市场大得多。在承销商声誉衡量方面，目前比较流行的方法包括JM排名法、CM排名法以及MW排名法等。我国现有的有关承销商声誉等级的排名基本类似于MW的排名系统，即都以各承销商的承销新股的业务量来衡量承销商在证券承销领域中的实力和声誉，在比较和借鉴这些排名的基础上，本文也基于MW方法，将我国承销商历年来的承销金额与承销次数综合反映在排名中，并以此作为承销商声誉等级的衡量指标。根据承销商的综合得分，将我国承销商分成四个等级，初步考察不同声誉等级承销商所承销的IPO抑价的基本情况，结果表明在我国证券发行市场中，承销商声誉与其所承销IPO抑价之间的关系并不是一种严格的负相关关系，这与理论论证的观点并不十分吻合，有待进一步进行回归分析。接着，本文以2001年4月23日到2007年12月31日发行上市的451只A股为研究对象，提出假设，参照Carter等（1998）构建的经典模型建立多元线性回归模型，对IPO抑价率与承销商声誉的关系进行实证检验，结果表明：我国证券市场上，承销商声誉与IPO抑价率之间存在着正相关关系（统计显著性不强），即承销商声誉越高越有利于降低证券发行市场中信息不对称的程度，从而由其承销的IPO的抑价水平越低的结论在我国证券发行市场上得不到支持。此外，本文尝试在回归模型的构建中加入制度虚拟变量，考察保荐制的实施对IPO抑价的影响，结果表明我国股票发行市场中的抑价现象并没有因为保荐制的实行而有所改观。针对实证检验的结果，结合声誉机制的一般原理，本文认为导致我国承销商声誉机制弱化，承销商声誉与IPO抑价之间关系扭曲的原因除了我国的证券市场仍然属于新兴市场，不稳定的因素较多之外，主要与我国声誉对承销商行为的约束力不强，监管当局监管不当，发行市场制度缺陷等因素有关。本文最后探讨了重塑我国证券承销商的声誉机制、规范我国证券承销商行为的政策取向，为实现我国证券发行市场的稳定发展提供了理论支持和政策依据。

In stock market, IPO underpricing is a long-standing problem which has critical impact on every participant in the market. Underpricing violates the Efficient Market Hypothesis, and lots of hypotheses do not explain the facts occurred in the stock market very well, so underpricing is regarded as “IPO mystery”. The abnormal high level of underpricing on the first day of IPO in the Chinese domestic stock market is not only much higher than western mature markets but also other new stock markets. The existence of this phenomenon has kept the Chinese domestic stock market away from healthy development. Lots of relevant foreign studies indicate that underwriter whose reputation is negatively related to IPO underpricing which plays a crucial role in IPO. However, since the Chinese domestic market is still a newly formed market with lots of uncompleted regulations, many theories applied in mature markets may not be tested for feasibility in China. Then, in the new Chinese stock market, what exactly is the relationship between the high IPO underpricing rate and the reputation of underwriter? Will the IPO underpricing rate of the high-reputation underwriter be relatively low?In recent decades, the commonly-existed phenomenon of asymmetric information in stock market has gradually become the focus of IPO underpricing study. According to the different study objects, it is divided into the information asymmetry between issuers and investors, the information asymmetry between investors and the information asymmetry between underwriters and issuers. To solve the problem, it is necessary to seek a third party to produce and transfer information which is undertaken by underwriter who relies on signal displaying and transferring function of the reputation system according to “Certifying Agent Theory”. Therefore, the basic relationship between IPO underpricing level and the reputation of underwriter can be conferred theoretically. If underwriter’s reputation can well play the role of signal mechanism, then underwriter with higher reputation rank can more effectively mitigate asymmetry of information between investor and issuing firm. Thereby, when the other factor is constant, the better underwriter’s reputation, the lower IPO’s underpricing.The research object of this dissertation is those new stocks issued under Approval System. Government fades out from the role of recommender, and underwriter is authorized to select and recommend the stocks correspondently. The number of underwriter is larger, and the issue market is more competitive as the supervision is looser. Under this system, reputation gradually becomes the root of survival for underwriter, and its role as a signal will play better. Next, this dissertation gives a general introduction on IPO underpricing in the Chinese stock market under Approval System by using market-adjusted underpricing rate: Compared with Licensing System phase, the general IPO underpricing level in the Chinese stock market is going down, but still higher than mature markets with relatively big fluctuation. There are three popular methods to evaluate the underwriter reputation including JM Measure, CM Measure and MW Measure. The dissertation uses MW Measure to rank underwriters about their new stock underwriting. According to comprehensive scores, this dissertation divides the underwriters into four ranks to observe the underpricing phenomenon preliminary, and it indicates that the negative relationship between underwriter reputation and IPO underpricing is not strictly, which is not identical with the former theoretical argument. Next, hypothesis and model is made to perform a regression to analyze IPO underpricing rate and the reputation of underwriter basing on 451 A stocks data collecting from 23, April, 2001 to 31, December, 2007. The analysis shows that there is a positive relationship between the reputation of underwriter and IPO underpricing rate (not strongly obvious though) which proves the theory not feasible in the Chinese stock market: the higher reputation a underwriter possesses, the easier it is to reduce the level of information unsymmetrical, and also the lower the IPO underpricing is. In addition, this dissertation adds a dummy variable to the regression model to evaluate the affect of issue system. The result shows that the underpricing phenomenon is not improved obviously by the implementation of Sponsor System.According to the result of the empirical analysis, this dissertation comes to a conclusion that besides the unstable factors under newly-formed market environment, the weakness of the Chinese reputation system of underwriters are mainly because of lack of constraint on underwriters, ineffective surveillance from relevant authorities, flawed issue system, etc. At last, this dissertation also provides polices of how to reestablish underwriter’s reputation mechanism and regulate underwriter’s behavior in order to boost steadily development of China’s security market.