【作者】 张宁；

【导师】 郑少锋；

【作者基本信息】 西北农林科技大学 ， 农林经济管理， 2015， 博士

【摘要】 分散股权结构下,公司所有权和控制权彼此分离导致股东与管理层之间出现代理问题(即“第一类代理问题”),该问题的有效解决途径之一便是集中股权,形成控股股东。在集中股权结构下,终极控股股东不仅享有利润最大化的所有权利益,还有足够的控制权以实现预期利益,出现了剩余索取权和剩余控制权的统一,从而解决了第一类代理问题。但是,所有权和控制权集中于终极控股股东会产生另一个问题:控股股东往往试着以其他股东和雇员的利益损失为代价而获取收益,当其控制权超过所有权时尤是如此,此即第二类代理问题。两类代理问题是公司治理研究中的核心,而股权结构则是公司治理中的重要治理机制,代理问题是否解决和股权结构是否合理最终均会反映于公司业绩之上。因此,股权结构、两类代理问题和公司业绩存在着非常密切的关系。尽管现有理论研究对三者中两两之间的单向关系进行了研究,但往往囿于直接股权结构层面,未能追溯到直接大股东背后的终极控股股东,进而以终极股权结构的深层角度切入研究两两之间的关系,也未能从“质”的角度探讨代理问题严重性,更未同时探讨终极股权结构、两类代理问题和公司业绩三者之间的相互关系。现实中,农业上市公司作为连接农业和资本市场的纽带,肩负着实现农业产业化和现代化的重任,关系到“三农问题”是否能得到妥善解决。因此,农业上市公司具有十分重要的特殊地位。然而,现有农业上市公司频频暴露出股权结构失衡(主要表现为国有股“一股独大”)、董事会和监事会规模偏小且独立性不高、两职合一现象较多影响管理层独立性、偏离农业主营业务过度多元化经营以及公司业绩不佳等诸多问题。追根溯源,这些问题恰是农业上市公司中终极股权结构设置不合理、股东与管理层之间以及终极控股股东与中小股东之间存在代理问题的反映。为了补充终极股权结构、两类代理问题和公司业绩理论研究,探究现实中农业上市公司相关治理问题的解决之道,本文立足于公司治理中的经典理论——委托代理理论和逐渐取得主流地位的终极股权结构理论,参照中国农业大学“大农业”上市公司分类方法,筛选出118家农业类上市公司2004-2013年间共计1180个平衡面板样本数据,通过对这些数据构建适当的数学模型,应用多种统计分析软件和方法对模型进行估计,旨在研究清楚现有农业类上市公司中终极股权结构安排,具体代理问题及其严重程度和代理成本,以及终极股权结构、代理问题和公司绩效三者之间的复杂关系,以此为提高公司治理质量寻得解决之道。具体而言,本文首先采用均值、标准差和四分位数等基础统计方法对样本公司终极股权结构依次进行全样本和子样本分析,从而对农业类上市公司的终极股权结构有了整体性和细致化的认识。其次,兼顾代理问题“质”和“量”两方面,分别采用因子分析法和面板数据多元回归方法分别测算出两类代理问题严重程度综合因子得分和两类代理问题所引发的成本,从而全面深刻理解两类代理问题严重程度与代理成本。再次,在上述研究的基础上,建立面板数据固定效应变截距或随机效应变截距模型对代理问题严重程度与公司业绩、终极股权结构与公司业绩、终极股权结构与两类代理问题严重程度等两两关系进行模拟分析,试图检验终极股权结构、两类代理问题严重程度和公司业绩两两之间单向关系是否成立和显著,获得此类关系的初步判断依据。最后,在单向简单关系研究的基础上,顾及终极股权结构、两类代理问题严重程度和公司业绩之间可能存在的内生性问题,构建联立方程模型模拟分析三者之间的相互关系,系统全面地捋清终极股权结构、两类代理问题严重程度和公司业绩之间的复杂关系。主要研究结论如下:第一,对终极股权结构的全样本统计结果发现,国家及其地方政府作为最为重要的终极控股股东之一,控制了超过70%的样本公司;终极控股股东的两权偏离程度较高;终极控股股东普遍建立两层以上金字塔式股权结构实施终极控制。对子样本的剖析,发现国家及其代理机构对超过三分之二的传统农林牧渔业、食品饮料业以及化肥农药业上市公司进行着终极控制,而这些上市公司终极控股股东的两权偏离程度、金字塔链层级与条数与总样本相对值基本持平;位于中部地区的涉农上市公司两权偏离程度高于西部,东部地区样本公司终极控股股东两权偏离程度最高且建立的金字塔式股权结构最复杂。第二,超过40%的样本公司存在严重的第一类代理问题,而接近半数的样本公司第二类代理问题较为严重;传统种植业第一类代理问题严重于第二类代理问题,林、牧、渔业上市公司的第二类代理问题严重性高于第一类代理问题;东部地区上市公司两类代理问题均最为严重,西部地区次之,中部地区两类代理问题最轻微。样本公司的代理成本研究结果显示,第一大股东持股比例和股权集中程度分别与第一类代理成本呈现负相关和正相关关系;金字塔式股权结构的样本公司第二类代理成本低于非金字塔式股权结构的公司;终极控股股东为自然人及其家族的公司存在更大的第二类代理成本。第三,终极控股股东两权偏离程度与第一类代理问题严重性、公司业绩与第二类代理问题严重性两两之间呈现显著的双向负相关关系;终极控股股东两权偏离程度越高,第二类代理问题越严重;第二类代理问题越严重,两权偏离程度越低;两权偏离程度越高则公司业绩也越高;公司业绩越高,两权偏离程度越低;第二类代理问题越严重则第一类代理问题也越严重。更多还原

【Abstract】 After financial crisis in East Asia the scholars owed this crisis to the faultiness of corporate governance mechanism, the last 15 years saw an upsurge of interest in the field of corporate governance. Nowadays, with the prevail of corporate governance,the item "corporate governance " is becoming hot words for common people to talk about. Apparently, citizens and scholars focus on the different point although they have the same passion to discuss the problem of corporate governance. As ordinary investor, who concern the most is how to assure they can get a return in their investment. In their opinion, the structure of corporate governance constituted by board of directors and board of supervisors would be a feasible way to guard their money in firms. While, it is far enough to understand internal governance mechanism for researchers, what they also have to undertake is studying external governance mechanism like products market, labor market and reputation, even find out the intricate relationships between governance mechanism and capital structure, agency problems and firm performance. Theoretically, ultimate ownership structure, agency problems and firm performance should be closely related for the reasons as follow. Firstly, figuring out agency problems has been always viewed as the core target of corporate governance. Secondly,(ultimate) ownership structure is not only the starting point of corporate governance, but also a crucial governance mechanism. Lastly, the effects of solving agency problems and corporate governance are all reflected by firm performance. Here are some questions: whether or not the relationships discussed above can be tested and verified by empirical data? Does ultimate ownership really affect agency problems? Are agency problems and ultimate ownership reflexes of corporate performance? Further on, are there only sample and unidirectional relation between ultimate structure, agency problems and corporate performance? Realistically, as bond between agriculture and capital market, agricultural listed firms shoulder the responsibility of achieving agriculture industrialization and agricultural modernization, influence whether three dimensional rural issues would or not be solved properly. Therefore, the performance of agricultural listed firms means much to China’s economy. Unfortunately, there are some problems about this kind of firms, such as irrational shareholding structure, small-scale and non-independence of board of directors and supervisors, CEO duality, deviating from primary business, and miserable performance. The reasons for this plight, actually, may root in ignoring agency problems, lacking the proper ways to alleviate them and unsuitable ownership structure.So this dissertation aims at clarifying ultimate ownership structure, finding out different agency problems, measuring the severity and the cost of these agency problems, understanding well the complex relationship between ultimate ownership structure, agency problems and firms performance, and finally providing some useful advice on promoting standard of existing agricultural corporate governance in China. The thesis is based on the principal-agent theory which is classic theory in corporate governance, and the theory of ultimate ownership structure which is becoming international mainstream. In terms of selecting sample, according to the study on “pan-agricultural listed firms” leaded by China Agricultural University,the author collect 118 firms during 2004-2013 to construct a balanced-panel database concluding 1180 observations in total. Generally speaking, there are lots of specific econometric methods adopted in this thesis. To begin with, I differentiate the whole sample and sub-sample sorted by industry, zone and period. On this base, I use the basic methods like mean, standard deviation and quartile to analyze ultimate ownership structure in the different samples, finally I find there are dramatic difference between the ultimate ownership structure in the whole sample, sub-sample and simple firms in the former study. Next, factor analysis method and multiple regression method are employed successively to measure severity of two different kinds of agency problems and costs of these agency problems. It is demonstrated that the degrees of severity of two agency problems are varied greatly in whole sample and the sub-samples. Besides, this dissertation analyzes the influence that two agency problems severity on firm performance, and how ultimate ownership structure affect firm performance by constructing fix effect model for panel-data. Not surprisingly, severity of two types of agency problem act on performance dis-similarly, and the nature of ultimate shareholder, cash flow rights(CFR), control rights(CR), degree of deviation between CFR and CR, pyramidal structure have the different degree effects on firms performance. At last, there-stage least square method is used to model the complicate relationship between ultimate ownership, severity of agency problems and firms performance. And the relationship is not unidirectional as former hypotheses without considering endogeneity problems. In summary, the thesis try to answer “there questions” pointed out at the beginning by adopting some reliable econometric method.Finally, it is drew that some safely conclusions provided by the process of analyzing.In terms of ultimate ownership structure in the whole sample of agricultural listed firms, as one of the most important ultimate shareholder, the state controls more than 70% sample firms; the degree of CFR-CR deviation is quite high; ultimate shareholders usually build pyramidal structure with more than two layers. In the sub-sample, the state and its affiliates dominate more than two-thirds of traditional agriculture, forestry, animal husbandry and fishery industry, food and beverage industry, chemical fertilizer and pesticide industry, in which industries the ultimate shareholder’s CFR-CR deviation, layers and links of pyramidal structure are almost same as corresponding values in the whole sample. As for samples in the different zones, listed firms in the west are lower than that in the midland when speaking to ultimate shareholder’s CFR-CR deviation. While the eastern listed firms have the highest degree of CFR-CR deviation and the most complex pyramidal structure.As for the severity of two types of agency problems, the outcomes are as follows. Agency problem among shareholders and managers(namely first agency problem) is quite serious in two-fifths firms, and nearly half of firms with severe agency problem between ultimate controlling shareholders and minority shareholders(namely second agency problem). Concretely, in forestry, animal husbandry and fishery industry, the severity of second agency problem exceeds that of first agency problem; while the second agency problem is much more serious than the first agency problem in traditional agriculture; both types of agency problems are most serious in eastern listed firms, followed by the firms in west, the severity of two types of agency problem are weakest in listed firms in the midland.Meanwhile, the outcomes of agency costs tell us that we should pay more attention on agency costs. Firstly, the holding ratio in hands of biggest shareholder is negative to the first type agency costs. On the contrary, the degree of holding concentration keeps a positive relation with the first type costs. Then, the second agency costs in the firms with pyramids are lower than that in the non-pyramid firms. When the ultimate controlling shareholders are natural person and their families, these firms ordinarily suffer greater second agency costs than other firms do.At last but not least, there-stage least square model obtains following empirical results. The first outcome is about the mutual relationship between two types of agency problems and ultimate ownership structure. It is dedicated that ultimate ownership structure has remarkable negative influence on the severity of first type agency cost, and is very remarkable positive to the severity of second type agency cost. And the deviation between two rights of ultimate controller is lower with declining the second agency problem’s servity. Then, it is very dramatic negative relationship between the gravity of second type agency problem and firm value. Next, it is providing evidence to certify that ultimate ownership structure has positive impact on firm’s performance, which, however, plays a negative role in ultimate ownership structure. At last, the second agency problem has great positive influence on the first agency problem.更多还原