【作者】 刘淼；

【导师】 赵振全；

【作者基本信息】 吉林大学 ， 数量经济学， 2004， 硕士

【摘要】 Modigliani和Miller（1958）的经典文献指出公司的资本结构政策和股利政策均与公司价值无关。Modigliani和Miller的结论是建立在一系列严格的前提假设基础之上的，包括共同预期、完美市场假设等。舍弃、放松或修改这些前提假设一直是公司财务理论研究的前沿，而其中最先考虑到的因素就是税收。西方有关公司资本结构税收效应的理论和实证研究文献十分丰富，而我国在这方面的研究仍接近于空白。西方学者的研究都是基于他们本国的税收环境，经过抽象假设得出结论。我国的税收制度与西方各国之间存在巨大差异，因此，根据我国税收制度的现状，探讨公司资本结构的税收效应是非常有必要的。本文以资本结构的税收效应为研究主线，从MM定理出发，对其无税的假设条件不断进行扩展，在前人研究的基础上，结合我国的税收制度与上市公司的财务数据，从理论和实证的角度分别探讨了我国上市公司资本结构的税收效应，并对某些具体行业如何更充分的利用税收效应给出了相应的建议。全文由四章组成：第一章回顾了MM定理及其修正、扩展模型。首先从Modigliani和Miller的资本结构无关论出发，在此基础上探讨了加入公司所得税和个人所得税的必要性，并分别介绍了具体模型形式。第二章在已有研究成果的基础上，通过引入资本利得税及股利支付率拓展了Miller（1977）的模型，根据对模型的分析，可判断出公司目前的资本结构是否已经充分利用了税收效应，对于未充分利用的公司（具有潜在税收效应的公司）可以通过提高债务融资比例来增加公司价值，而且该模型还可用来判断公司目前的股利政策对税收效应的影响。此外，依据我国实际的税收制度，建立了适合我国国情的税收效应模型。在如下假设条件下：1．公司的税后利润以b（）的比例支付股利，(1-b)的比例以<WP=41>资本利得的形式支付给股东；2．公司没有新增投资，即净投资等于零。得出了同时考虑公司所得税、股利所得税、利息所得税和资本利得税时的杠杆公司价值其中：、分别为杠杆公司和无杠杆公司的价值——杠杆公司的所得税率——股东股利收入的个人所得税税率——债权人利息收入的个人所得税税率——资本利得税率——杠杆公司的债务市场价值依据上述模型，得出如下结论：1．若，则如果公司保持目前的股利政策及相应税率，增加债务融资的比例会为公司带来税收效应；2．若，则如果公司保持目前的股利政策及相应税率，增加债务融资的比例会减少公司价值，以致损害股东利益；3．若，则如果公司保持目前的股利政策及相应税率，资本结构与公司价值无关，即公司不能通过资本结构变化来获得税收效应，应改变目前的股利政策或进行纳税筹划。给出了同时征收公司所得税和个人所得税时，杠杆公司的权益资本成本表达式，应等于无杠杆公司的权益资本成本加上经公司所得税和个人所得税调整后的财务风险溢价，即 <WP=42>其中：——杠杆公司的权益资本成本——无杠杆公司的（权益）资本成本 ——杠杆公司的债务成本（利息率）——杠杆公司的权益市场价值依据我国实际税率，建立了适合我国国情的杠杆公司价值模型 （企业所得税率为33%时）从中我们可以发现：（1）尽管我国目前的资本利得税率为零，但无论公司采取什么样的股利政策，债务融资仍然可以带来税收收益；（2）如果公司采用极端低的股利政策，即b=0时，债务融资的税收利益显著减小。也就是说，公司通过资本利得这种免税形式给予股东回报可以显著降低债务融资的税收利益。第三章探讨了如何权衡税收效应与公司最大承债能力之间的关系。根据第二章的分析，可以得出具有潜在税收效应的公司应增大债务融资的比例以获得税后效应的结论，但这种做法在增加税收效应的同时，也会引起公司更大的财务风险，所以此处推导出了公司最大承债能力，引出了负债能力系数，用来约束公司在正常经营情况下进行合理的债务融资，其值若小于1，表明公司目前的债务状态已经超过了最高界限，公司必将出现亏损，权益资本必将萎缩；如果大于1，则证明公司在其经营活动范围内还存在继续负债的空间，公司可以继续借债，以求更大的杠杆效益。第四章则在分析我国上市公司资本结构特点的基础上，利用我国沪市上市公司2002年的公告数据对上述模型进行了实证分析，模型实证结果表明：我国上市公司资产负债率普遍偏低，大多数上市公司目前还没有充分意识到负债融资的“税盾效应”给公司带来的益处，尤其是公用事业、交通运输、能源电力、酿酒食品、冶金、医药类这六个行业，具有潜在税收效应公司的比例更高，说明这些公司还没有充分地利用财务杠杆，发挥其税收效应。 <WP=43>通过负债能力系数分析，我们发现大多数具有潜在税收效应的公司都存在继续增加债务融资的潜力，尤其是上述六个行业，不仅存在继续增加负债的必要，而且也存在继续负债的可行空间。更多还原

【Abstract】 In their celebrated papers, Modigliani and Miller (1958) argued that the capital structure decision and dividend policy is a matter of irrelevance, affecting neither the value of corporation nor its cost of capital. Their theories are founded on a series of assumptions, such as homogeneous expectation; symmetric information; and perfect market. In a perfect market there is not tax at all, but in the modern society, taxation is inevitability for every individual and firm. Many scholars in foreign countries devoted themselves to the issue on how taxation affects the capital structure decisions. They have conducted a lot of theoretical studies and empirical analyses, and put forward various theories with different ideas. However, little research on this problem has been done in China so far. It is known that there exist many differences in taxation between China and other countries mentioned in the previous study, therefore, my paper try to perform a theoretical in the previous study, therefore, my paper try to perform a theoretical study about the question how taxation in China affects the capital structure policies. Based on the MM theorem, this paper takes the tax effect of capital structure decision as the master clew and extends the hypothesis that there is not tax at all step by step. The paper integrate the tax system in China with the financial data in the annals of listed corporations to investigate the tax effect of capital structure decision of public listed corporations in China from theory and empirical aspect respectively. According to the above analyses, the paper gives some corresponding suggestions about how to benefit from the tax effect of capital structure more sufficiently to certain concrete Industries. The paper is composed of four sections:In the section 1, I review the modified and extended models of the MM theorem. Based on the irrelevant theory of the MM theorem, I expatiate on that <WP=45>it’s necessary to add the individual income tax and the corporation tax to it ,then I present the extended model.In the section 2, founded on conclusion that some literatures have suggested I extend Miller’s model (1977) through bringing in the capital gains tax and dividends payout ratio. According to the information the model reflected we can estimate whether the tax effect of capital structure is used sufficiently at the present time. For the corporations that have not utilized the tax effect (they can be called ‘corporations with the potential tax effect’) we can increase the ratio of debt in the capital structure to increase the corporate value and also estimate how much the dividends policy influence the tax effect. Furthermore, according to the tax system at present I construct the model of tax effect of capital structure that assort with the situation of our country. On the assumption that:(1) Corporate aftertax gain will be paid the shareholders in the proportion of b(0≤b≤1), and as the capital gains the remainder part will be paid the shareholders through shares repurchase.(2)There is no new investment in the corporations, namely net investment is zero.And when taking corporate income tax , personal income tax and capital gains tax into account corporate value iswhere=Value of a levered corporation;= Value of a unlevered corporation;=Income tax rate of a unlevered corporation; <WP=46>=Dividend personal income tax rate;=Interest personal income tax rate;=Capital gains tax rate;= Debt value of equity of levered corporation; According to the above model I gain some conclusion as follow:(1) If (1- TC ) × [b× (1- TS )+ (1- b) (1- TG)]< (1-TB ), under keeping dividend policy the level of and corresponding tax rate at present, corporation will benefit from the tax effect of the capital structure decision trough increase the ratio of debt in the capital structure;(2) If (1- TC ) × [b× (1- TS )+ (1- b) (1- TG)] > (1-TB ), under keeping dividend policy and the level of corresponding tax rate at present, that increasing the ratio of debt in the capital更多还原