【作者】 姚佳；

【导师】 陈国进；

【作者基本信息】 厦门大学 ， 金融学， 2009， 博士

【摘要】 随着家庭理财意识的强化和金融产品的丰富，家庭资产组合选择问题重新进入学术视野，尤其是在家庭金融实践比较发达的欧美国家，家庭资产组合理论发展很快，成为与资产定价、公司金融等传统金融研究方向并立的一个新的独立研究方向。家庭金融研究具有其特有的复杂性，主要表现在：每个家庭都必须在长期且有限的期限内进行财务规划；每个家庭都拥有像劳动收入这样不可交易的资产；每个家庭都持有类似于房产这样的非流动性资产；每个家庭都面临着不同的借贷约束。而且不同家庭在资产组合配置方面有巨大的差异性，大量的家庭既没有参与股票投资也没有进行其他风险性资产投资，而有些家庭即使参与股票市场或投资其他风险性资产，资产组合也是千差万别。本文研究的主要目的，就是在借鉴中外经济学者已有研究成果的基础上，运用实证分析和规范分析相结合的方法就家庭资产组合选择进行系统分析。考察不同因素对家庭“持股之谜”现象的解释，对家庭参与投资风险性金融资产、风险性非金融资产和包括前两者在内的风险性资产的影响，以及不同因素对家庭持有各种风险性资产投资比重的影响。这些因素主要包括财富、风险性非金融资产投资，也被称为背景风险因素的房地产投资和实业投资、劳动收入、人口统计特征(年龄、受教育程度、职业、婚姻状况)。对美国家庭资产组合选择行为进行实证研究发现：家庭参与风险性金融资产投资的可能性和持有风险性金融资产比重会随着财富的增加而上升，财富效应明显。劳动收入的提高也会增加家庭投资概率和投资比例。风险性非金融资产投资，也被称为背景风险因素的房地产投资和实业投资对家庭参与风险性金融资产投资的可能性以及家庭持有风险性金融资产具有明显的“挤出”效应，会降低参与风险性金融资产的投资意愿，减少持有风险性金融资产的比重。家庭风险性资产参与和投资比例的生命周期效应也比较显著，家庭参与风险性资产的可能性和持有比例会随着年龄先上升，在退休前后达到最高点，然后开始下降。户主受教育程度高、职业稳定、已婚的家庭持有风险性金融资产的概率和比重相对更高。本文还考察了美国家庭股票投资对家庭参与风险性非金融资产和投资比例的影响，研究表明美国家庭参与股票投资或持有股票资产会对家庭风险性非金融资产产生“挤出”效应。此外，不同因素对家庭参与风险性资产和投资比例的影响也得到了检验，研究结果与对风险性金融资产的研究结果基本一致。本文通过国际比较，归纳总结出欧美各国家庭资产组合选择的规律和差异。然后根据我国宏观经济金融总量数据和12个城市家庭资产组合调查问卷数据，通过统计分析对我国家庭资产组合选择行为进行研究。研究发现，在我国高学历、高收入、年龄较大的家庭参与各项资产的投资更积极，家庭房地产投资仍是家庭主要的投资项目，占家庭总资产的比重最高。本文的创新主要体现在以下几个方面：第一，采用美国联邦储备委员会最新公布的SCF2007数据，运用离散选择模型(DCM，Discrete Choice Model)中的二元Probit选择模型和限值因变量模型(LDVM，Limited Dependent Variable Model)中的Tobit模型，考察不同因素对家庭参与股票市场投资和资产组合中股票持有比重的影响。第二，实证研究家庭风险性非金融资产，即家庭房地产投资和实业投资对家庭风险性金融资产选择和配置的影响。第三，以往的研究只考虑了风险性金融资产的影响因素，本文创新的考察了家庭投资风险性金融资产对家庭风险性非金融资产配置的影响，从而研究了风险性金融资产和风险性非金融资产的相互作用。更多还原

【Abstract】 With the improvement of household financing knowledge and enrichment of financial products, the problem of household portfolio choice has attracted much interest in academia. Theories of household portfolio achieved great progress, especially in America and European countries where the household financial practice is well developed. Household finance has already become a new research field which is paralleled with traditional research fields of asset pricing and corporate finance.Household financial problems have many special features that give the field its own character. Households must plan over long but finite horizons; they have important nontraded assets, notably their human capital; they hold illiquid assets, notably housing; they face constraints on their ability to borrow. Moreover, households show considerable heterogeneity in portfolio allocations. The majority of households hold neither common stock nor other risky assets. Others invest in stock almost exclusively. The extent to which risky asset holdings are diversified also varies greatly.The goal of this paper is to conduct systematic analysis of household portfolio with positive and normative methods by learning from the existed Chinese and foreign research works. The main task is to examine explanations of "Stockholding Puzzle" with different factors. Furthermore, effect on participation and holdings of risky financial assets, risky nonfinancial assets and risky asset by factors is to be examined. These factors include wealth, risky nonfinancial assets or back ground risk factors which are real estate and business equity, labor income and demographic characteristics which consist of age, education, occupation and marital status.Empirical results on American household portfolio show that there are strong positive effects of wealth, income and education on participation and holding shares of risky financial assets. Labor income also increases ownership and holding shares of risky financial assets. The life-cycle effect is hump-shaped, reflecting younger households are less likely to participate in investment of risky financial assets. It is worthwhile to notice that background risk factors "crowd out" household investment in risky nonfinancial assets. Also as expected, both ownership and share rise with married or employed-by-others households.The effect of stock investment on risky nonfmancial assets is investigated in chapter five. Results show that households will invest a lower proportion of its wealth in risky nonfmancial asset, if households hold stocks. Empirical consequences of household ownership and shares of risky assets are similar to those of research on risky financial assets.Through international comparison, the similarities and difference in household portfolio choice of various countries are summarized. Then according to macro economic and financial data and micro data of China, statistical analysis on household portfolio choice is conducted.Results show that households with higher education, higher income and prime age are more likely to invest in each asset. Real estate is the major investment item, which accounts for the largest proportion in household total assets.The main contributions of this paper include:Firstly, different factors’ effects on household stock participation and share of stocks in portfolio are examined with recently published the Federal Reserve Board’s Survey of Consumer Finances for 2007 by using discrete choice model and limited dependent variable model.Secondly, empirical research on household risky nonfmancial asset is conducted in this paper.Thirdly, this paper extends previous studies by considering interdependent relationship of household risky financial assets and risky nonfinancial asset.更多还原