【作者】 舒建平；

【导师】 胡培；

【作者基本信息】 西南交通大学 ， 管理科学与工程， 2006， 博士

【摘要】 风险度量是证券投资领域的核心课题之一。长期以来，有关证券的风险度量存在着两大理论体系——全域风险理论和非对称风险理论： 全域风险理论将风险定义为收益的不确定性。常见的风险度量指标包括基于随机游走的方差、β指标以及基于收益长期相关性特征的赫斯特指数、λ参数、C指标等。实证结果显示收益序列具有复杂的相关性特征：有的收益序列中仅存在着长期相关性或短期相关性，有的则既存在着长期相关性也存在着短期相关性，上述风险度度量指标难以对此进行完备的刻画。 非对称风险理论则认为，风险是投资者关于收益不确定性的心理感受，并因投资者对收益不确定性的不同部分具有不同的心理感受而表现出非对称性。该风险理论又包括下方风险理论和组合偏差风险理论，其中，前者将风险视为投资的损失，后者则将损失和收益(相对于特定的目标或基准收益水平)都纳入风险度量的范畴，但二者的作用不同。比较而言，组合偏差风险理论更全面地反映了投资者关于风险的实际心理感受。但是，现有组合偏差风险度量指标存在不足之处，即，它们没有将损失和收益的波动性考虑在内。 证券(市场)投资价值分析本质上是对收益和风险的某种综合评价。现有研究主要是从定价的角度来考察证券(市场)的投资价值，即，研究证券(市场)价格同其基本价值之间的关系，常用的方法包括股利贴现模型、市盈率模型、风险溢价模型等。关于中国股市的投资价值研究大都采用上述方法进行，其不足主要表现在两个方面：第一，它仅从投资者个体的视角来考察投资者能否获利，而忽略了证券市场重要功能——资源配置的有效性，因而并不全面；第二，所采用的研究方法除了本身固有的不足之外，更主要的缺点是忽略了投资者对证券(市场)价值判断的主观因素，它们或许可以应用于发达国家的成熟市场中，但在不成熟的新兴市场(如中国股市)却难以得出精确、一致的结论。 在对现有研究进行总结的基础上，本论文进行了如下研究： (1)在全域风险理论框架下，从时间序列预报的角度将风险定义为收益的非预测性，并用收益序列中的噪声对其进行度量，将具有复杂收益序列特征的证券风险度量纳入统一的分析框架，具有较强的适应性。 (2)结合展望理论，构造了一种同时考察上(正)偏差、下(负)偏差及其波动性的组合偏差风险度量指标，比现有的组合偏差风险度量指标更符合投资者关于风险的决策心理。 (3)拓展了投资价值研究的内涵，提出了从市场配置资源功能角度来研更多还原

【Abstract】 Risk measurement is one of the most important issues in security investment. For a long time, there are two theoretical systems in risk measurement, they are full domain risk theory and asymmetric risk theory.In full domain risk theory, risk is defined as the uncertainty of returns. Variance and β based on random walk and Hurst Index, index C and parameter X based on returns’ long-term correlation are common risk measurement indexes in full domain risk theory. The empirical results showed that the correlation of return series were very complicated, such as some returns series have only short-term correlation or long-term correlation, while others have both, which can not be described perfectly by the risk indexes above mentioned.In asymmetric risk theory, risk is investor’s psychology feeling to return’s uncertainty and shows asymmetry, for investor has different psychology feeling to different part of security return’s uncertainty. There two kinds of risk theories in asymmetric risk theory framework, one is down-side risk theory, the other is compounded deviation risk theory. Down-side risk theory considers risk as the loss of investment, while compounded deviation risk theory thinks that both loss and gain should be included in risk measurement though they have different influence. Comparing compounded deviation risk theory with down-side risk theory, we find that the former can reflect investor’s real psychology about risk more accurately. But, the default of existed compounded deviation risk indexes is that the fluctuation of loss and gain are not included in.The analysis of security(market) investment value is essentially some kind of comprehensive assessment about return and risk. The existed research analyzes the security(market) investment value from the viewpoint of pricing, that is, studying the relation of security(market) price and its fundamental value. Dividend discounted model, price earning ratio model and risk premium model are most used in empirical analysis, and also used in the research of investment value about China security market. The defaults of existed research are mainly in two sides: First, it cares about whether investors can profit from investment but neglects the efficiency of resource allocation which is a very important function of security market. Second, besides their own limits of the ways used, the fact that they ignore更多还原