【作者】 何彦；

【导师】 赵华；

【作者基本信息】 厦门大学 ， 统计学， 2020， 硕士

【摘要】 资产定价一直为金融学研究的热门领域,但学界同时考虑资产及市场的上行风险及下行风险,针对不同情况下资产波动与市场波动之间的关系对资产收益率影响的研究较为欠缺,且涉及资产上行系统性风险及下行系统性风险的研究较少。因此,本文试图利用最新的已实现半协方差拆分方法,使用资产与市场的已实现半协方差将已实现贝塔进一步拆分并重组,从而探究资产波动、市场波动之间的关系及资产上行系统性风险、下行系统性风险对资产预期收益可能产生的影响。基于上述背景,本文选取沪深A股高频数据构建已实现贝塔,使用已实现半协方差将其拆分,根据计算已实现半协方差所使用的市场收益率和资产收益率的符号将拆分而成的分项命名为正贝塔(资产收益和市场收益均大于0)、负贝塔(资产收益和市场收益均小于0)、混合正贝塔(市场收益大于0,资产收益小于0)及混合负贝塔(市场收益小于0,个股收益大于0)。同时根据个股收益率符号,将拆分得到的贝塔系数进行重整,即将正贝塔及混合负贝塔的和定义为好贝塔,负贝塔及混合正贝塔的和定义为坏贝塔,并将它们作为衡量资产上行(下行)系统性风险的变量。研究发现,正贝塔、混合负贝塔及好贝塔有着负的风险报酬,负贝塔、混合正贝塔及坏贝塔有着正的风险报酬。综合来看,正贝塔、混合负贝塔及好贝塔对收益率的解释力度更加稳健,其中混合负贝塔和好贝塔的风险报酬始终显著为负。而在控制其他变量的影响后,坏贝塔的风险报酬显著为正。同时我们发现,投资者承担资产上行系统性风险会要求负的风险报酬,承担下行系统性风险会要求正的风险报酬。本文还根据上述结果构建了投资组合策略,最后结果表明基于正贝塔、混合负贝塔以及好贝塔的各策略均明显优于基于传统已实现贝塔的投资策略及买入持有策略,且综合所有指标来看,基于好贝塔的策略表现最优。更多还原

【Abstract】 Asset pricing has always been a hot area of financial research,but considering both the upside and downside risks of assets and the market,the research on the impact of asset volatility and market volatility on the return on assets in different situations is relatively lacking,and there is less research on the upward systemic risk and downward systemic risk of assets.Therefore,this article attempts to use the latest realized semicovariance split method to further separate the realized beta,so as to explore the impact of the relationship between asset volatility and market volatility and upward systemic risk and downward systemic risk of assets on the expected assets returns.Based on the above background,this article selects the high-frequency data of Shanghai and Shenzhen A shares to construct the realized beta.The realized beta is split using the realized semi-covariance.According to the symbols of the market returns and assets returns used for the calculation of realized semi-covariance,the split items are named positive beta(both asset returns and market returns are greater than 0),negative beta(both asset returns and market returns are less than 0),mixed positive betas(market returns greater than 0,asset returns less than 0)And mixed negative beta(market returns less than 0,stock returns greater than 0).At the same time,according to the individual stock return symbol,the beta coefficient obtained by the splitting is reformed,that is,the sum of positive beta and mixed negative beta is defined as good beta,and the sum of negative beta and mixed positive beta is defined as bad beta.And use them as variables to measure the upside(downside)systemic risk of assets.The study found that positive beta,mixed negative beta and good beta have negative risk rewards,and negative beta,mixed positive beta and bad beta have positive risk rewards.Taken together,the impact of the positive beta,mixed negative beta and good beta on asset return is more robust,and risk rewards of mixed negative beta and good beta is always significant.When controlling other variables,the risk reward of bad beta is significantly positive.And Investors who undertake systemic risks on assets will require negative risk rewards,and those who undertake systemic risks on downsides will require positive risk rewards.At the same time,this article also builds investment strategy based on the above results.The results show that the strategies based on positive beta,mixed negative beta and good beta are significantly better than the investment strategies based on traditional realized betas and Buy&Hold strategy.And the investment strategy based on good beta performs best.更多还原