中国A股市场多因素选股模型实证分析

发布时间：2018-07-26 10:23

【摘要】：随着信息技术的不断发展以及中国资本市场的日渐成熟,量化投资在国内受到越来越多的关注。相比于传统投资模式,量化投资具有独特的优势,它不仅打破了传统投资的范围限制,而且克服了投资者主观因素的影响,是一种更加系统化科学化的投资模式。在量化投资策略中,多因素选股由于具有市场容量大、收益稳定的特征而受到机构投资者的青睐,本文就是在这样的背景下,尝试构建符合A股市场特征的多因素选股模型,以期能够持续稳定的战胜市场。本文中我们以多因素理论为基础,通过对股票收益与若干风险因子进行多元线性回归,构建多因素选股模型。具体而言,我们首先通过对64个可能影响股票市场收益的基础指标进行显著性检验,通过检验我们筛选出了39个显著指标,并通过降维处理得到了规模、价值、成长、财务质量、盈利收益、运营、动量、流动性以及波动性等9个风格因子,同时为规避行业因子影响,我们也将行业因子加入,与风格因子一起构建我们的多因素选股模型。实证中我们发现,我们的多因子策略能够显著的战胜市场,在6年的回溯期内年化超额收益高达26.8%,即使去除2013年结构性牛市的影响,策略组合依然能够达到22.4%的年化超额收益。此外,我们分年度、分市场阶段对模型进行了分析,我们发现不管是在牛市、熊市还是震荡市中,策略组合都能明显的战胜市场指数,而且震荡市中的表现要远好于牛市和熊市中的表现。当然我们的模型也存在一些不足,如回测中出现了几次较大的回撤,风险指标也并未完全覆盖所有可能影响股票收益的因子,这也将是我们进一步研究中重点研究的方向。
[Abstract]:With the development of information technology and the maturation of Chinese capital market, quantitative investment has been paid more and more attention in China. Compared with the traditional investment mode, quantitative investment has unique advantages. It not only breaks the limit of traditional investment scope, but also overcomes the influence of investors' subjective factors. It is a more systematic and scientific investment model. In the quantitative investment strategy, multi-factor stock selection is favored by institutional investors because of its large market capacity and stable returns. This paper tries to build a multi-factor stock selection model in accordance with the characteristics of A-share market under this background. With a view to a sustained and stable victory over the market. In this paper, based on the theory of multi-factors, we construct a multi-factor stock selection model by multivariate linear regression between stock returns and some risk factors. To be more specific, we first tested 64 basic indicators that may affect the returns of the stock market. By testing, we screened out 39 significant indicators, and through dimensionality reduction, we got scale, value, growth. Financial quality, earnings, operations, momentum, liquidity and volatility are nine style factors. In order to avoid the impact of industry factors, we also add industry factors, and build our multi-factor stock selection model together with style factors. We find that our multi-factor strategy can significantly overcome the market, in the six-year retroactive period, the annualized excess return is as high as 26.8.Even excluding the impact of the structural bull market in 2013, the strategy portfolio can still reach 22.4% annualized excess return. In addition, we analyze the model year by year and market stage, and we find that in bull market, bear market or shock market, the strategy combination can obviously beat the market index. And shock market performance is much better than the bull market and bear market performance. Of course, there are some shortcomings in our model, for example, there are several large recoveries in the back test, and the risk index does not completely cover all the factors that may affect the stock returns, which will also be the focus of our further research.
【学位授予单位】：南京大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F832.51

【相似文献】