Feature Mining for Financial Texts and Research on Dynamic Factor Fusion Strategy

doi:10.3778/j.issn.1002-8331.2112-0142

Abstract

Abstract: Current analysis of financial texts is limited by non-normative financial texts, and the extracted financial features are not effective enough. To solve this problem, an normative finanical text feature mining（NFTFM） model that studies brokerage research reports is proposed to extract valid financial feature factors from normative financial texts. Firstly, the normative finanical text sentiment dictionary（NFTSD） is proposed to fully mine the semantics of brokerage reporting. Then, it uses the [K]-nearest neighbor algorithm（KNN） to classify the attitude tendency of the report authors. Finally, it integrates the attitude classification results into two financial characteristic factors which include rate consistency（RV） factor and rate consistency（RC） factor according to the temporal dimension. Aiming at the problem that the factor weight of traditional quantitative multi-factor model can not adapt to the market change, the fusion factor strategy of dynamic optimization is proposed, and the weight of the factors are dynamically optimized by genetic algorithm. In order to verify the effectiveness of normative financial characteristic factors and the effect of dynamic optimization fusion factor strategy, the RC and RV factors are taken as the basic factor set to construct a multi-factor strategy instance for China Securities 500 stocks and carry out historical cycle backtest. The results show that the strategy returns have significantly improved compared with the benchmark returns, and have good adaptability to different market environments, indicating that NFTFM model effectively extracts standardized financial feature factors, and the factors under the dynamic optimization of the fusion factor strategy have the ability to adapt to market changes.

Key words: normative financial text, data analysis, [K]-nearest neighbor（KNN）, multi-factor strategy, genetic algorithm

摘要： 目前的金融文本分析受到非规范性金融文本的局限性，所提取的金融特征有效性不足。为解决这一问题，提出了以券商研究报告为研究对象的规范性金融文本特征挖掘模型（normative finanical text feature mining，NFTFM），通过构建规范性金融情感词典（normative finanical text sentiment dictionary，NFTSD）充分挖掘券商报告语义，并采用[K]邻近算法[（K]-nearest neighbor，KNN）实现报告作者评价态度分类，将态度分类结果按照时序维度整合为评价一致性因子（rate volatility，RC）和评价特征因子（rate consistency，RV）两类金融特征因子；针对传统量化多因子模型的因子权重无法自适应市场变化的问题，提出动态优化的融合因子策略，通过遗传算法动态优化因子权重。为验证规范性金融特征因子的有效性以及动态优化融合因子策略的效果，以RC、RV因子为基础因子集合，针对中证500股票构建多因子策略实例并展开历史周期回测。结果表明，策略收益相比于基准收益有明显提升，且对于不同的市场环境都具有较好的适应能力，表明NFTFM模型有效地提取了规范性金融特征因子，且动态优化的融合因子策略下的各类因子具有自适应市场变化的能力。

关键词: 规范性金融文本, 数据分析, [K]邻近算法（KNN）, 多因子策略, 遗传算法

ZHANG Wei, ZHU Hanqing, GAO Zhigang. Feature Mining for Financial Texts and Research on Dynamic Factor Fusion Strategy[J]. Computer Engineering and Applications, 2023, 59(8): 297-305.

张伟, 朱汉卿, 高志刚. 金融文本特征挖掘及动态融合因子策略研究[J]. 计算机工程与应用, 2023, 59(8): 297-305.

References

[1] BLEI D M.Probabilistic topic models[J].Communications of the ACM，2012，55（4）：77-84.
[2] HU M，LIU B.Mining and summarizing customer reviews[C]//Proceedings of the Tenth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining，2004：168-177.
[3] BACCIANELLA S A，SEBASTIANI F.SentiWordNet 3.0：an enhanced lexical resource for sentiment analysis and opinion mining[C]//Proceedings of the International Conference on Language Resources and Evaluation，2010：2200-2204.
[4] NGUYEN T H，SHIRAI K，VELCIN J.Sentiment analysis on social media for stock movement prediction[J].Expert Systems with Applications，2015，42（24）：9603-9611.
[5] WUTHRICH B，CHO V，LEUNG S，et al.Daily stock market forecast from textual web data[C]//IEEE International Conference on Systems，Man，and Cybernetics，1998.
[6] DING X，ZHANG Y，LIU T，et al.Deep learning for event-driven stock prediction[C]//Twenty-fourth International Joint Conference on Artificial Intelligence，2015：2327-2333.
[7] 丁效.基于社会媒体的市场行情预测方法研究[D].哈尔滨：哈尔滨工业大学，2016.
DING X.Research on methodology of market trends prediction based on social media[D].Harbin：Harbin Institute of Technology，2016.
[8] ZHANG W，SKIENA S.Trading strategies to exploit blog and news sentiment[C]//Proceedings of the Fourth International Conference on Weblogs and Social Media，2010：375-378.
[9] KAZEMIAN S，ZHAO S，PENN G.Evaluating sentiment analysis evaluation：a case study in securities trading[C]//Proceedings of the 5th Workshop on Computational Approaches to Subjectivity，2014：119-127.
[10] LI X，XIE H，CHEN L，et al.News impact on stock price return via sentiment analysis[J].Knowledge-Based Systems，2014，69：14-23.
[11] BOLLEN J，MAO H，ZENG X.Twitter mood predicts the stock market[J].Journal of Computational Science，2011，2（1）：1-8.
[12] SI J，MUKHERJEE A，LIU B，et al.Exploiting topic based twitter sentiment for stock prediction[C]//Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics，2013：24-29.
[13] LIU L，WU J，LI P，et al.A social-media-based approach to predicting stock comovement[J].Expert Systems with Applications，2015，42（8）：3893-3901.
[14] ZHANG X，SHI J，WANG D，et al.Exploiting investors social network for stock prediction in China’s market[J].Journal of Computational Science，2018，28：294-303.
[15] 郭卿.基于情感分析的股指预测技术的研究与实现[D].长沙：国防科学技术大学，2016.
GUO Q.Research and implementation of stock index prediction technology based on sentiment analysis[D].Changsha：National University of Defense Technology，2016.
[16] 韩雨薇.基于深度学习的多因子股票风险预测方法研究[D].杭州：浙江大学，2020.
HAN Y W.Research on multi-factor stock risk prediction method based on deep learning[D].Hangzhou：Zhejiang University，2020.
[17] 曾叶.面向金融舆情分析技术的研究与应用[D].成都：电子科技大学，2020.
ZENG Y.Research and application of financial public opinion analysis technology[D].Chengdu：University of Electronic Science and Technology of China，2020.
[18] FAMA E F，FRENCH K R.Common risk factors in the returns on stocks and bonds[J].Journal of Financial Economics，1993，33：3-56.
[19] CARHART M M.Common risk factors in the returns on stocks and bonds[J].Journal of Finance，1997，52（1）：57-82.
[20] FAMA E F，FRENCH K R.A five-factor asset pricing model[J].Journal of Financial Economics，2015，116（3）：1-22.
[21] 刘睿智.基于改进Fama-French模型的市场流动性定价实证检验[J].统计与决策，2015（11）：152-155.
LIU R Z.Empirical test of market liquidity pricing based on improved Fama-French model[J].Statistics and Decision，2015（11）：152-155.
[22] 柳垣仰.基于日内行情数据的因子模型与实证研究[D].大连：东北财经大学，2017.
LIU Y Y.Factor model and empirical research based on intraday market data[D].Dalian：Dongbei University of Finance and Economics，2017.
[23] 张信东，李建莹.盈利因子与投资因子具有定价能力吗?—来自中国股市的实证[J].金融与经济，2018（2）：10-18.
ZHANG X D，LI J Y.Do profit factor and investment factor have pricing power?—empirical evidence from China’s stock market[J].Journal of Financial Economics，2018（2）：10-18.
[24] 赵胜民，刘笑天.引入投资者偏好的多因子模型—基于前景理论视角的分析[J].中国经济问题，2019（2）：106-121.
ZHAO S M，LIU X T.Multi-factor model with investor preference—analysis based on the perspective of prospect theory[J].China Economic Studies，2019（2）：106-121.
[25] KAZEMIAN S，ZHAO S，PENN G，et al.Evaluating sentiment analysis evaluation：a case study in securities trading[C]//Meeting of the Association for Computational Linguistics，2014：119-127.
[26] REKABSAZ N，LUPU M，BAKLANOV A，et al.Volatility prediction using financial disclosures sentiments with word embedding-based IR models[C]//Meeting of the Association for Computational Linguistics，2017：1712-1721.
[27] YANG H，ZHU Y，HUANG Q，et al.A multi-indicator feature selection for CNN-driven stock index prediction[C]//International Conference on Neural Information Processing，2018：35-46.
[28] 张耀华.基于港资持仓因子LSTM分析的股票预测系统[D].大连：大连理工大学，2020.
ZHANG Y H.A stock forecasting system based on LSTM analysis of Hong Kong holdings[D].Dalian：Dalian University of Technology，2020.
[29] 金元浩，房勇，卢焱.基于K-NN最邻近点算法的因子提取方法及量化策略研究[J].数学的实践与认识，2021，51（19）：106-119.
JIN Y H，FANG Y，LU Y.Research on factor extraction method and quantization strategy based on K-NN nearest neighbor point algorithm[J].Mathematics in Practice and Theory，2021，51（19）：106-119.