大数据背景下金融风险模型的统计分析  

Statistical Analysis of Financial Risk Model under the Background of Big Data

作　　者：刘成良 吴盟 王绪钰 甄肖燕 

机构地区：[1]山东农业工程学院国土资源与测绘工程学院,山东 淄博 [2]山东农业工程学院基础课教学部,山东 淄博

出　　处：《应用数学进展》2024年第12期5275-5289,共15页Advances in Applied Mathematics

基　　金：国家自然科学基金面上项目(12071237);山东省统计科研课题(KT23068)。

摘　　要：本文聚焦于金融风险管理的统计预测,运用VaR理论结合蒙特卡洛模拟,对金融风险进行了量化分析。研究中,以上证指数数据为样本,精确计算了上证指数的VaR值。基于资产定价理论,结合沪深市场数据,深入剖析了关键风险因子,并利用时间序列预测模型(如SMA)预测了主要股票指数的未来趋势及潜在亏损风险。此外,我们还创新性地引入社交媒体金融舆情分析,通过微博热搜数据评估社会风险系数,拓宽了风险监测的维度。在贵金属与外汇市场方面,我们同样进行了详尽的风险系数计算,剔除了市场正常波动,确保分析的准确性和针对性。基于上述综合分析,我们构建了一个集成化的金融风险反应模型,并利用Matlab平台结合机器学习技术,设计了一套金融证券风险反应程序。该程序能够实时接入证券等相关数据,对当前金融风险进行精确计算和深入分析,为金融业的稳健发展提供有力的指导和风险防范措施。This paper focuses on the statistical prediction of financial risk management, and uses VaR theory combined with Monte Carlo simulation to quantitatively analyze financial risks. In the study, the VaR value of the Shanghai Composite Index was accurately calculated using the Shanghai Composite Index data as a sample. Based on the asset pricing theory, combined with the Shanghai and Shenzhen market data, the key risk factors are analyzed in depth, and the future trend and potential loss risk of the main stock indexes are predicted by using the time series prediction model (such as SMA). In addition, we also innovatively introduce social media financial public opinion analysis, and evaluate the social risk coefficient through Weibo hot search data, which broadens the dimension of risk monitoring. In the precious metals and foreign exchange market, we also carried out a detailed calculation of the risk coefficient, eliminating the normal fluctuations in the market to ensure the accuracy a

关 键 词：金融风险 人工智能 蒙特卡洛模拟 时间序列模拟 

分 类 号：F83[经济管理—金融学]