基于B样条分位数高频金融数据建模及应用  

Modeling and Application of High Frequency Financial Data Based on B-spline Quantile

作　　者：赵月旭 徐伟旗 ZHAO Yuexu;XU Weiqi(School of Economics,Hangzhou Dianzi University,Hangzhou 310018,Zhejiang,China)

机构地区：[1]杭州电子科技大学经济学院,浙江杭州310018

出　　处：《信息与管理研究》2023年第6期28-40,共13页Journal of Information and Management

基　　金：教育部人文社会科学研究项目(21YJA910005)。

摘　　要：基于2018一2022年碳价、股价、外汇汇率和大宗商品小时单位数据,利用变系数模型和线性ARCH模型,构建了变系数风险度量模型,以资产相对成交量作为样条基函数,使用B样条分位数方法估计模型系数,测度高频金融数据的风险值,最后利用蒙特卡洛模拟验证了该估计方法的有效性和准确性。实证分析表明:资产相对成交量影响金融资产风险变动,B样条分位数方法的拟合成功率和误偏、误差值最小,拟合精度最高。Based on carbon price,stock price,foreign exchange rate,and commodity hourly unit data from 2018 to 2022,a variable coefficient risk measurement model is constructed by a variable coefficient model and a linear ARCH model.With the relative asset turnover as a spline basis function,the Bspline quantile method is utilized to estimate the model coefficients and measure the risk value of highfrequency financial data.Finally,Monte Carlo simulation is applied to verify the effectiveness and accuracy of the estimation method.Empirical analysis shows that the relative turnover of assets affects the risk changes of financial assets,and when the B-spline quantile method has the smallest fitting success rate,error bias,and error value,the fitting accuracy is the highest.

关 键 词：B样条 蒙特卡洛模拟 高频金融数据 分位数回归 

分 类 号：F831.5[经济管理—金融学]