基于收缩回归的可转债上市首日定价分析  

Convertible Bonds′ the First Listing Day Pricing Based on Shrinkage Regression

作　　者：王晓强 陈文斌[2] WANG Xiaoqiang;CHEN Wenbin(Foreign Languages Scholl Aff.To Longhua Academy of Edu.Sci.,Shenzhen,Guangdong 518110,China;School of Mathematical Sciences,Fudan University,Shanghai 200433,China)

机构地区：[1]深圳市龙华区教育科学研究院附属外国语学校,广东深圳518110 [2]复旦大学数学科学学院,上海200433

出　　处：《数学建模及其应用》2021年第3期12-22,75,共12页Mathematical Modeling and Its Applications

摘　　要：可转债是一种复杂的金融衍生工具,其定价受到学界与业界的广泛关注.目前主流的TF98定价模型对我国可转债定价的适用性有一定局限.本文利用收缩回归对我国2019年和2020年上市的两百余只可转债建模并分析,对比了回归模型与TF98定价模型的结果,分析了传统定价模型的局限性,并验证了回归模型的定价有效性.本文回归模型的拟合优度大于0.77,回归模型在测试集上的定价精度和定价稳定性明显优于传统定价模型.Convertible bonds are complex hybrid securities,their pricing model a hot topic in academia and industry.The main theoretical model——TF98 isn′t necessarily suitable for Chinese convertible bonds.This thesis analyses Chinese convertible bonds listed between 2019 and 2020 by using shrinkage regression.The regression results are compared with the traditional model,to clarify the conventional model′s limitations and to validate the regression model.The goodness-of-fit of the regression model is more than 0.77.The pricing accuracy and pricing stability of the regression model are also superior to the traditional model.

关 键 词：可转债定价 影响因子 收缩回归 TF98 

分 类 号：F832.5[经济管理—金融学]