摘要
本文利用我国随机分析与计算领域的国际领先成果,结合有关风险与不确定性的哲学一经济学经典理论,创建了新的概率统计分布模型——随机极限正态分布。进而,在此基础上提出了审慎性风险监管指标R-VaR和R-ES。论文旨在为解决长久困扰金融监管界与实业界的厚尾风险建模测度问题提供开拓性的理论与实证支持。
The paper introduces a new distribution to improve tail risk modelling. We first demonstrate that the fundamental model of risk metrics, like VaR and ES, which leads to their inability to measure risk in a realistic, dynamic economic environment. Then, random limit normal distribution model is proven to be more effective for measuring and managing risk in the real business world. By employing the new distribution, we then propose more prudential risk metrics —— R-VaR and R-ES.
出处
《经济研究》
CSSCI
北大核心
2014年第9期135-148,共14页
Economic Research Journal
基金
国家自然科学基金项目--基于非线性数学期望的系统性风险测度与防控方法研究(批准号:71371109)的资助
关键词
审慎监管
风险测度
尖峰厚尾
在险价值
预期损失
Risk Models
Prudential Risk Management
High Peak and Fat Tails
Value at Risk
Expected Shortfall