MANAGEMENT SCIENCE · VOL. 63, NO. 4· APRIL 2017

近年金融危机发生之前和发生期间股票指数收益的贝叶斯分析

作者：Steven Kou (National University of Singapore), Cindy Yu (Iowa State University), Haowen Zhong (Columbia University)

摘要：笔者试图在本文中回答两个问题：（i）2008-2009年金融危机后股权收益跳跃如何变化——特别是跳跃率、跳跃规模或两者同时，是否存在显著变化？（ii）如果跳跃规模更大，仿射跳跃扩散模型的表现是否能够得到提升，即以比正态分布更肥的尾部跳跃？为了回答第二个问题，本文发现，具有随机波动率和双指数跳跃的简单仿射跳跃扩散模型能够很好拟合从1980年至2013年的标准普尔500指数和纳斯达克-100日收益率；在危机期间，该模型优于现有的方差-γ跳跃或波动跳跃等模型，并且其在危机之前至少与现有模型是相当的。 对于第一个问题，基于模型和数据集，本文观察到，在危机期间负向跳跃率显著上升，尽管平均负向跳跃规模几乎没有变化。

关键词：仿射跳跃扩散模型；股权收益；金融危机；贝叶斯马可夫链；蒙特卡罗

Jumps in Equity Index Returns Before and During the Recent Financial Crisis: A Bayesian Analysis

Steven Kou (National University of Singapore), Cindy Yu (Iowa State University), Haowen Zhong (Columbia University)

ABSTRACT

We attempt to answer two questions in this paper: (i) How did jumps in equity returns change after the 2008–2009 financial crisis—in particular, were there significant changes in jump rates or in jump sizes, or both? (ii) Can the performance of affine jump-diffusion models be improved if jump sizes are larger, i.e., jumps with tails heavier than those of the normal distribution? To answer the second question, we find that a simple affine jump-diffusion model with both stochastic volatility and double-exponential jumps fits both the S&P 500 and the NASDAQ-100 daily returns from 1980 to 2013 well; the model outperforms existing ones (e.g., models with variance-gamma jumps or jumps in volatility) during the crisis and is at least comparable before the crisis. For the first question, on the basis of the model and the data sets, we observe that during the crisis, negative jump rate increased significantly, although there was little change in the average negative jump size.

Keywords: affine jump-diffusion models; equity returns; financial crisis; Bayesian Markov chain; Monte Carlo

原文链接：

http://pubsonline.informs.org/doi/abs/10.1287/mnsc.2015.2359

翻译：景薇