基于Fisher信息的Heston模型不确定性估计
统计金融
2016-10-19 v2
摘要
我们通过Fisher信息矩阵探讨欧式期权价格关于波动率的信息内容。我们假设观测到的期权价格以Heston模型提供的理论价格为中心,并受到加性高斯噪声的扰动。我们将似然函数拟合到VIX指数的组成部分上,即到期日超过23天且少于37天、具有非零买价的标普500指数近月和次月看跌与看涨期权,并根据Heston模型中的希腊字母计算它们的Fisher信息矩阵。我们发现,只要波动率足够大,期权价格就能以可忽略的不确定性对波动率进行可靠估计。有趣的是,如果波动率降至某个临界值以下,从期权价格中进行推断将变得不可能,因为Vega——欧式期权对波动率的导数——几乎消失。
引用
@article{arxiv.1610.04760,
title = {Uncertainty Estimates in the Heston Model via Fisher Information},
author = {Oliver Pfante and Nils Bertschinger},
journal= {arXiv preprint arXiv:1610.04760},
year = {2016}
}
备注
29 pages, 11 figures. The text overlap of the first version with arXiv:1609.02108 by other authors has been removed. It occurred in the introduction (page 2 line 22-37) of the first version. A part of the introduction of arXiv:1609.02108 was incidentally adopted. The issue is removed in the second version: the introduction is independent and arXiv:1609.02108 is cited properly