More than You Ever Wanted to Know About the VIX: Bringing Together Serial Correlation and Volatility Clustering

30 Pages Posted: 26 Feb 2014

See all articles by Stefan Rostek

Stefan Rostek

University of Tuebingen - Faculty of Economics and Business Administration

Date Written: February 25, 2014

Abstract

In this paper, I show that the volatility index VIX is not model-free as soon as the diff usion term is not Brownian motion even when correcting for jumps. In a stock index model that allows for temporary periods of under- or overreaction, such as a multifractional model, a wrong annualization procedure associated with the VIX, may lead to severe mispricings as the three concepts of the VIX, the volatility per annum and (the square-root of) realized variance do not coincide any longer.

As a byproduct, this article proves that multifractionality is a parsimonious and economically sound explanation for the observable phenomenon of volatility clustering as it suggests that clusters of abnormally high (low) volatility can be attributed to periods of underreaction (overreaction). In this way, the paper brings together two observable patterns in fi nancial markets: the transitory existence of periods with serially correlated asset returns and the phenomenon of volatility clustering.

Keywords: volatility clustering, serial correlation, multifractionality, model-free volatility

JEL Classification: G10, G13, G14

Suggested Citation

Rostek, Stefan, More than You Ever Wanted to Know About the VIX: Bringing Together Serial Correlation and Volatility Clustering (February 25, 2014). Available at SSRN: https://ssrn.com/abstract=2401043 or http://dx.doi.org/10.2139/ssrn.2401043

Stefan Rostek (Contact Author)

University of Tuebingen - Faculty of Economics and Business Administration ( email )

Mohlstrasse 36
D-72074 Tuebingen, 72074
Germany

Register to save articles to
your library

Register

Paper statistics

Downloads
164
Abstract Views
670
rank
181,500
PlumX Metrics