Cross‐market spillovers with ‘volatility surprise’

• Published date: 01 November 2014
• Author: Julien Chevallier,Sofiane Aboura
• Date: 01 November 2014
• DOI: http://doi.org/10.1016/j.rfe.2014.08.002

Cross-market spillovers with ‘volatility surprise’

Sofiane Aboura

a

, Julien Chevallier

b,

⁎

a

DRM Finance,Université ParisDauphine, Place du Maréchalde Lattre de Tassigny,75775 Paris Cedex 16, France

b

IPAG BusinessSchool (IPAG Lab), 184Boulevard Saint-Germain,75006 Paris, France

abstractarticle info

Articlehistory:

Received12 November 2013

Accepted11 August 2014

Availableonline 20 August 2014

JEL classification:

C32

C4

G15

Keywords:

Cross-marketrelationships

Volatilitysurprise

Volatilityspillover

ADCCX

Asset management

This article adopts the asymme tric DCC with one exogenous variable (ADCCX) mode l developed by Vargas

(2008), by updating theconcept of ‘volatility surprise’to capture cross-marketrelationships. Current methods

for measuringspillovers donot focus on volatilityinteractions, and neglectcross-effects betweenthe conditional

variances. This paper aims to fill this gap. The dataset include s four aggregate indices representing equities,

bonds,foreign exchange rates andcommodities from 1983 to2013. The results providestrong evidence of spill-

overeffects coming from the‘volatility surprise’componentacrossmarkets. Against thebackground of the recent

financial crisis, theaim is to contribute to the literature on the interdependenciesof financial markets, both in

conditionalmeans and (co)variances.In addition, asset management implications are derived.

© 2014 ElsevierInc. All rights reserved.

1. Introduction

The study of volatility interaction is of interest to both academics and

practitioners. Changes in varianceare said to reflect the arrivalof infor-

mation, and the extent to which the marketevaluates and assimilates

new information.

1

The transmission pattern in va riance provides an

insight concerning the characteristics and dynamics of economic and

financial prices, and such information can be used to construct better

econometric models describing th e temporal dynamics of the time

series.

A risingresearch interest isdirected toward the topicof internation-

al transmission mechanisms –attrib utable to the ever-increasing

degree of interdependence amon g world financial markets –which

seem to become more pronounced during financial crises. Regarding

returns transmission, the study of returns co-movements begins with

the investigation of the benefits from internati onal diversification at

various frequencies (Schwe rt (1989),Susmel and Engle (1994) ,

Andersen and Bollerslev (1997) ). Returns, volatility and correlation

changes are closely related in financial models (Orlowski (2012) ,

Bekiros (2013)).

Regarding volatility transmission, Ross (1989) shows that it is the

volatility of an asset price, not the asset's price change, that is related

to the rate of information flow to the market. This empirical justifies

the study of international volatility transmission, in additionto returns

contagion. Schwert,French, and Stambaugh (1987) and Campbell and

Hentschel(1992) introduce the notion of the volatility feedback effect:

volatilityis typicallyhigher after a stockmarket decreasethan after it in-

creases, whichexplains the negativecorrelation between stockreturns

and futurevolatility. In the same field,various studies examinethe vol-

atility spillover effects with univariate and multivariate GARCH models

(Lin,Engle, and Ito (1994)). Thesemodels typicallyprovide practicalap-

plications for optimal portfolio sele ction or option pricing (Al Janabi

(2012),Konermann, Meinerding, and Sedova(2013)).

Let us now discuss theconcept of volatility surprise. In finance,the

attentionis usually focused onthe predictable variance,such as the con-

ditional varianceor the implied variance. However,according to Engle

(1993),it is the difference that cannotbe forecast betweenthe squared

residualsand the conditional variancethat is worthy of interest. Sucha

quantityhas been coined a ‘volatility surprise’.Hamao, Masulis,and Ng

(1990) were the first to interpret this quantity as a volatility surprise

since it lags behind the conditional variance. This new concept paved

the way for numerous studies(Kim and Rogers (1995),Chan-Lau and

Ivaschenko(2003)).

Reviewof Financial Economics 23 (2014)194–207

⁎Correspondingauthor. Tel.:+ 33 1 49 40 73 86; fax: +33 1 49 40 72 55.

E-mailaddresses: sofiane.aboura@dauphine.fr(S. Aboura),

julien.chevallier04@univ-paris8.fr(J . Chevallier).

1

Ross(1989) shows that,in a no-arbitrageeconomy, the varianceof price changes is di-

rectly relatedto the rate of information flow to the market. Engle et al. (1990) attribute

movements in variance to the time required by market participants in processing new

information.

http://dx.doi.org/10.1016/j.rfe.2014.08.002

1058-3300/©2014 Elsevier Inc. All rightsreserved.

Contents listsavailable at ScienceDirect

Review of Financial Economics

journal homepage: www.elsevier.com/locate/rfe