The effect of volatility persistence on excess returns

• Date: 01 January 2017
• DOI: http://doi.org/10.1016/j.rfe.2016.11.003
• Published date: 01 January 2017

The effect of volatility persistence on excess returns

Ajeet Jain

a,

⁎, Sascha Strobl

b

a

AlabamaA&M University, Departmentof Finance and Economics,Normal AL 35810, UnitedStates

b

Universityof Vaasa, Departmentof Accounting and Finance, P.O.Box 700, FI-65101 Vaasa, Finland

abstractarticle info

Articlehistory:

Received8 December 2015

Receivedin revised form 24 August 2016

Accepted14 November 2016

Availableonline 15 November 2016

JEL Code:

G1

In this paper, we examine the effect of volatility persistence in explaining excess returns in conjunction with

establishedfactors. We use an I-GARCH modelto estimate volatility persistence for each companyon the NYSE

for each year between 1989 and 2014. We find that volatil ity persistence is significant in explaining excess

returns formedium to high turnover portfolios.We also find a similar relationshipfor portfolios sortedon size.

This study tries todisentangle the effects of various informationasymmetry aspects in asset pricing and show

that not only volatility itself but alsoits persistence is important in explainingreturns.

© 2016 ElsevierInc. All rights reserved.

Keywords:

Volatilitypersistence

Volatility

Liquidity

Informationasymmetry

1. Introduction

The Dow JonesIndustrial Average (DJIA)plummeted on August 24,

2015, losing N1000 points at one point in time. In fact most of the stock

markets inthe world crashed with fears of a slowingChinese economy

and regained some of these losses in the following days. Volatility in the

stock market has been and still is a topic of immense inter est to both prac-

titioners and researchers alike. There is an extant literature on volatility,

its causes and effects.

1

However, theliterature on volatility persistence

has so far concentrated on improving the estimationand neglected its im-

pact on explaining stock returns. In this paper, we try to fill this gap and

investigate the effect of volatility persistence in explaining excess returns.

The importance of volatility in explaining stock returns is we ll docu-

mented. Bollerslev, Tauchen, and Zhou (2009) find evidence that the

variance risk premium can explain aggregate market returns.Carr and

Wu (2009) show that individual stock returns are related to the vari-

ance risk premium. Researchers ha ve also looked into factors that

causes volatility which ranges fr om information asymmetry, unin-

formed traders, size and the list goes on. According to Wang (1993),

an increasein volatility can be attributed to the information asymmetry

associatedwith a firm. Daigler and Wiley(1999) attribute volatility and

volumeto less informed traderswho are away fromthe market and suf-

fer from information asymmetry. Volatility of an individual stock ca n be

separated into idiosyncratic vol atility and systematic volatility. In

addition to that, researchers have documented that idiosyncratic risk

is an important explanatory vari able in explaining excess returns.

Merton (1987) suggests a positive relationshipbetween idiosyncratic

risk and expected returns when investors hold a single stock or their

portfolio is not fully diversified.Xu & Malkiel (2003) argues that idio-

syncraticvolatilityof individual stocksis dependent on thelevel of own-

ership by financialinstitutions andexpected earnings growth.Monthly

stock returnshave a negative relationship with one monthlagged idio-

syncratic volatility (Ang, Hodrick, Xing,& Zhang, 2006). However,idio-

syncratic volatility varies s ubstantially over time and cannot be

assumed to bepersistent due to infrequentevents like earnings disclo-

sureswhich are periodicalin nature or seasonalvariations in supplyand

demandfor a particular firm (Fu,2009). In a diversifiedportfolio setting,

idiosyncratic risk can theoretically be diversified away.

Researchershave uncovereda plethora of determinantsthat explain

excess returnsin the stock market ranging from market risk premium,

book-to-market, illiquidity, size,corporate governance, adverse selec-

tion costs and more. However, the effect of vola tility persistence on

stock returns at different information asymmetry levels has not been

examined so far.To do so, we employ an I-GARCH model as developed

by Engle and Bollerslev (1986) to esti mate volatility persistence for

each company on the NYSE in each year between 1989 and 2014 and

find that volatility persistence is positively related to higher levels of

volatility, more illiquidit y and fewer analyst recommendations –in

short volatility persistenceis associated with information asymmetry.

Next, we extract unexpected volatility by regressing volatility on volatil-

ity persistence.Thus, we separate volatilityinto a long term and a sur-

prise component. We find that volatili ty persistence is significant in

Reviewof Financial Economics 32 (2017)58–63

⁎Correspondingauthor.

E-mailaddresses: ajeet.jain@aamu.edu (A. Jain),sascha.str obl@uva.fi(S.Strobl).

1

See Cremers, Halling, and Weinbaum (2015);Wachter (2013);Whitela w (1994);

Schwert(1989);Mankiw, Romer,and Shapiro (1985) and many others

http://dx.doi.org/10.1016/j.rfe.2016.11.003

1058-3300/©2016 Elsevier Inc. All rightsreserved.

Contents listsavailable at ScienceDirect

Review of Financial Economics

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