Characteristics of mutual funds with extreme performance

• DOI: http://doi.org/10.1016/j.rfe.2017.04.003
• Author: Dmitry A. Shapiro,Patrick J. Schorno,Jason P. Berkowitz
• Date: 01 September 2017
• Published date: 01 September 2017

Characteristics of mutual funds with extreme performance☆

Jason P. Berkowitz

a,

⁎, Patrick J. Schorno

b

, Dmitry A. Shapiro

c

a

St. John'sUniversity, 8000 UtopiaParkway, Queens, NY 11439,United States

b

ModelValidation Executive,Ally Financial, 440 S. Church Street,Charlotte, NC 28202, UnitedStates

c

Universityof North Carolina Charlotte,9201 University CityBoulevard, Charlotte,NC 28223, United States

abstractarticle info

Articlehistory:

Received26 August 2015

Receivedin revised form 11 April 2017

Accepted15 April 2017

Availableonline 18 April 2017

JEL classifications:

G11

G23

We focus on mutual fund characteristics associatedwith periods of extreme performance. We find that funds

with eitherpositive (hot-hand) ornegative (icy-hand) persistence tend to haveportfolio similaritiesconsistent

with riskierpositions: comparedto no-streak funds, theyhold fewer stocks, invest morein top ten holdings, and

havea higher portfolio beta.Also both hot-hand and icy-handfunds have significantlyhigher asset turnoverthan

benchmark funds.Icy-hand funds tend to be more extremewith riskier positions and asset turnoverthan hot-

hand funds. At the same time, icy-hand (hot-hand) funds tend to have larger (smaller) management teams,

and are less (more)likely to be managed by one person. Finally,we do not observe many funds changingtheir

managementteams either before or afterextreme performance. Thatis, we find no evidence that the beginning

of an extreme performance period is associ ated with changes in management or that it induces changes in

management.

© 2017 Elsevier Inc. All rights reserved.

Keywords:

Mutualfunds

Extremeperformance

Fund characteristics

1. Introduction

Mutual fund performance has continually been a focus for both inves-

tors and academic researchers. Startingwith Jensen (1968), whodocu-

mented negative average fund alphas net of expenses and trading

costs, the academic literature has accumulated a large body of evidence

relating a fund's performance to its characteristics. For instance, empiri-

cal evidence suggests that larger funds (Chen, Hong, Huang, & Kubik,

2004; Ferreira, Keswani, Miguel, & Ramos, 2013; Indro, Jiang, Hu, & Lee,

1999; Yan, 2008) and funds with higher fees (Carhart, 1997; Gil-Bazo

& Ruiz-Verdu, 2009; Prather, Bertin, & Henker, 2004; Volkman &

Wohar, 1995) perform worse. Additionally, there is evidence of

short-term persistence in fund performance (Hendricks, Patel, &

Zeckhauser, 1993; Huij & Verbeek, 2007), as well as a “smart

money”effect (Gruber, 1996; Zherng, 1999; Keswani &Stolin, 2008).

The focusof this paper is on the characteristics of mutualfunds that

experiencedcontinuous periods of eitherextremely good or extremely

bad performance. Following the terminologyof Gilovich, Vallone, and

Tversky(1985), we referto a period of continuousextreme positiveper-

formanceas a hot-hand streak and a periodof continuous extremeneg-

ative performance as an icy-hand st reak. We are interested in two

groups of questions. First, whether funds with hot-han dor ic y-hand

streaks differ before, during, or after their streaks from funds that do

not experience extremeperformance. Second, whether characteristics

differ between funds with posit ive extreme performance and fu nds

with negativeextreme performance.

Using data fromMorningstar® Principia®,we form a panel data set

consisting of 10,898 “growth”and“growth and income”funds world-

wide from the first quarter of 1999 throughthe third quarter of 2011.

We define a fund as havinga streak of extremepositive (negative)per-

formanceif it had quarterly returnsin the top (bottom) 10% forat least

four consecutive quarters. Funds with extreme performance are then

compared to two benchmarks: (1) all funds without extreme perfor-

mance and (2) funds that are matched wi th extreme performance

funds using propensity score matching.

1

Our resultsare as follows. First,we show that there are common fac-

torsin how hot-hand fundsand icy-hand fundsdiffer from boththe pro-

pensity scorematched benchmark funds andall other funds. Both hot-

hand andicy-hand funds havehigher asset turnoverand they hold risk-

ier portfoliosthan benchmark funds.The latter result manifests itself in

thatfunds with extremeperformance haveportfolios thatare less diver-

sified and havehigher beta than thebenchmark. As properlydiversified

Reviewof Financial Economics 34 (2017)50–60

☆The views expressed in this paper are those of the authors and do not necessarily

reflectthose of Ally Financial.

⁎Correspondingauthor.

E-mailaddresses: berkowij@stjohns.ed u (J.P.Berkowitz), patrick.schorno@ally.com

(P.J.Schorno), dashapir@uncc.edu(D.A. Shapiro).

1

Morningstarhas 36 objectivecodes, many with onlya small portion of funds.Follow-

ing the literature, we focu s on the two largest objective codes: “growth fun ds”and

“growthand incomefunds.”These two categoriesalso havethe largest frequenciesof both

hot-handand icy-hand streaks. Although we focuson these two categories of funds, our

resultsstill hold when looking at allfund types together.

http://dx.doi.org/10.1016/j.rfe.2017.04.003

1058-3300/©2017 Elsevier Inc. All rightsreserved.

Contents listsavailable at ScienceDirect

Review of Financial Economics

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