Performance of mutual funds an international perspective (finance and capital markets)

Bechmann and Jesper Rangvid 3.5 Costs and performance of Danish mutual funds 6 On the Relationship Between Price and Quality in the US Mutual Fund Industry: Evidence from the 1992–2003

Performance of Mutual Funds

An International

Perspective

Greg N Gregoriou

Edited by

ADVANCES IN RISK MANAGEMENTASSET ALLOCATION AND INTERNATIONAL INVESTMENTSDIVERSIFICATION AND PORTFOLIO MANAGEMENT OF MUTUAL FUNDS

Performance of Mutual Funds

An International Perspective

Edited by

GREG N GREGORIOU

Individual chapters © contributors 2007

All rights reserved No reproduction, copy or transmission of

this publication may be made without written permission.

No paragraph of this publication may be reproduced, copied or

transmitted save with written permission or in accordance with

the provisions of the Copyright, Designs and Patents Act 1988,

or under the terms of any licence permitting limited copying

issued by the Copyright Licensing Agency, 90 Tottenham Court

Road, London W1T 4LP.

Any person who does any unauthorized act in relation to this

publication may be liable to criminal prosecution and civil

claims for damages.

The authors have asserted their rights to be identified as

the authors of this work in accordance with the Copyright, Designs

and Patents Act 1988.

First published 2007 by

PALGRAVE MACMILLAN

Houndmills, Basingstoke, Hampshire RG21 6XS and

175 Fifth Avenue, New York, N.Y 10010

Companies and representatives throughout the world

PALGRAVE MACMILLAN is the global academic imprint of the Palgrave

Macmillan division of St Martin’s Press, LLC and of Palgrave Macmillan Ltd Macmillan® is a registered trademark in the United States, United Kingdom and other countries Palgrave is a registered trademark in the European

Union and other countries.

ISBN-13: 978-0-230-01914-0

ISBN-10: 0-230-01914-5

This book is printed on paper suitable for recycling and

made from fully managed and sustained forest sources.

A catalogue record for this book is available from the British Library.

Library of Congress Cataloging-in-Publication Data

Performance of mutual funds : an international perspective / edited by

Greg N Gregoriou.

p cm — (Finance and capital markets series)

Includes bibliographical references and index.

ISBN 0-230-01914-5 (cloth : alk paper)

1 Mutual funds 2 Mutual funds–Europe I Gregoriou, Greg N., 1956—II Series: Finance and capital markets

HG4530.P427 2006

16 15 14 13 12 11 10 09 08 07

Printed and bound in Great Britain by

Antony Rowe Ltd, Chippenham and Eastbourne

my father Nicholas

3 Danish Mutual Funds: Description, Costs, Performance,

Ken L Bechmann and Jesper Rangvid

3.5 Costs and performance of Danish mutual funds

6 On the Relationship Between Price and Quality in the US

Mutual Fund Industry: Evidence from the 1992–2003 Period 108

Javier Gil-Bazo and Pablo Ruiz-Verdú

7 Yaari’s Dual Theory of Choice, Generalized

Gini’s Mean Differences, and Performance

Wolfgang Breuer and Marc Gürtler

7.3 A dual measure of risk and (generalized) Gini’s mean

9 Performance Persistence of Unit Funds: Evidence

Valerio Potí and Eoghan Duffy

10 What Is Behind the Financial Performance of

Ethical Funds? A Study of the American Market 183

Radu Burlacu, Isabelle Girerd-Potin and Denis Dupré

10.5 Performance and management style 19010.6 Ethical strength, investment style and performance 194

Silke Ber, Alexander Kempf and Stefan Ruenzi

12.2 The relationship between fund size and performance 231

I would like to thank Stephen Rutt, Publishing Director, and AlexandraDawe, Assistant Editor, at Palgrave Macmillan for their suggestions, effi-ciency and helpful comments throughout the production process, as well asKeith Povey (with Elaine Towns and Rona Gundry) for copy-editing andeditorial supervision of the highest order.

I thank the numerous anonymous referees who have commented on thepapers which made it into this book, as well as their selection

In addition, I would like to thank the numerous anonymous referees in the

US and Europe during the review and selection process of the articles proposed for this volume

Acknowledgments

xi

the peer-reviewed journal Derivatives Use, Trading and Regulation, published

by Palgrave Macmillan He has authored over fifty articles on hedge funds,and managed futures in various US and UK peer-reviewed publications,

including the Journal of Portfolio Management, Journal of Futures Markets, European Journal of Finance, Journal of Asset Management, European Journal of Operational Research and Annals of Operations Research.

The Contributors

Ken L Bechmannis an Associate Professor at the Department of Finance,Copenhagen Business School, Denmark He has studied at the University ofAarhus and received his PhD in 1999 Ken has published numerous articles onvarious aspects of the Danish financial market and has published in interna-

tional finance journals such as the Journal of Financial Markets, European Journal

of Law and Economics and Journal of Derivatives Accounting His research areas

are financial markets, corporate finance, and mutual funds

Silke Berworks as a research assistant at the Department of Finance at theUniversity of Cologne, Germany She is the coordinator of the Graduate School

of Risk Management Prior to this, She completed an apprenticeship at a bank.She holds a diploma in Business Administration from the University of

Notes on the Contributors

xii

Cologne Her main research areas are empirical and theoretical research offinancial markets, in particular mutual funds Currently, she is workingempirically on various studies of the German mutual fund market She pres-ents her research regularly at academic conferences such as the GermanFinance Association’s Annual Meetings.

Wolfgang Breuerhas been since March 2000 a full Professor of Finance atthe RWTH Aachen University, Germany’s leading technical university.From October 1995 to February 2000 he was a full Professor of Finance atthe University of Bonn He earned his PhD degree in February 1993 and hishabilitation degree in July 1995, both at the University of Cologne After hisdiploma in 1989 he worked for one year in Frankfurt as a consultant withMcKinsey & Co before continuing his academic career He has writtenabout a dozen books, contributed to more than thirty other books and writ-ten numerous peer-reviewed journal articles covering a great variety oftopics in the field of finance His current research interests focus on portfo-lio management issues, and in particular on performance evaluation formutual funds as well as behavioral corporate finance and internationalfinancial management

Radu Burlacu is Associate Professor of Finance at the University PierreMendès France, Grenoble, France His thesis entitled “Issuing ConvertibleBonds: A Signal of Firms’ Quality” won him the “Best Thesis in Finance –2001” prize, awarded by EURONEXT Stock Exchange and the French FinanceAssociation His research fields are in asset pricing (information asymmetry;rational expectations equilibrium), portfolio management (mutual fund per-formance; ethical investment) and investment decisions with hybrid securi-ties He has published several articles in these different fields

Eoghan Duffyworked with Valerio Potì at Dublin City University, Ireland,where he graduated with a thesis on performance persistence

Denis Dupré is Associate Professor of Finance at the University PierreMendès France of Grenoble, France and member of the CERAG (“Centredes Recherches Appliquées à la Gestion”) research centre His main researchfields are in banking (asset liability management; IAS and Bâle II norms),portfolio management (ethical investment, temporal diversification, pen-sion funds) and risk management (modelling; applications for bankingmanagement) He has published several books and more than twenty arti-cles in these different fields

José L Fernández-Sánchezis Researcher in Ethical Funds and CorporateSocial Responsibility in the Department of Business Administration fromthe University of Cantabria, Spain He has an MA in economics from

Queens College (CUNY), New York, and a Master’s degree in marketingfrom ESIC (Madrid) He is now working towards a PhD in BusinessAdministration at the University of Cantabria The topics he writes on mostfrequently are corporate social responsibility, firm reputation, ethical fundsand social information.

Javier Gil-Bazoreceived his BA in Business Administration and his PhD inEconomics from the University of the Basque Country, Spain He has alsoundertaken research at the Wharton School of the University of Pennsylvania,and has been an Assistant Professor of Finance at Carlos IIIUniversity, Madridsince October 2000 His research interests cover financial econometrics, assetpricing theory and institutional investment, and his research has been pub-

lished in journals such as the Journal of Financial Econometrics and Studies in Nonlinear Dynamics and Econometrics.

Isabelle Girerd-Potinis Associate Professor of Finance at the UniversityPierre Mendès France in Grenoble, France and member of the CERAG(“Centre des Recherches Appliquées à la Gestion”) research centre Her pri-mary research areas include efficiency anomalies, chaos theory, mutualfund performance, portfolio risk management and socially responsibleinvestment She has published in academic and professional journals, andhas co-authored a book on portfolio management

Marc Gürtler has been since 2002 a full Professor of Finance at the TechnicalUniversity of Braunschweig, Germany Before coming to Braunschweig hewas an Assistant Professor of Finance at the RWTH Aachen University Heearned his PhD degree in 1997 at the University of Bonn and his habilitationdegree in 2002 at the RWTH Aachen University From 1993 to 1994 heworked as a risk manager in the department of asset management of AXAColonia Insurance Company, Cologne His research interests include, in par-ticular, portfolio management, credit risk management and internationalfinancial management He has written several books and peer-reviewedjournal articles, and contributed to other books

Vijay Jog is a Chancellor Professor at Carleton University in Ottawa,Canada, where he teaches corporate finance and value-based management atthe Sprott School of Business He has published extensively, with over 100research papers and books/monographs to his name, and has won manybest-paper awards for his publications and research He has been a recipient

of over $1 million in research grants He is invited frequently to lead shops and as a keynote speaker in conferences and national association meet-ings, and is actively involved in executive learning and workshops for bothpublic- and private-sector clients across the world National Post recognizedhim in 2001 as a “Leader in Management Education” in Canada His current

work-research includes taxation and corporate finance, corporate governance andperformance, and capital market efficiency, including market for corporatecontrol.

Alexander Kempf previously held positions at the Universities of Mannheimand Frankfurt (Oder) before going to the University of Cologne, Germany Hehas been head of the Department of Finance there since 1999 Since 2003, hasalso been the spokesman of the Graduate School of Risk Management and is aboard member of the Institute of Empirical Economic and Social Research atthe University of Cologne He is dedicated to empirical and theoreticalresearch of financial markets, his main research interests being asset manage-ment and risk management He is a popular discussion and project partneramong academic professionals He acts as advisory consultant for companiesand scientific organizations as well as a referee for national and international

journals, and has been the editor of the journal Die Betriebswirtschaft since 2002.

Ladislao Lunais Professor in Business Administration at the University ofCantabria, Spain He gained a PhD in Economics and Business Sciencesfrom the University of Oviedo, Spain 1993 His areas of interest are busi-ness, corporate social responsibility and ethical funds, and the productionand distribution of agriculture

Dimitri Margaritis is Professor of International Finance in the Faculty ofBusiness at Auckland University of Technology He was previously Professor

of Economics at the University of Waikato, Hamilton, New Zealand and has

in the past held academic appointments at SUNY-Buffalo, Southern IllinoisUniversity, the University of Washington, and the University of BritishColumbia in the USA He served as Adviser and Manager of Research at theReserve Bank of New Zealand in the early 1990s and was subsequentlyappointed as the Bank’s Senior Research Fellow and returned to teach atWaikato University He was a member of the World Bank’s project onFinancial Reform and is currently the leader of the New Zealand EnterpriseEfficiency and Productivity project funded by the Foundation for Research,Science and Technology He has published extensively in the international ref-ereed literature in monetary policy, international finance, health economics,economic growth and productivity

Juan Carlos Matallín-Sáezis a Professor in the Department of Finance andAccounting in Spain at the University Jaumel He obtained his PhD infinancial economics in 2000 He has carried out and published research in management and performance of mutual funds and pension funds, and thedynamics of asset prices and their implications for mutual fund manage-

ment He has published in international journals such as Applied Financial Economics, Applied Economics, and the International Journal of Finance.

David Morenohas a PhD in economics and is a Professor in the Department

of Business and Administration in the University Carlos III, Madrid, Spain.David’s research interests center on Mutual Funds Performance Evalua-tion and new portfolio theories (based on downside risk measures) Hisresearch examines issues such as the persistence of mutual funds perform-ance, the results of considering nonlinear techniques to create new classifi-cations in mutual funds He has published in international journals – for

example, the European Journal of Operational Research and Applied Economics, and in national journals such as Boletín ICE.

Roger Ottencompleted his PhD on mutual funds in 2001 and currentlyworks as an Assistant Professor of Finance at Maastricht University in theNetherlands He researches and consults in this specialized field, and has

published in a number of leading academic journals including the Journal of Banking & Finance, European Financial Management, Journal of Performance Measurement, Journal of Asset Management, Accounting and Finance, and Managerial Finance In 2002, Roger was awarded the best-paper award for his work on European mutual funds for the European Journal of Financial Management.

Valerio Potígraduated from Bocconi University of Milan and worked formany years as a derivatives trader He later taught International Finance atQueen’s University Belfast, Northern Ireland and is now a Finance lecturer

at Dublin City University, Ireland He will shortly defend his PhD thesis atTrinity College Dublin He has publications forthcoming in internationalpeer-reviewed journals on asset pricing, and on the estimation of the volatil-ity and co-dependency of asset returns He is now working on the pricing ofnon-linear strategies and alternative investment performance evaluation

Jesper Rangvidis an Associate Professor at the Department of Finance,Copenhagen Business School, Denmark He has studied at the University

of Copenhagen and received his PhD from Copenhagen Business School in

1999 He was awarded the Silver medal from the University of Copenhagenand the Tietgen Prize from Copenhagen Business School He has published

in financial and economics journals such as the Journal of Financial Economics, European Economic Review, Economic Letters, and the International Journal of Forecasting His research interests cover the areas of mutual funds,

international finance and asset pricing

Stefan Ruenzi is an Assistant Professor of Finance in the Department ofFinance at the University of Cologne, Germany He wrote his PhD thesis onmutual fund families and has worked extensively on empirical studies deal-ing with risk-taking behavior and the performance of mutual fund managers.His research interests include theoretical and empirical financial markets He

presents his research regularly at international conferences such as theEuropean Finance Association or the Financial Management Association, andserves as a referee for many conferences and academic journals – for example,

the Journal of Economic Behavior and Organization His paper entitled

“Tourna-ments in Mutual Fund Families” won a best-paper award at the GlobalFinance Association’s Annual Meeting in 2004 He has received variousawards, including several travel grants from the Centre for Financial Research(CFR), Cologne, as well as research grants from the German AcademicExchange Service and the BSI Gamma Foundation

Pablo Ruiz-Verdúis Assistant Professor of Management at the Department

of Business Administration at Carlos III University, Madrid, Spain Hereceived his BA in economics from Carlos IIIUniversity and his PhD in eco-nomics from Stanford University, Calif., USA His research interests center

on corporate governance, institutional investors, and the relationshipbetween trade unions and firm performance

Roberto Savonais Assistant Professor of Financial Markets and Institutions,Department of Business Studies, at the University of Brescia, Italy Hereceived his PhD in finance from the University of Udine, Italy in 2002 Healso teaches at the Master MF of Brescia and collaborates with SDA andNewfin at Bocconi University His current research interests include mutualfunds, hedge funds and performance measurement, and he presented hisworks at EFMA and FMA He also organized the Euro Working Group ofFinancial Modelling Conference held in Brescia in May 2005

Rajeeva Sinhahas a PhD from Warwick Business School, UK, and teachesfinance at the Odette School of Business, University of Windsor, Ontario,Canada His doctoral thesis was on corporate governance, and he examinedthe role of board structure in disciplining top management He has hadpapers published in international journals and in edited volumes on corpo-rate governance He has also researched and published on the role of gov-ernment procurement in national technological development and presentedpapers at several international conferences His current research interests arethe performance of the money management industry, the role of hostiletakeovers in corporate governance, and the behavior of mutual fundinvestors He is widely travelled, and has taught and researched in universi-ties in India, the UK and Canada

Alireza Tourani-Radis the Chair and Professor of Finance at the School ofBusiness, Auckland University of Technology, New Zealand He is a SeniorAssociate of the Australasian Institute of Banking and Finance, and the co-director of GARP-NZ He is a past Professor of Finance at WaikatoUniversity, New Zealand, adjunct Professor of International Finance at the

University of Liege, Belgium, and Associate Professor of Finance, MaastrichtUniversity in the Netherlands He has also been a Senior Research Fellow

at the Maastricht Research School of Economics of Technology andOrganizations; the Executive Vice-President of the European FinancialManagement Association; and the Secretary of the Limburg Institute of

Financial Economics He has been on the editorial board of European Financial Management and a guest editor of several issues of Managerial Finance He has

published widely on aspects of European financial markets, corporatefinance and mutual fund performance, and completed works on Pacific-Basin financial markets

In Chapter 1 the authors examine the performance of Canadian mutualfund managers, and find that their performance is indifferent when com-pared with some well-recognized bench marks such as the TSE 300 and the90-day T-Bill rates, and is even lower when one accounts for the timing ofentry and exit by mutual fund investors They attribute this to the lack ofperformance persistence However, unlike some US studies, they do notfind evidence suggesting that Canadian mutual fund investors chase winners and are reluctant to exit from losing funds; while investors do allo-cate funds based on past performance, the allocations do not favor starfunds disproportionately Poor performers experience significant fundwithdrawals They attribute this to the differences in the tax treatment of retirement-related savings – the principal source of mutual funds assetgrowth.

Chapter 2 applies data envelopment (DEA), a mathematical programmingtechnique, to measure the performance of equity retail funds in New Zealandover the period 1998–2003 An analysis of fifty-two equity mutual funds,national and international, shows significant differences in their performances,with an average DEA efficiency score of 0.72 Applying regression analysis fur-ther shows that funds with an international asset allocation strategy have hadlower efficiency scores, and that larger funds have had higher efficiency scores.Chapter 3 examines Danish mutual funds The authors describe what isspecial about Danish mutual funds, as well as the dimensions along whichDanish funds are comparable to other European funds They discuss howDanish mutual funds have performed in absolute terms and in relation toother European mutual funds, and focus also on the costs to the investor ofpurchasing Danish mutual funds certificates Finally, the authors compareDanish fund costs with the mutual fund costs in other European countries.Chapter 4 discusses the recent evidence suggesting that behind invest-ment strategies there is a latent philosophy featuring the market in whichmoney managers operate Starting from this insight, the study explores the

Introduction

xix

styles and performance of Italian managers over the period 1999–2004,making as clear as possible all the significant idiosyncrasies the authorsfind by scrutinizing the return patterns over time.

Chapter 5 aims to supplement the existing literature on Spanish equitymutual funds’ seasonality and how it is related to the relative performance

of these funds In addition, the authors investigate whether there is anyempirical evidence of window-dressing or performance hedging strategies

in their sample

Chapter 6 investigates the relationship between fees and performance inthe US market for domestic equity mutual funds The analysis shows thatprice and quality have been related negatively in this market during theperiod 1992–2003 The result holds for different measures of performance,across fund categories, and across time periods, with a single exception: the dot-com bubble, when more expensive funds delivered higher-than-average abnormal returns

Chapter 7 examines a set of performance measures derived for the eral case of mean-risk-preferences based on a portfolio-theoretical frame-work As an application of general analysis, the authors use Yaari’s dualtheory of choice to develop a specific “measure of dual risk,” which leads

gen-to the consideration of generalized Gini mean differences The authors trast the resulting performance measures with performance evaluations inthe case of traditional mean-variance and mean-variance-skewness analy-sis via an empirical study of the German capital market

con-Chapter 8 examines the efficiency of large US stocks, bonds and anced funds using a data envelopment analysis (DEA) approach Theauthor uses different DEA models to rank and compare their efficiency, andthen compares the efficiency of the funds of the various DEA models withthe well-known risk-adjusted measure known as the Sharpe ratio

bal-Chapter 9 analyzes the persistence of Irish mutual funds using a gency table methodology The authors find little evidence of performance,but discover that risk adjustments are important in evaluating performance.Chapter 10 investigates whether it is possible to reconcile ethical andfinancial performance? Using a new measure of ethical strength, the authorsfind that US equity mutual funds exhibit a highly significant negative rela-tion between the two, suggesting the existence of ethical costs Ethical fundsseem able, nevertheless, to compensate ethical costs with superior financialperformance

contin-Chapter 11 is concerned with the development and current structure ofthe German market, and provides an overview of the products offered Thechapter also outlines possible consequences of the changing market struc-ture for the future development of the German fund market

Chapter 12 examines the relationship between mutual investment fundsize and fund financial performance, using the Spanish mutual fund mar-ket The results of the chapter show that there is a relationship betweenfund size and performance consistent with the past literature

1.1 INTRODUCTION

With nearly $440bn in assets and 51 million account holders by the end ofthe year 2003 in Canada, mutual funds now occupy a prominent positionamong financial intermediaries The 1990s witnessed an explosive growth inmutual funds in Canada; the number of accounts grew nearly tenfold duringthis period A similar growth in mutual fund assets has been reported inmany countries around the world

This phenomenal growth notwithstanding, there are serious concernsabout the value added by mutual funds, and the ability of investors to earnsuperior risk-adjusted returns The pioneering work of Jensen (1968) and

the more recent works by Malkiel (1995), Elton et al (1996) and Gruber

(1996) of US-based mutual funds cast a long shadow over the ability ofmoney managers to add value Studies by Odean (1998) of investors’ trad-ing activity, and Sirri and Tufano (1998) of fund flows also suggest thatinvestors are being seriously short-changed by their proclivity to chasewinners and their reluctance to let go of losers

The finding that mutual fund investors chase above-average performingfunds (Sirri and Tufano, 1998), are reluctant to book losses (Odean, 1998),and the evidence on declining performance persistence amongst mutualfunds (Malkiel, 1995) raises the possibility that returns to mutual fundinvestors (IRR) may be lower than returns reported by mutual funds (RR).This study provides the first-ever evidence on the magnitude of the differ-ence between IRR and RR for mutual fund investors We also test for the

asymmetric relationship between past performance and subsequent fundsflow, using panel data techniques.

The study is organized as follows: Section 1.2 provides a brief literaturereview of the discussion on mutual funds and the trading behavior of mutualfund investors Section 1.3 examines some methodological and measurementissues that underpin the validity of the findings Section 1.4 discusses thesample; section 1.5 reports the findings; and section 1.6 concludes the study

BEHAVIOR OF MUTUAL FUNDS

A number of studies have examined the performance of mutual funds(Jensen, 1968) While these studies have typically concentrated on thereported returns by mutual funds, three strands of literature lead to thepossibility that the IRR may be lower than RR The first group of studiesanalyzes the sensitivity of capital flows into funds as a function of per-formance Studies by Chevalier and Ellison (1997), Sirri and Tufano (1998)provide extensive evidence in support of an inverse relationship betweenpast performance and current fund flows Odean (1998), in a study of trad-ing behavior of more than 30,000 households, found that investors usedpast returns as a positive signal of fund quality and future performance.This has been referred to as “representative heuristic” in behavioralfinance An above-average performance by a mutual fund in the previousyear is likely to induce a greater inflow of funds in the current year

The strategy of investing in out-performing funds has been described

as the “hot hands” phenomenon Hendricks et al (1993), Goetzmann and

Ibbotson (1994) and Brown and Goetzmann (1995) suggest that mutualfunds showing above-average performance in one period will follow it upwith an above-average performance in the following period Thus, accord-ing to these studies, mutual fund investors will get higher returns if theychoose mutual fund investors that are past winners However, Malkiel(1995) in a study of US mutual funds, found that while there appeared to bepersistence of returns in the 1970s, there was no similar significant persist-ence during the 1980s In the 1980s, the performance decay was character-istic, and past performance was no predictor of future performance Theevidence on persistence is important for the IRR and RR relationship IRRwill be greater than RR if there is performance persistence, and less than RR

in the absence of performance persistence

Finally, a study by Odean (1998) documents the reluctance by investors

to realize losses This loss aversion will have the implication of wideningthe gap between RR and IRR Using a unique data set on the trading behav-ior of 30,000 households, Odean (1998) found that investors are reluctant torealize losses by selling under-performing funds This is an example of the

disposition effect (Shefirin and Statman, 1985) The combined implication

of the evidence on investors chasing past winners, lack of performance sistence and reluctance to realize losses will be that the IRR is lower than RR.Nesbitt (1995) examined the impact of market timing by mutual fundinvestors, by compiling the dollar-weighted returns of seventeen categories

per-of mutual funds, and found that the dollar-weighted returns were less thanthe time-weighted returns for every category of mutual funds Nesbitt con-cluded that investors suffer a shortfall in return because of the ill-timedmovement of funds

We report raw returns or RR defined as the percentage change in the fund’svalue for the period, including any dividends given out and net of

expenses The use of raw returns or RR is in line with Brown et al (1996),

and Chevalier and Ellison (1997), who have shown that peer-group orwithin-sector comparisons of raw returns provide a valid basis for theassessment of managerial effort in the mutual fund industry

As pointed out earlier, the asymmetric fund flows to past returns, ble lack of performance persistence, and the reluctance of investors to real-ize their losses give rise to the distinct possibility that IRR may be lowerthan RR This measure reflects the effects of the timing of investors’ pur-chase and sale of mutual funds units in the context of the fluctuation ofsecurity markets

possi-The formula for the calculation of IRR is

where

CF n  cash flow in period n

IRR  internal rate of return

The above formula gives the monthly IRR.1To annualize IRR, the following

calculation is used:

Annualized IRR  (1  IRR)12 1

As in the case of RR, IRR is calculated for year 1 and the average of years 2,

n

and the proclivity to chase winners We provide evidence on the lack of formance persistence among Canadian mutual funds As our sample com-prises all funds established upto 2001, and uses both long-term and short-termperformance averages, we compute performance averages over 2, 3, 5, 10and 15 years The inferences drawn from the tables are robust to changes inmarket sentiments and different stages of the business cycle.

per-To determine the relationship between past returns and funds flow weuse panel data methodology that allows us to account for errors in esti-mation arising out of multicollinearity and heterogeneity The sources ofheterogeneity could be factors specific either to the mutual fund or to the changes in policy environment and in business cycles In principle, thepanel data technique allows for more sophisticated models in assessing therelationship between past performance and funds flow in the presence ofother variables and with less restrictive assumptions The use of panel datahas a number of advantages in this regard First, it allows us to improve theefficiency of the estimators through the use of n t observations; “n” beingthe number of mutual funds and “t” being the time period It also alleviatesthe problem of multicollinearity, as the explanatory variables can beallowed to vary in both dimensions This is a significant improvement overthe traditional OLS techniques given the high level of correlation expectedbetween various performance measures The panel data technique allows

us to make a distinction between residual heterogeneity associated withchanges over time (period effects) and across funds (group effects)

The basic relationship using this methodology can be depicted as follows(Table 1.5):

NIF it  (P it1, NP it, Star or Loser Dummy)it it

P it1, and NP itare independent variable groups used to assess the behavior

of the dependent variable NIF it NIF itis a measure of the fund flowing into

fund i in period t P it1is the performance measure used to assess

perform-ance of the fund i in period t  1 The fund flows NIF itis also a function of

non-performance variables NP itsuch as lagged values of fund flows, agement expense ratio, size of the fund and its family, and so on

man-There are three components of the error term in the estimated ship:iis the firm-specific error component or sources of variation in per-formance changes that are specific to the firm; tis the period-specific errorcomponent or time effects that reflect the impact of policy or macroeco-nomic developments on fund flows over a period of time; and itis the nor-mal error term or the pure error term The model has been estimates usingLIMDEP version 8.0 by Econometric Software Inc

relation-The standard formulation of the dependent variable is:

NIF i,t  {TNA i,t  TNA i,t1(1 R i,t1)}/TNA i,t

where, for fund i and time period t (the frequency of period t is monthly), NIF  net inflow of funds

TNA  total net assets

To assess the long-term and short-term impacts of performance on fund

flows, NIF is measured for month 1 and the average of 3, 6 and 12 months.

As the fund flows are found to be seasonal and related to the end of the tax

year, only the estimates with the 12-month averages of NIF as the

depend-ent variable are reported in the tables

The study is based on a comprehensive sample of Canadian 914 ended mutual equity funds The data set provided by Fundata andFundmonitor.com includes live and dead funds.2The oldest fund for which

open-we have a record was established in 1950 There is no establishment dateavailable for 111 of the 914 funds in the sample However, a closer exami-nation of the dataset leads us to conclude that most of these 111 funds wereestablished prior to 1988, as 69 (62 percent) of these funds are dead Itappears that we have establishment dates for all funds established after

1988 Deaves’ (2004) study of Canadian equity funds records 190 new fundsestablished between 1988 and 1998 Our own data set shows that 193 fundswere established during this period Fundata records are near complete forthe latter part of the 1990s Therefore, it is reasonable to conclude that mostfunds in the dataset with no establishment dates available were establishedprior to 1988 We have the establishment dates of 114 dead and alive fundsbetween 1950 and 1987 The 111 funds for which establishment dates arenot available were founded either during the 1950 to 1988 period or before

We can claim, within reason, that our sample covers nearly all equity fundsestablished in Canada, dead or alive, upto the end of 2001 The total assets

of the Canadian equity funds included in the sample are $103.95 Canadianbillion, which is approximately 26.56 percent of all assets invested in mutualfunds in Canada at the end of 2002

An examination of the growth in assets and number of funds shows thatthere is a clear divide between funds in the 1970 and 1980s compared tofunds in the 1990s The average asset size of mutual funds had grown from

$19.50 million funds in the 1970s to $175.67 million in the 1990s The 1990swas a period of rapid expansion in both the number of mutual funds andthe assets invested into them Thus, out of the 800 funds for which we haveestablishment dates available in the sample, 559 were established betweenthe years 1989 and 2002 Mutual funds assets grew at an impressive annualrate of 27 percent, and the number of accounts grew annually by 39 percentduring the 1990s For Canadian equity funds (dead and alive), the annual

rate of growth in the number of funds was slightly higher, at 29 percent.The assets of live Canadian equity funds grew at a rate of 31 percent, andthe assets of dead funds grew annually at 12 percent Largely because of thestock market crash in the year 2000 and the onset of recession, in 2002, forthe first time, there was a decline in the market value of assets, and in 2003there was a decline in the number of mutual fund accounts.

Accordingly, we examine returns for investors in two stages First, wereport on long-term comparisons of returns of RR with TSE 300 and T-Billreturns; and second, we examine the relationship between RR and IRR.Table 1.2 profiles the performance of Canadian mutual funds and compares

it to two benchmarks, the TSE 300 index and the 3-month T-Bill rates The

Table 1.1 Percentages of funds and investors

outperform-ing their benchmark

Year 1 Year 2 Year 3 Year 5 Year 10 Alpha a

7

table shows that for the majority of mutual funds, performance is superior

to TSE 300 in the 1–3-year horizon ending in 2002 This was also a periodthat was more turbulent than at any time in the history of the TSE 300, andwhere movement in one stock (Nortel) accounted for 35 percent of themovement in the TSE 300 index at its peak It is possible that by simplyunderweighting in Nortel stocks because of internal policy constraints,many funds outperformed the TSE 300 However, these percentages fallsharply when we look at 5, 10 and 15-year returns In the very long run(10 to 15-year horizon) we find that most funds outperform the 3-month T-Bills but not the TSE 300 Clearly, for the live funds as at 2002, their long-term performance has been less than stellar

These figures have to be interpreted in the context of the studies onthe holding period behavior of mutual fund investors The InvestmentCompany Institute (ICI) study of US mutual fund investors (ICI, 2001)found that the median holding period of a typical mutual fund investor is

7 years The ICI survey further found that 45 percent of the small percentage

of shareholders who sold shares did not do so for the purposes of adjustingtheir portfolio but rather because they needed to release the money to buy

a house or meet educational or other expenses Odean’s (1998) study of30,000 households notes that 25 percent of the investors never sold shares

during the five and a half years of their study sample period Choi et al.’s

(2000) study found that more than half of the participants in 401(K) plansnever made a trade during the three years covered by the study TheICI (2001) also reports that 73 percent of participants in large employee-sponsored retirement plans made no changes in their asset allocations overthe 10-year period covered by the study Only 3 percent of the participantsmade six or more transactions during the sample period Although there is

no study documenting the portfolio behavior of Canadian mutual fundinvestors, it is likely to be along the lines of the US investors Even if weassume an average holding period of 5 years as opposed to 7 years for the

US investors, it is apparent that the average mutual fund returns have beenlower than the TSE 300, only 12 percent of the mutual funds over a 5-yearperiod, 13 percent over a 10-year period, and 26 percent funds over a15-year period outperformed the TSE 300

An important contribution of this study is performance assessment interms of IRR or the returns accruing to investors As can be seen fromTables 1.1 and 1.2, there is a consistent pattern of IRR being lower than RR.The mean levels of differences between RR and IRR (RR IRR) is nearly

2 percent on average, and tends to increase for long-term average ance While it is true that the majority of fund managers outperform theirchosen indexes on a RR basis, when we take alpha net of the differencebetween RR and IRR we find less than a quarter of the investors outper-form their associated indexes Thus performance may be superior on arisk-adjusted basis from the perspective of mutual fund managers, but not

perform-from the perspective of investors, as only a quarter of funds outperform theadjusted alpha.

Next, we examine the short- and long-term persistence in performance ofmutual funds The intention is to find out whether mutual fund managers dif-fer in quality and whether good managers (funds) consistently outperformthe rest of the funds in the sample Typically, persistence in long-term per-formance is assessed using the approach of Goetzmann and Ibbotson (1994)and Malkiel (1995) In assessing the scope of performance persistence inCanadian equity mutual funds, a winner (loser) is defined as a fund that hasachieved a rate of return over the calendar year that exceeds (is less than) themedian fund return Performance persistence or “hot hands” occurs whenwinning is followed by winning in the subsequent year(s) Thus if a winnercontinues to post returns greater than the median returns in the years 2, 3and 5 we include it among repeat winners We follow each fund up to 5years to investigate persistence in performance We also assess the short-termpersistence in performance of mutual funds We rank firms using monthlydata on returns in the top 5%, 10%, 15% and 25% for each month, then wefollow these funds for the following 3 months, 6 months and 12 months.Performance persistence is measured for each of the years 1970 to 2001.Tables 1.3 and 1.4 present the evidence on the long-term and short-termperformance persistence of mutual funds The long-term performance ofmutual fund investors is not persistent Winners do not repeat We find thattypically for funds that are alive, investors have a 1 in 2 chance of choosing

a repeat winner in the second year, a 1 in 4 chance of choosing a repeat

Table 1.3 Long-term performance persistence of mutual funds

Persistence in performance Persistence in performance

wins for wins for wins for wins for wins for wins for

winner in the third year, and a 1 in 20 chance of picking a repeat winner inthe fifth year The performance decay of dead funds over the years, is muchhigher, as might be expected, than that of live funds Table 1.4 focuses onthe short-term performance of mutual funds The short-term performance

of mutual funds also lacks persistence The table shows that from a corpus

of 2,557 monthly returns that were in the top 5 percent of the returns for aparticular month, fewer than 378 funds continued to be in the top 5 percentfor 3 months The number drops dramatically to 4 percent over a six-monthperiod, and none of the funds could hold on to the top 5 percent slot over a12-month period Even when we take the top quartile in terms of monthlyperformance, the number shows a sharp decline from 15,067 funds inmonth 0 to 5,202 funds over a 3-month period The number of funds drops

to 430 over a 6-month period, and to 0 over a 12-month period

This lack of performance persistence in both short-term and long-termperformance is significant and has important implications for active money

Table 1.4 Short-term performance persistence of mutual funds*

Performance No of funds No of funds No of funds No of funds (no of funds) in the from column from column from column

group (A) continue to be continue to be continue to

in the same in the same be in the same

group for 3 group for 6 group for 12 months after months after months after they were identified they were they were

in the relevant identified in identified in performance group the relevant the relevant

in column (A) performance performance

management, and money management based on market timing There islittle basis to assert the superiority of active money management One canunderstand the futility of chasing past winners; however, at the same time

it is difficult to justify exiting past losers as a rational response The lack ofperformance persistence can be a possible justification investment in passiveindexed funds This passive strategy is also supported by the long-termperformance comparisons of mutual funds with the TSE 300 and T-Billrates, as shown in Table 1.2

We now focus on the relationship between performance and funds flow.Table 1.5 reports the panel data estimates for the fund flows and perform-ance relationship As discussed in the methodology section, panel data esti-mates are more robust in dealing with multicollinearity and fund-specificfactors that may affect the fund flow and performance relationship Theestimates are corrected for autocorrelation We do not impose a premedi-tated regression model in the derivation of the estimates The choicebetween OLS and panel data estimates and in panel data estimates betweenrandom and fixed effects is made on the basis of statistical tests and diag-nostics reported in Table 1.5 The following conclusions apply to all thetables The Lagrange test statistics show that the panel data and not theOLS is a more appropriate model specification The estimated regressionsshow that there are large and significant fund-specific unobserved sources

of variation that affect the estimated relationships Thus the use of ordinaryleast squares or pooled data techniques where the error structure is assumed

to be homogenous will not provide robust estimates The Hausman tics comparing the hypothesized error structure of the estimated regressionsshows that the fixed-effect specification is superior to the random effectsmodel All the estimates were tested for period effects using time-relateddummies, and the test statistics showed the absence of period effects in allthe regressions Thus we can conclude that the estimated coefficients arenot affected in any systematic way by changes in the economic environmentand impacted by policy changes Therefore our inferences are based onpanel data fixed effects models with significant group effects and no signif-icant period effects

statis-Since mutual fund inflows are related to the tax year and tend to peak atthe end of it, we only report the 12-month averages of the standardized

variable (NIF it) in measuring fund flows Thus all the variables reported inthe regression including the performance measures that vary monthly are12-month averages We use 12-month lagged averages of the performancemeasure individually in separate regressions.3

Table 1.5 presents the regressions estimating the relationship betweenfunds flow and various performance measures A star (loser) fund is onewhose performance is in the top (bottom) 10 percent, has a track record of

at least two years and belongs to a fund family with at least twelve memberfunds Panel data estimates show that riskiness of the fund and its size are

Table 1.5 Stars and losers among individual funds top and bottom 10 percent panel data estimates*

performance

suggesting significant mutual fund specific heterogeneity in the role of fund characteristics and their performance for the net inflow of funds (NIFi,t) The estimates do

not show any significant period effects The estimates have been corrected for first order autocorrelation

* 0.05

The basic relationship that is estimated is as follows:

NIF it  (Pit 1, NP it, Star or Loser Dummy)  i  t  it

P it1, and NP it are independent variable groups used to assess the behavior of the dependent variable NIF it NIF it is a measure of the fund flowing into fund i in period

t The standard formulation of the independent variable – Net inflow of funds, is NIF i,t  {TNAi,t  TNAi,t 1 (1 Ri,t1)}/TNAi,t Where, for fund i and time period t (period t is monthly), NIF  Net inflow of funds, TNA  Total Net assets, R  Monthly Return The estimates with the 12-month averages of NIF as the dependent

variable have been reported in the table Standard deviation of returns and performance variables are lagged by 3, 6 and 12 months corresponding respectively to 3,

6 and 12-month averages of NIF used as dependent variable P it 1 is the performance measure used to assess performance of the fund i in period t 1 The fund flows NIF itis also a function of non-performance variables NPitlike lagged values of fund flows, management expense ratio Size of the fund and its age We also use a star

or loser dummy in the regression There are three components of the error term in the estimated relationship: VIis the firm-specific error component or sources of variation in performance changes that are specific to the firm; t is the period specific error component or time effects that reflect the impact of policy or macroeconomic developments on top fund flows over a period of time; it is the normal error term or the pure error term Stars and losers are defined based on 12- month lagged moving arithmetic average of monthly returns.

positively related to fund flows The net inflow of funds based on a12-month average is also significant and positively related to the laggedmonthly inflow of funds The size of the fund family is also positivelyrelated to the net inflow of funds variable Visibility of the fund and pastasset allocations appears to have an important role in the direction of newcapital flows All measures of performance except excess returns are signif-icant and positively related to the flow of funds.

It is possible that investors rank funds within the star (top 10 percent)and the loser (bottom 10 percent) funds by only rewarding the superstars(losers) This is the evidence from US funds that finds the relationshipbetween fund performance and fund flows to be asymmetric To examinethe hypothesis that mutual fund investors may award winners dispropor-tionately and are reluctant to quit losers, we include a dummy variable inthe sub-sample estimates SUPER is a dummy variable that takes the value

1 if the fund is in the top 10 percent (bottom 10 percent) by performancerank within the star (loser) sub-sample

We found that, in the star funds sub-sample, in none of the estimatedequations is the SUPER dummy variable that takes the value 1 for star fundssignificant Thus there is no evidence to suggest that investors prefer the starfunds in their incremental investment decision Contrary to the existingempirical results on US mutual funds, however, we do not find that investorsare reluctant to quit losing funds We find that the SUPER dummy that takesthe value 1 in funds in the bottom 10 percent sub-sample is consistentlynegative and related significantly to the net inflow of funds In the case of thereturns and alpha performance measure, the coefficients are significant at0.01 percent, and in the case of the Sharpe and excess return performance,the measure of the relationship is significant at 10 percent Thus, the signif-icance of the estimated coefficients of the stars and losers do not supportthe asymmetry argument in the funds flow and performance relationship

Mutual funds have become an important part of the Canadian investors’savings The value added by money managers on a long-term basis is mea-ger and inconsistent For a holding period of five years or more, only aquarter of mutual funds outperform the market Even these returns are notrealized by the mutual fund investors, as the performance of mutual fundinvestors is not consistent and a mere of 5 percent of the funds can beexpected to perform higher than the median levels of returns of the sampleafter five years This lack of performance persistence, both short-term andlong-term, and the asymmetric response to performance changes is reflected

in the lower value of the returns to investors (IRR) when compared to thereported returns of the mutual funds (RR)

In our direct examination using panel data, we find that investors do notinvest disproportionately into winning funds, and they do seem to punishlosing funds Our estimates also show that past performance and past assetallocations, as well as fund size and the size of the fund family are signifi-cant determinants of current fund flows One possible explanation ofinvestors’ willingness to move funds out of losers may be explained by thefact that a large fraction of mutual fund investments are through tax-exempt registered retirement savings plan (RRSP) accounts Our calcula-tion of monthly net cash flows suggests that 60 percent of the net cash flowinto mutual funds is in the months of January, February and March, and

95 percent of Canadian equity funds are RRSP-eligible As long as theseinvested funds continue to be held in RRSP accounts, the movement offunds in and out of them have no tax implications The load structure ofmutual funds facilitates this process Nearly 31 percent of the Canadianequity funds are no-load funds Out of the 69 percent of the funds that haveloads, 54 percent have no back-end fees and 41 percent have no front-endfees It is possible that Canadian investors have greater freedom than USinvestors to move funds into and out of existing funds Our findings alsohighlight the importance of widening the empirical base of research onmutual funds

NOTES

1 As an example, suppose an investor made just two transactions in his or her portfolio over a twelve-year period The initial investments of $10,000 were made on Jan 1, 1990 and let’s assume that the portfolio grew by 15% per year for the next eight years Subsequently, another $500,000 was added on January 1, 1998 Let’s assume that in the two years following the second investment, the portfolio fell in value by a total of 20% On January 1, 2000, the overall value of the portfolio would stand at $424,472 The cumulative (simple) return would read 17% while the internal rate of return (IRR) would be a much lower, at 58% The IRR figure reflects the fact that most of the money was invested at a high, and a large portion of it was lost over a relatively short period of time.

2 We gratefully acknowledge the support of Fundmonitor.com for the data on IRR.

3 We also ran these regressions using 3, 6 and 12-month averages of fund flows, formance, and other variables of fund characteristics, respectively The significance of the reported coefficients is not affected by the choice of a systematic averaging period.

per-REFERENCES

Brown, S J and Goetzmann, W N (1995) “Performance Persistence”, Journal of Finance,

50(2): 679–98.

Brown, K C., Harlow, W V and Starks, L T (1996) “Of Tournaments and Temptations:

An Analysis of Managerial Incentives in the Mutual Fund Industry”, Journal of Finance,

51(1): 85–109.

Chevalier, J and Ellison, G (1997) “Risk Taking by Mutual Funds as Response Incentives”,

Journal of Political Economy, 105(6): 1167–200.

Choi, J., Laibson, D and Metrick, A (2000), “Does the Internet Increase Trading? Evidence

From Investor Behaviour in 401(K) Plans”, NBER Working Paper no 7878.

Deaves, R (2004) “Data Conditioning Biases, Performance, Persistence and Flows: The Case

of Canadian Equity Funds”, Journal of Banking and Finance, 28(3): 673–94.

Elton, E., Gruber, M and Blake, C (1996) “Survivorship Bias and Mutual Fund

Performance”, Review of Financial Studies, 9(4): 1097–120.

Goetzmann, W and Ibbotson, R (1994) “Do Winners Repeat? Patterns in Mutual Fund

Behaviour”, Journal of Portfolio Management, 20(2): 9–18.

Gruber, M (1996) “Another Puzzle: The Growth in Actively Managed Mutual Funds”,

Journal of Finance, 51(3): 783–807.

Hendricks, D., Patel, J and Zeckhauser, R (1993) “Hot Hands in Mutual Funds:

Short-Run Persistence of Relative Performance”, 1974–1988, Journal of Finance, 48(1): 93–130.

ICI (Investment Company Institute) (2001) “Redemption Activity of Mutual Fund

Nesbitt, S L (1995), “Buy High, Sell Low: Timing Errors in Mutual Fund Allocations”,

Journal of Portfolio Management, 22(1): 57–60.

Odean, T (1998) “Are Investors Reluctant to Realize Their Losses?”, Journal of Finance,

53(5): 1775–98.

Shefrin, H and Statman, M (1985) “The Disposition Effect”, Journal of Finance, 40(3): 777–90 Sirri, E R and Tufano P (1998) “Costly Search and Mutual Fund Flows”, Journal of

Finance, 53(5): 1589–622.

2.1 INTRODUCTION

The enormous growth in the number of mutual funds and the volume ofinvestment in them worldwide has led to an increasing demand for tech-niques to evaluate their performance Risk measurement and performanceevaluation of mutual funds are of vital importance for investors and fundmanagers alike The performance of mutual funds has been investigatedwidely in finance literature, both theoretically and empirically, since the1970s

The majority of earlier studies – for example, Sharpe (1966) and Jensen(1968) among others – showed that the performance of mutual funds is belowcomparable market indices But later studies reported opposing results fromthe above findings, in that fund managers have access to private informa-tion sufficient to cover fund expenses and fees (Henriksson, 1984; Ippolito,1989) The findings of these studies have, however, been shown to be sensi-

tive to the methodologies applied (Elton et al., 1993) This is evident in more

recent studies, such as that by Ambachtsheer (1998), where the average formance of active investors was found to be less than the corresponding

per-benchmark returns This is further supported by Kosowski et al (2001),

who found that, after controlling for luck, only a minority of fund managerswere able to cover their costs

On the whole, the main outcome of the vast analysis in the finance ature is that the majority of mutual funds have not been able to performbetter than the indices against which they have been compared Yet thereappears to be no clear agreement on the best way to measure and comparefund performance.

liter-The conventional evaluation methods, while valuable, have a number ofproblems and shortcomings associated with their use The performancemeasurement techniques are primarily within the risk–return framework,based on the Capital Asset Pricing Model (CAPM), assuming that a fund’sinvestment behavior can be explained by a single market index The mostcommon are developed by Treynor (1965), Sharpe (1966) and Jensen (1968)

A larger Treynor ratio indicates greater returns for a certain level of risk,using beta as a measure of systematic risk The Sharpe ratio is similar toTreynor’s measure, but uses total risk as measured by the standard devia-tion of the fund’s return Jensen’s measure, better known as alpha, involvescalculating the average return of the portfolio over and above what isexpected by the CAPM, resulting in a percentage excess return

Recent literature on the cross-sectional variations in stock returns,

par-ticularly by Fama and French (1992, 1996), and Chan et al (1996) and Carhart

(1997) have, however, questioned the adequacy of a single index asset-pricingmodel to explain the performance of mutual funds Finding an appropriateasset-pricing model is essential to measure and compare the performance

of investment funds The performance measurement is susceptible to thechoice of the asset-pricing models employed (Lehman and Modest, 1987;

Elton et al., 1993) There is an enduring debate among the researchers

regard-ing the best asset-pricregard-ing model, with no consensus yet in sight

A related issue when applying asset-pricing models is the assumption of

a constant beta coefficient over the sample period under study However,whenever fund managers change their risk strategy by changing their assetallocation, known as market timing – see, for example, Treynor and Mazuy(1966), and Fabozzi and Francis (1979) – an estimation bias is introducedinto the benchmark models, rendering computed measures unreliable.Finally, the traditional measures of performance cannot easily incorpo-rate the transaction costs and management fees charged by the funds Withsome exceptions, notably Grinblatt and Titman (1989), most studies inves-tigate the performance of mutual funds net of transaction costs Grinblattand Titman (1989) argue that transaction costs could be surrogates for fundmanagers’ private information If fund managers have superior informa-tion or abilities, they can generate adequate returns to cover their highercharges The funds’ gross returns can be a complex function of transactioncosts (Choi and Murthi, 2001)

In this chapter, we apply a powerful non-parametric tool, data

envelop-ment analysis (DEA) (Charnes et al., 1978), that has recently been applied

by researchers in finance as an alternative to traditional measures of

performance It was originally suggested by Murthi et al (1997) as a measure

of portfolio performance DEA can easily capture the multidimensionalaspect of mutual fund performance and alleviate some of the problemsassociated with the traditional performance measures First, it does notrequire the specification of a functional form, in our case an asset-pricingmodel Second, it allows an estimation of the return of funds by incorpo-rating the transaction costs into the analysis without requiring a specific func-tional relationship between returns and costs DEA gives a single measure

of performance that takes into account the multiple dimensions of zational activity

organi-In the context of mutual fund performance analysis, DEA has alreadybeen applied to US mutual funds by Choi and Murthi (2001) and McMullenand Strong (1998); to Australian funds by Galagedera and Silvapulle (2002),

to the ethical funds by Basso and Funari (2003), to the real estate mutual

funds by Anderson et al (2004), to the hedge funds by Gregoriou et al (2005).

The main contribution of this study is to apply the DEA methodology tomeasure the performance of NZ mutual funds Several inputs, in addition toproxies for risk, are taken into account These, in particular, are size, expenseratio and the load factor By applying a “two-stage” procedure, first the rel-ative efficiency of New Zealand (NZ) mutual funds is calculated by applyingDEA Then regression analysis is employed to study how the differences inperformance among the funds can be explained by the fund attributes notincluded in the DEA analysis

The rest of this chapter is organized as follows An overview of the NZmutual fund market and the scant literature on NZ fund performance ispresented in section 2.2 The sample and its characteristics are described inSection 2.3, followed by the methodology in Section 2.4 Our empiricalfindings and their discussion are reported in Section 2.5 Section 2.6 con-cludes the chapter

Table 2.1 presents the characteristics of the major global mutual fund kets including those of New Zealand The total size of the New Zealandmarket is less than US$10 bn and is by far the smallest in the world This ispartially because of the accessibility of Australian mutual funds to NewZealand investors: the Australian fund market is much larger and offers awider range of alternatives The size of the retail funds and the number offunds in New Zealand are again well below international averages.Compared to the USA and Europe, fund management in New Zealand has

mar-a short history with limited published mar-anmar-alysis Boustridge mar-and Young (1996)examined the risk-adjusted performance of NZ funds from 1989 to 1995and found more than 80 percent of active fund managers underperformed