PERFORMANCE AND NEW MONEY CASH FLOWS IN REAL ESTATE MUTUAL FUNDS

• DOI: http://doi.org/10.1111/jfir.12019
• Author: Abhay Kaushik,Anita K. Pennathur
• Published date: 01 December 2013
• Date: 01 December 2013

PERFORMANCE AND NEW MONEY CASH FLOWS IN REAL

ESTATE MUTUAL FUNDS

Abhay Kaushik

Radford University

Anita K. Pennathur

Florida Atlantic University

Abstract

Empirical evidence on the performance of real estate mutual funds is mixed. Moreover,

despite the growth of new money cash flows into the funds, research on investor

rationales for the same is limited. Recent research in diversified mutual funds suggests

that conventional statistics give misleading results and wrongly attribute fund

performance to “skill”rather than to “luck.”Using cross‐sectional bootstrap

methodology for 1990–2009, our estimates show evidence of underperformance.

Moreover, it appears that ordinary least squares regressions understate the number of

underperforming funds. Our analysis of new money cash flows shows evidence of

momentum in fund flows and returns‐chasing behavior.

JEL Classification: C15, G11

I. Introduction

Real estate mutual funds have grown tremendously since the early 1990s, increasing from

fewer than 10 funds in 1990 to 409 funds by the end of 2009. Real estate funds typically

invest in real estate investment trusts (REITs). Although including a real estate

component is touted as a hedge against losses in capital markets, the pitfalls of leverage,

illiquidity, and risk have made direct investment in real estate less attractive to the average

investor. Real estate mutual funds (REMFs) provide a cheaper, safer, and more liquid

alternative to investing directly in real estate assets. Our interest in REMFs is sparked by

the theory that sector fund managers should presumably possess informational

advantages about the sector to earn superior returns, as documented by Dellva,

DeMaskey, and Smith (2001), who find evidence of momentum and superior

performance in their study of sector funds. It appears, at least at first glance, as if

some REMF managers have the ability to successfully pick income‐producing REITs.

1

We thank Tom Springer (associate editor) and an anonymous referee for their valuable comments. We also

thank participants at the Financial Management Association (2011) and Southern Finance Association (2010)

meetings and are grateful to Donna Woodward at the SAS Institute for her timely support. All remaining errors are

ours.

1

According to David Lee, portfolio manager of the T. Rowe Price Real Estate fund, one of the top‐performing

real estate funds for 2010, returning 29.9%, “While investors tend to look at REITs as a group, mutual fund managers

know to slice the market into areas that behave differently during economic cycles”(“Real Estate Funds Defy

Forecasts with 27.6% 2010 Gain,”USA Today, January 6, 2011).

The Journal of Financial Research Vol. 36, No. 4 Pages 453–470 Winter 2013

453

© 2013 The Southern Finance Association and the Southwestern Finance Association

Recently, Hartzell, Muhlhofer, and Titman (2010) document that the market

capitalization of REIT mutual funds grew at an annual compound rate of 40% over their

sample period of 1994–2005. They further note that the growth rate in real estate funds

outpaces that of other sector funds. Two interesting research questions arise based on this

enormous growth of funds. The first question deals with the performance of these funds.

The second question deals with the impact of the new money flows on the performance of

these funds. In this article, we examine the performance and impact of new money cash

flows on REMFs for 1990–2009.

Although several prior studies examine real estate funds, empirical support on the

evidence of over‐or underperformance is mixed. Our article adds to the discussion and

literatureon REMFs by asking if the evidenceof overperformancefound in previous studies

represents REMF manager skill. Recent research by Kosowski et al. (2006), Cuthbertson,

Nitzsche, and Sullivan (2008), and Fama and French ( 2010) argues that the use of

conventionaltest statistics in measuringmutual fund performance givesmisleading results,

with overperformance being inaccurately attributed to “skill”rather than to mere “luck.”

These authors recommend cross‐sectional bootstrap methodology to distinguish between

skillful and lucky fund managers. Kosowski et al. document the economic significance of

their findings—that skillful subgroups of funds add significantly to the wealth of fund

investors while poorly performing funds destroy investor wealth. Examini ngt he difference

in value added for funds that are skillful versus those that are lucky, the authors document

that funds returning an alpha of 4% per year through skill alone generate approximately

$1.2 billion in wealth, whereas funds that underperform destroy about $1.5 bi llion in

investors’wealth. Given the perception that sector fund investors are a more sophisticated

clientele who make smart asset allocation decisions, we ask an important question: do

investors obtain the performance they seek and be lieve they receive? Given the mixed

evidence of REMF performance, we examine whether the evidence of over‐or

underperformance suggested by conventional ordinary least squares (OLS) alphas is

substantiated by a more stringent measure of performance, namely, bootstrapped alphas.

ThegrowthinREMFs—in both the number of funds and assets under management

(e.g., Hartzell, Muhlhofer, and Titman 2010; Kallberg, Liu, and Trzcinka 2000)—suggests

that investors believe these funds can deliver a better return than the return generated by

passive portfolios. Investors’interest in actively managed REMFs suggests that they believe

fund managers have the skills to identify those REITs and therefore will generate higher

returns for fund investors compared to passive funds. Therefore, our first contribution to the

literature on REMFs is the use of bootstrap methodology to investigate the selectivity

abilities in fund managers. Our study also fits into the literature on stylized fund performance

such as Bello (2005) and Atkinson, Baird, and Frye (2003).

The second question of interest stems from the tremendous growth in assets under

management and involves the flow‐induced performance of REMFs. REMFs provide an

interesting platform in which to examine fund flows as they perform as a fund of funds.

We discuss the flow of funds into REMFs in a Ling and Naranjo (2003, 2006) framework

to examine whether capital flows into REMFs affect returns, and we simultaneously

examine whether past REMFs’returns affect subsequent cash flows into the funds.

Although Ling and Naranjo examine this at the aggregate level, we analyze new money

cash flows at the individual fund level. Furthermore, we decompose the incremental fund

454 The Journal of Financial Research