Cautious Risk Takers: Investor Preferences and Demand for Active Management

• DOI: http://doi.org/10.1111/jofi.12747
• Published date: 01 April 2019
• Date: 01 April 2019

THE JOURNAL OF FINANCE •VOL. LXXIV, NO. 2 •APRIL 2019

Cautious Risk Takers: Investor Preferences and

Demand for Active Management

VALERY POLKOVNICHENKO, KELSEY D. WEI, and FENG ZHAO∗

ABSTRACT

Despite their mediocre mean performance, actively managed mutual funds are dis-

tinct from passive funds in their return distributions. Active value funds better hedge

downside risk, while active growth funds better capture upside potential. Since such

performance features may appeal to investors with tail-overweighting preferences,

we show that preferences for downside protection and upside potential estimated

from the empirical pricing kernel can help explain active fund flows in the value

and growth categories, respectively.This effect of investor risk preferences varies sig-

nificantly with funds’ downside-hedging and upside-capturing ability, with levels of

active management, and across retirement and retail funds.

DESPITE THE MEDIOCRE PERFORMANCE of actively managed mutual funds rela-

tive to their passively managed counterparts, assets under active management

continue to significantly outweigh those of index funds.1This puzzle has at-

tracted considerable interest in the mutual fund literature. While some stud-

ies attempt to rationalize investments in actively managed funds by modeling

state-dependent managerial effort or skill, in this paper we explore components

in the demand for active management that stem from investor preferences for

upside potential or downside risk protection. We show that, compared with pas-

sive benchmarks, active value funds achieve greater downside hedging while

∗Valery Polkovnichenko is with University of Texas at Dallas and the Federal Reserve Board

of Governors. Kelsey D. Wei and Feng Zhao are with University of Texas at Dallas. We thank

Ken Singleton (the Editor); the anonymous Associate Editor; two anonymous referees; and Angie

Andrikogiannopoulou, Jules van Binsbergen, Vincent Glode, Kris Jacobs, Rachel Pownall, and

Clemens Sialm for comments and suggestions. We also thank conference participants at the 2015

AFA conference, 2013 Lone Star Finance Symposium, 2014 FIRS annual conference, 2014 Confer-

ence on Professional Asset Management at Rotterdam School of Management, 2013 FMA Asian

Annual Meetings, 2014 Mitsui Finance Symposium at the University of Michigan, and University

of Oregon Finance Department 2014 Conference on Institutional Investors; as well as seminar par-

ticipants at American University, Central University of Finance and Economics, Federal Reserve

Board, Harbin Institute of Technology, Temple University, University of Delaware, University of

Houston, University of St. Gallen, and U.S. Securities and Exchange Commission. The views pre-

sented in this paper are solely those of the authors and do not necessarily represent those of the

Federal Reserve Board or its staff. We have read the disclosure policy of the Journal of Finance

and have nothing to disclose.

1For example, Fama and French (2010) estimate that, during the 1984 to 2006 period, active

equity mutual funds underperformed benchmark portfolios by approximately 1% annually,roughly

the average cost of investing in mutual funds.

DOI: 10.1111/jofi.12747

1025

1026 The Journal of Finance R

active growth funds deliver stronger upside potential. We then link variation

in investor preferences to fund flows and show that proxies for investor pref-

erences have significant explanatory power for flows into actively managed

growth and value funds. These findings contribute to our understanding of

investor demand for actively managed mutual funds.

We begin by constructing bootstrapped samples of monthly returns of actively

managed mutual funds and their passive benchmarks to compare their return

distributions.2To isolate the active component of fund returns, we construct

the market-neutral excess returns of active funds over passive benchmarks.

We use the skewness of returns and excess return (idiosyncratic) skewness to

measure the upside potential of active funds. To measure downside protection

of active portfolios, we use co-skewness with market returns, downmarket

betas of excess returns, and excess returns’ loadings on aggregate jump and

volatility risk factors constructed from straddle portfolios of S&P 500 index

options (Cremers, Halling, and Weinbaum (2015), CHW henceforth).3

The results of our analysis of active fund returns and their excess returns

over passive benchmarks suggest that the return distributions of active funds

are quite different from those of their passive benchmarks in terms of their

downside-hedging and upside-capturing properties. While both active growth

and value funds possess some of these return features, their extent varies con-

siderably.On balance, active value funds provide better downside-hedging abil-

ity.Relative to passive benchmarks, active value funds have lower co-skewness

with the market, have lower downside beta, and provide significant hedging

against aggregate jump risk. On the other hand, active growth funds exhibit

more pronounced upside-capturing features. Relative to passive benchmarks,

they have considerably higher skewness and positive idiosyncratic skewness.

In addition, while their overall market beta tends to be high, active growth

funds also reduce beta during market downturns. However, unlike active value

funds, active growth funds do not provide hedging against jump risk.

In addition to using the market portfolio as the simplest benchmark, we

employ two other benchmarking methods. The first benchmark is constructed

based on holdings-matched portfolios. Relative to return-based benchmarks,

holdings-based benchmarks have the advantage of abstracting from factors af-

fecting fund returns that are unrelated to active equity investment strategies

(e.g., fees and cash reserve). In addition, they allow us to control for differences

in diversification (including number of different stocks in the portfolio) and

security characteristics between active and passive funds (Bollen and Busse

(2001), Huang, Sialm, and Zhang (2011)). To construct this benchmark, every

quarter for each active fund in our sample, we replace each of its holdings with

a randomly drawn stock with the same size, book-to-market, and momentum

(SBM) quintile ranks. The returns of this hypothetical portfolio are used as our

passive benchmark. To construct the other benchmark, we follow Berk and van

2We report results with total net-of-fee returns and equity portfolio holding gross returns using

the bootstrap procedure. Our conclusions are not affected by the type of return used.

3We thank Martijn Cremers for providing jump and volatility risk factor data.

Cautious Risk Takers 1027

Binsbergen (2015) and use four-factor model (Fama-French three factors plus

momentum) predicted returns. Regardless of the benchmark used, our results

point to consistent and marked differences in the return distribution across ac-

tive and passive funds and across different investment objectives. The distinct

properties of active fund returns cannot be explained by the potentially differ-

ent characteristics of securities held by active versus passive funds, nor can

they be attributed to lower diversification of active funds’ portfolios compared

to those of passive funds.

Given the observed differences in the returns of active versus passive funds,

we hypothesize that active funds may appeal to investors who value upside po-

tential and downside protection. While we refer to investors with such prefer-

ences by an oxymoron (“cautious risk takers”) in the title, these preferences are

widely supported by experimental evidence (Camerer and Ho (1994), Camerer

(1995), Starmer (2000)). Several theoretical models of portfolio choice demon-

strate that investors who overweight tail probabilities of return distributions

may prefer portfolio returns with limited downside risk and high upside poten-

tial.4Changes in investor expectation, risk attitude, or investor composition

would naturally generate flows in and out of different styles of funds. Based on

the distributional properties of active funds’ returns, we hypothesize that flows

into actively managed value funds increase with investor aversion to downside

risk while flows into actively managed growth funds increase with investor

preference for upside potential.

To explore whether this channel of demand for active funds is theoretically

plausible, we consider a portfolio choice problem of an investor who has access to

a risk-free asset and to active and passive funds with a joint return distribution

as bootstrapped from the data. We assume that the investor maximizes rank-

dependent expected utility (RDEU; Quiggin (1982), Yaari (1987)) with a Prelec

(1998) inverse-S probability weighting function. This weighting function has

two parameters, which allows for a variety of risk attitudes toward tail events,

including joint preferences for upside potential and downside protection.

We solve numerically for the optimal portfolios and examine the comparative

statics of the demand for active management with respect to the parameters of

the probability weighting function. Our results demonstrate that the observed

differences in return distributions between active and passive funds are quan-

titatively relevant to an RDEU investor. In particular, we show that demand

for active value funds increases when the investor has greater preference for

downside protection, while demand for active growth funds increases when the

investor exhibits greater upside-seeking preference. We also use the model to

compute the implied value of active funds and find that an investor would be

willing to pay reasonable fees for access to the active fund portfolios.

4For example, Shefrin and Statman (2000) show that such investors would construct optimal

portfolios with downside protection containing upside lottery-like security. Polkovnichenko (2005)

and Mitton and Vorkink(2007) find that investors with tail-overweighting preferences, while being

risk-averse overall, may invest in less diversified assets to increase the upside potential of portfolio

returns.