When fund management skill is more valuable?

Published date01 March 2020
AuthorJohn A. Doukas,Feng Dong
DOIhttp://doi.org/10.1111/eufm.12234
Date01 March 2020
Eur Financ Manag. 2020;26:455502. wileyonlinelibrary.com/journal/eufm © 2019 John Wiley & Sons Ltd.
|
455
DOI: 10.1111/eufm.12234
ORIGINAL ARTICLE
When fund management skill is more
valuable?
Feng Dong
1
|
John A. Doukas
2,3
1
Siena College, School of Business,
Loudonville, New York,
Email: fdong@siena.edu
2
Old Dominion University Graduate
School of Business, Norfolk, Virginia,
Email: jdoukas@odu.edu
3
Judge Business School, University of
Cambridge, Cambridge, UK
Abstract
Does fund management skill allow managers to identify
mispriced securities more accurately and thereby make
better portfolio choices resulting in superior fund
performance when noise trading a natural setting to
detect skill is more prevalent? We find skilled fund
managers with superior past performance to generate
persistent excess riskadjusted returns and experience
significant capital inflows, especially in high sentiment
times, high stock dispersion, and economic expansion
states when price signals are noisier. This pattern
persists after we control for lucky bias, using the false
discovery rateapproach, which permits disentangling
manager skillfrom luck.
KEYWORDS
fund manager skill, investor sentiment, mispricing, mutual fund
performance
JEL CLASSIFICATION
G11; G14; G20; G23
EUROPEAN
FINANCIAL MANAGEMENT
We thank two anonymous referees, European Financial Management Association (EFMA) 2018 conference
participants, and Russ Wermers for valuable comments and the Conference Board for kindly providing us with the
sentiment data. We also are grateful to Jeffrey Wurgler and Malcolm Baker for making their sentiment index publicly
available. We also thank Stambaugh, Yu, and Yuan (2012) for providing their stock mispricing data.
noise creates the opportunity to trade profitably, but at the same time makes it difficult to trade profitably. (Fisher Black,
1986, p. 531).
1
|
INTRODUCTION
Apparently increasing evidence has accumulated regarding the extent to which financial
markets are informationally inefficient as a result of investor cognitive biases related to the
noisiness of information (Hilbert, 2012; Tversky & Kahneman, 1974). Undoubtedly, noise
trading activity is mostly prevalent when markets are crowded by sentimentdriven noise
traders and it has been held responsible for much of the market uncertainty during the last two
decades. However, its impact on the performance of actively managed mutual funds remains
unknown.
1
Shedding light on this issue is very important given that the bulk of investment
activity has been delegated to professional fund managers. Hence, in this study we examine the
extent to which market sentiment, as the most prominent state of market noise relative to
others (i.e., high market dispersion states; economic expansions; stock mispricing, using a set of
11 market anomalies to identify overpriced stocks (Stambaugh et al., 2012); and market
volatility), affects the performance of mutual fund managers and, especially, that of skilled fund
managers? As Black (1986) argues, noise traders and professional investors with low levels of
skill cannot discern valuable information from market noise and may treat market noise as
useful information and trade based on it. We address this question by investigating whether
market sentiment influences fund alphas, as noted by Miller (1977), which, in turn, based on
the insight of Black (1986), makes it less (more) difficult for high(low) skilled fund managers
to carry out profitable trades. Specifically, we examine the capacity of US domestic equity fund
managers to add value during periods of high sentiment, used as an acidtest of management
skill, when it is more difficult to identify profitable stocks. Unlike previous studies, we assess
the added value of fund skill in high sentiment periods conditional on fundspast performance,
capital inflows, high market dispersion, economic expansions, equity mispricing, based on a set
of 11 market anomalies to identify overpriced stocks (Stambaugh et al., 2012), and market
volatility conditions that make it more (less) likely for high(low) skilled fund managers to
improve and sustain fund performance.
While a large body of the literature arrives at the conclusion that actively managed funds, on
average, underperform passively managed funds, recent papers find that funds with certain
characteristics can outperform the market benchmark.
2
However, previous studies do not
examine whether fund managers create added value when markets are crammed by high levels
of noise as a result of investor sentiment, which is widely recognized as a market state of high
noise trading activity. Other studies document that active fund manager skill varies over time
with macroeconomic conditions (e.g., Avramov & Wermers, 2006; Christopherson, Ferson, &
Glassman, 1998). This literature emphasizes that fund managers may have particular skills in
managing stocks during a certain phase of the business cycle.
3
Whereas these studies show that
fund alpha in various industry sectors is related to real economic forces, they do not shed light
on the important question whether fund alpha is linked to fund managersskill, particularly in
high investor sentiment states of the equity market when high(low) skilled fund managers are
expected to execute more (less) profitable trades. Unlike the previous literature, this paper aims
1
Campbell and Vuolteenaho (2004) separate stock market returns into that due to changing forecasts of future cashflows (which tend to be nonmeanreverting)
and changing forecasts of market discount rates (which are related to investor sentiment and are meanreverting). They find that changes in discount rate
forecasts (i.e., changes in investor sentiment) are a major driver of market volatility.
2
For example, Kacperczyk, Sialm, and Zheng (2005), Cremers and Petajisto (2009), and Cremers, Ferreira, Matos, and Starks (2016) document that industry
concentrated portfolio holdings or a large deviation of holdings away from a benchmark can outperform such a benchmark, net of expenses.
3
For instance, a technology manager may produce a higher alpha when shortterm interest rates are high and the credit spread is low, while a consumer staples
manager may produce alphas when short rates are low and the credit spread is high.
456
|
EUROPEAN
FINANCIAL MANAGEMENT
DONG AND DOUKAS
to quantify whether fund management skill delivers high alpha during high sentiment times,
when more managerial talent and resources (i.e., costs of gathering and processing private
information is higher) are required to separate real information and market noise. That is, when
the signaltonoise ratio is lower. While a higher level of investor sentiment presents a more
difficult environment for all active managers to pick valuable stocks and sectors, it should be
less daunting to skilled fund managers who possess the ability (better informed traders) to
discern noise from real information.
As explained by Black (1986), noise trader participation in the market, triggered by investor
optimism, can move asset prices away from fundamental values. In fact, since investor
sentiment has been shown to influence noise trader investment behavior and, in turn, asset
prices (Białkowski, Etebari, & Wisniewski, 2012; Dowling & Lucey, 2005; Edmans, Garcia, &
Norli, 2007; Hirshleifer & Shumway, 2003; Kaplanski & Levy, 2010; 1977), it stands to reason
that skilled fund managers would outperform the market benchmark and their lowskilled
counterparts by being able to better discern information from noise when the market is
crowded by noise traders. Additionally, since noise trader activity is asymmetric across
optimistic and pessimistic sentiment periods, unsophisticated investors are more likely to enter
the stock market during high sentiment times (Grinblatt & Keloharju, 2001; Lamont & Thaler,
2003). Moreover, based on the insights of Stambaugh et al. (2012) and Antoniou, Doukas, and
Subrahmanyam (2013), noise traders are more likely to drive prices away from intrinsic values
by being reluctant to realize losses (sell losers) during periods of optimism to avoid regret and
cognitive dissonance. Therefore, to the extent that skilled fund managers trade more on
(private) information about the true value of financial assets under management (AUM), in
contrast to the market benchmark and their lowskilled counterparts, they are expected to
deliver more value during high sentiment periods. In sum, the important question whether
fund managersperformance is affected by noise trading activity, prominent in high investor
sentiment times, a natural setting to detect if fund managers possess skill, warrants
investigation. To shed light on this issue we posit that the variation of skill across fund
managers is a key determinant of fund performance (alpha).
Moreover, in contrast to the previous literature that examines whether fund managers try to
exploit investor sentiment by deploying sentimentbased (timing) strategies in order to attract
capital flows (Massa & Yadav, 2015) or whether funds tilt their portfolios toward better
performing stocks when they buy (sell) stocks that are highly sensitive to market sentiment,
preceding an increase (decrease) in investor sentiment (Cullen, Gasbarro, Le, & Monroe, 2013),
in this study we treat sentiment as a market condition, not as a risk factor that skilled managers
actively time investor sentiment by modifying fund strategies based on their sentiment
prediction.
4
While our evidence is consistent with the previous literature showing that skilled
fund managers outperform their lowskilled peers across time (Amihud & Goyenko, 2013; Berk
& van Binsbergen, 2015), we distinctly document that fund managersstockselectivity skill is
more profitable during high than low sentiment periods due to short selling limitations (Shleifer
& Vishny, 1997) and costlier valuerelevant information that permit asset prices to drift away
from intrinsic values. Unlike Kacperczyk, Nieuwerburgh, and Veldkamp (2016), who argue that
the timevarying fund performance is caused by fund managersoptimally choosing to process
information about aggregate shocks in recessions and idiosyncratic shocks in booms, we treat
investor sentiment as a noisy market condition which allows us to determine whether skilled
4
Specifically, Massa and Yadav (2015) consider the preferences of fund managers for holding stocks that react in a contrary manner to the level of investor
sentiment or display a contrarian sentiment behavior.
DONG AND DOUKAS EUROPEAN
FINANCIAL MANAGEMENT
|
457

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT