Mutual fund skill in timing market volatility and liquidity

• DOI: http://doi.org/10.1002/ijfe.1580
• Published date: 01 October 2017
• Author: Jason Foran,Niall O'Sullivan
• Date: 01 October 2017

RESEARCH ARTICLE

Mutual fund skill in timing market volatility and liquidity

Jason Foran

1

| Niall O'Sullivan

2

1

Centre for Investment Research,

University College Cork, Cork, Ireland

2

Department of Economics and Centre for

Investment Research, University College

Cork, Cork, Ireland

Correspondence

Prof. Niall O'Sullivan, Department of

Economics, University, College Cork,

Cork, Ireland.

Email: niall.osullivan@ucc.ie

Funding information

Irish Research Council

JEL Classification: G11; G12; G14

Abstract

We investigate both market volatility timing and market liquidity timing for the

first time among UK mutual funds. We find strong evidence that a small

percentage of funds time market volatility successfully, that is, when condi-

tional market volatility is higher than normal, systematic risk levels are lower.

The evidence around market liquidity timing ability is similar although it is

slightly less prevalent compared to volatility timing. Here, funds lower the fund

market beta in anticipation of reduced market liquidity. We also find a positive

relation between liquidity timing ability and fund abnormal performance

where skilled liquidity timers outperform unskilled timers by around 3% p.a.

—though this finding is driven by poor liquidity timing funds going on to yield

negative alpha. However, despite the evidence of volatility and liquidity timing

ability among funds, we fail to find in support of persistence in this timing. We

find little evidence supporting market return timing ability.

KEYWORDS

Liquidity, mutualfund performance, timing, volatility

1|INTRODUCTION

In this paper, we investigate whether UK equity mutual

funds are able to time fluctuations in both market volatil-

ity and market liquidity, assuming that managers attempt

to do so in their investors' best interests. There is a size-

able extant literature on funds' ability to time fluctuations

in market return, in particular, in the US and UK fund

industries, for example, Treynor and Mazuy (1966),

Henriksson (1984), Ferson and Schadt (1996), Jiang

(2003), and Cuthbertson, Nitzsche, and O'Sullivan

(2010). However, less work has been undertaken on mar-

ket volatility timing, for example, Busse (1999) and

Giambona and Golec (2007), while there has been a

dearth of study on market liquidity timing (Cao, Simin,

& Wang, 2013). To our knowledge, we are the first paper

to set about the dual task of evaluating funds' skill in both

market volatility and market liquidity timing in the UK

mutual fund industry.

There are a number of reasons why market volatility

timing is of interest. First, risk‐averse investors are con-

cerned about both risk and return. If funds can decrease

(increase) beta when market volatility rises (falls), they

have the potential to deliver returns with relatively low

volatility. Busse (1999) finds that 80% of his 230 US funds

sample time volatility in this way.

Busse (1999) theorises a probably negative relation

between market return and market volatility. This is

empirically confirmed in the US data, which show a

monthly correlation between the S&P 500 return and

standard deviation of −0.47 between 1985 and 1995. This

incentivises fund managers to reduce the fund market

beta in anticipation of higher market volatility. Busse also

documents a significant relation between volatility timing

and fund performance: funds that reduce systematic risk

when conditional volatility is high earn higher risk‐

adjusted returns. Our UK data show a similar negative

We are grateful for financial support from the Irish Research Council as

well as the Strategic Research Fund at University College Cork, Ireland.

Received: 28 September 2016 Revised: 11 June 2017 Accepted: 20 June 2017

DOI: 10.1002/ijfe.1580

Int J Fin Econ. 2017;22:257–273. Copyright © 2017 John Wiley & Sons, Ltd.wileyonlinelibrary.com/journal/ijfe 257

relationship between market volatility and market

returns: the monthly correlation between the FTSE All

Share return and standard deviation is −0.50 over the sam-

ple period under consideration (January 1997–February

2009). Based on this intuition, if managers can time

market volatility, we expect a negative relation between

a fund's systematic risk (market beta) and market

volatility.

Second, most measures of performance are risk

adjusted (e.g., the Sharpe ratio and multifactor model

alphas). Because risk‐adjusted performance affects fund

cash flows (Kacperczyk & Seru, 2007; Massa, 2003;

Nanda, Wang, & Zheng, 2004), how funds manage risk

has implications for assets under management, fund fees,

and manager compensation. Studies such as Brown,

Harlow, and Starks (1996); Chevalier and Ellison (1997);

and Golec and Starks (2004) show that compensation

incentives affect fund managers' risk choices. Therefore,

fund managers may also attempt to time market volatility

independently of its relationship with market return.

Finally, although stock returns may be unpredictable,

there is persistence and predictability in volatility over

time (Bollerslev, Chou, & Kroner, 1992; Busse, 1999).

Therefore, there is greater reason, a priori, to believe that

fund managers may attempt to time market volatility, if

not market returns.

Cao et al. (2013), a study of the US domestic equity

mutual fund industry, is the only study of mutual fund

liquidity timing. There are also a number of reasons to

study market liquidity timing. First, clearly, liquidity is

of concern to mutual fund managers because an impor-

tant function performed by mutual funds is to provide

liquidity to investors. Second, the 2008 financial crisis

established a link between market‐wide liquidity and

fund performance where reduced liquidity was accompa-

nied by dramatic stock market declines. More formally,

asset pricing literature has identified market liquidity as

a risk factor in asset pricing.

1

As with volatility timing, if

fund managers can anticipate market liquidity conditions,

they can adapt their portfolio exposure accordingly to

alleviate losses and improve performance.

Market liquidity, similar to market volatility, is persis-

tent (Amihud, 2002; Chordia, Roll, & Subrahmanyam,

2000; Hasbrouck & Seppi, 2001; Pastor & Stambaugh,

2003), which again rationalises fund managers' attempt

to time it.

Acharya and Pedersen (2005) develop a theoretical

model of how asset prices are determined by three types

of illiquidity risk.

2

The model shows that because illiquid-

ity is persistent, it predicts future returns and is inversely

related to contemporaneous returns. This is because an

illiquidity shock predicts greater future illiquidity, this

raises future required returns and lowers

contemporaneous prices and contemporaneous returns.

Hence, a market illiquidity shock is associated with low

contemporaneous returns. Following Acharya and Peder-

sen (2005), Cao et al. (2013) contends that as market illi-

quidity shocks are contemporaneously and inversely

related to market returns, fund managers with (il)liquid-

ity timing ability reduce market exposure prior to higher

market illiquidity and we expect a negative market beta/

market illiquidity relation.

The above discussion links the sensitivity of a fund's

market beta to anticipated market volatility and market

liquidity. In a related stream of research, Huang (2015)

examines the relationship between expected market vola-

tility and mutual funds' demand for liquidity. During

periods of market volatility, there is a need for funds to

create a liquidity cushion to manage liquidity risk

(Scholes, 2000). As described by Huang (2015), periods

of high market volatility are associated with higher prob-

ability and magnitude of investor redemptions. Liquidity

mitigates against the adverse effects of investor outflows

by enabling funds to better meet redemptions without

the need to liquidate investments, which can be costly

(Chen, Goldstein, & Jiang, 2010; Nanda, Narayanan, &

Warther, 2000).

Huang argues that to the extent that expected market

volatility can serve as a signal of future redemptions, as

part of a liquidity risk timing strategy fund managers

can reduce the fund market beta (e.g., higher cash hold-

ings) during periods when expected market volatility is

high. Therefore, fund managers may attempt to time mar-

ket illiquidity independently of its relationship with mar-

ket return.

In this paper, we address a number of questions. We

investigate whether market volatility and market liquidity

conditions motivate fund managers' stock selection and

asset allocations decisions and whether managers possess

the ability to time these conditions. We further evaluate

whether managers have private timing skill, that is, an

ability to predict market volatility and liquidity beyond

that which may be predicted by publicly available infor-

mation. Third, we examine whether there is persistence

in these skills—is it possible ex ante to select skilful

timing funds? Finally, we ask whether volatility and

liquidity timing ability predict fund performance. To our

knowledge, we are the first paper to investigate these

questions in the UK fund industry.

We find that a small percentage of funds are skilful

volatility timers and reduce systematic risk in advance of

higher conditional market volatility. A slightly smaller

number of funds are similarly found to reduce the fund

beta in anticipation of market illiquidity. It is clear that

these timing abilities are private in that they remain in

our sample of funds after controlling for the predictive

258 FORAN AND O'SULLIVAN