Relative Liquidity, Fund Flows and Short‐Term Demand: Evidence from Exchange‐Traded Funds

• DOI: http://doi.org/10.1111/fire.12159
• Date: 01 February 2018
• Author: Pauline Shum,Markus S. Broman
• Published date: 01 February 2018

The Financial Review 53 (2018) 87–115

Relative Liquidity, Fund Flows

and Short-Term Demand: Evidence

from Exchange-Traded Funds

Markus S. Broman∗

Syracuse University

Pauline Shum

York University

Abstract

We show that highly liquid Exchange-Traded Funds (ETFs), especially those that are

more liquid than their underlying basket of securities (i.e., positive relative liquidity), are

particularly attractive to investors. Using three definitions of liquidity, we find that relative

liquidity predicts net fund flows, as well as inflows and outflows positively and significantly.

We further document a liquidity clientele among institutional investors:(i) relative liquidity is

significantly more important for short- than for long-term investors; and (ii) relative liquidity

is inversely related to investors’ average holding duration in the ETFs. These two findings

provide evidence that relative liquidity encourages short-term demand.

Keywords: ETF, liquidity clientele, institutional investors, short-term trading

JEL Classifications: G10, G14, G23

∗Corresponding author: Finance Department, Martin J. Whitman School of Management, Syra-

cuse University, 721 University Avenue, Syracuse, NY 13244-2450; Phone: (315) 443-4503; E-mail:

msbroman@syr.edu.

We wish to thank the Editor (Richard Warr), an anonymous referee, Kee-Hong Bae, Yisong Tian, Ming

Dong, Susan Christoffersen, Andrei Semenov, Walid Hejazi, Mark Kamstra, Francesco Franzoni, Lee

Cohen, Jaewoon Choi and seminar participants at YorkUniversity and the 2014 FMA Annual Meeting in

Nashville (U.S.) for helpful comments.

C2018 The Eastern Finance Association 87

88 M. S. Broman and P. Shum/The Financial Review 53 (2018) 87–115

1. Introduction

As of May 2017, the global Exchange-Traded Fund (ETF) market has garnered

almost $4 trillion in assets,1outsizing the global hedge fund industry by almost a

trillion,2despite the fact that hedge funds predate ETFs by 41 years. ETF trading

volume accounts for 42% of all equity dollar volume on the NYSE Arca in the first

nine months of 2014.3In addition, many ETFs are more liquid than their underlying

basket of securities; that is, they havepositive relative liquidity.One dramatic example

in terms of the proportional bid-ask spread is the SPDR S&P 500 ETF (Ticker: SPY).

Between 2006 and 2012, SPY was trading at spreads of approximately 18% of those

for its underlying basket.4The unique ETF structure also provides an extra layer

of liquidity, allowing institutional investors to create or redeem ETF shares via the

primary market whenever it is more cost efficient,or whenever there is a demand and

supply imbalance.

In this paper, we examine whether the improved liquidity (if any) of ETFs

relative to its underlying basket encourages trading and the formation of a liquidity

clientele in ETFs. John Bogle, the founder of the Vanguard Group famously warned

that: “Unlike traditional index funds, ETFs trade on an exchange, and can be bought

and sold at any time. That creates a temptation to trade” (Newlands, 2015).

We measure liquidity using three secondary market measures: quoted spreads,

price impact, and turnover. Our analysis focuses primarily on relative liquidity (i.e.,

ETF liquidity minus underlying basket liquidity), because institutional investors have

the ability to choose between investing in an ETF and its underlying basket. For the

retail segment, the transaction costs involved in trading the underlying basket are

prohibitive. Therefore, the liquidity of the ETF itself, rather than the relativeliquidity,

should matter more. In general, our main findings remain robust to using measures

of ETF liquidity as well. To capture the common component across the individual

measures of liquidity,we focus primarily on a composite index of liquidity, calculated

as the first principal component across the three aforementioned individual measures,

for ETF, underlying portfolio, and relative liquidity.

In order to provide a clean comparison of liquidity across ETFs and their un-

derlying securities without time-zone differences, we focus on a sample of 167 ETFs

that are traded in the United States and track only U.S. equity indices. The total assets

under management (AUM) of our sample as of the end of 2012 is $544 billion—

1Including both ETFs and Exchange-Traded Products, from just $426 billion in 2005. Source: ETFGI

database.

2The amount of global hedge fund assets under management was $3.07 trillion in the first quarter of 2017.

Source: Hedge Fund Research (HFR) Global Hedge Fund Report First Quarter 2017.

3“Measuring ETF Liquidity: Looking beyond Trading Volume” SPDR University,2015, SPDR.com.

4Other examples include iShares Russell 2000 (Ticker: IWM), the SPDR S&P Midcap 400 ETF (Ticker:

MDY),the iShares Russell 2000 Growth (Ticker: IWO) and Value(Ticker: IWN), all of which were trading

at spreads of 11–42% of those for their respective underlying baskets.