ETF Premiums and Liquidity Segmentation

AuthorLouis R. Piccotti
Date01 February 2018
Published date01 February 2018
The Financial Review 53 (2018) 117–152
ETF Premiums and Liquidity Segmentation
Louis R. Piccotti
University at Albany, State University of New York
Exchange traded funds (ETFs) provide a means for investors to access assets indirectly
that may be accessible at a high cost otherwise. I show that liquidity segmentation can explain
the tendency for ETFs to trade at a premium to net asset value (NAV) as well as the life-cycle
pattern in premiums. ETFs with larger NAV tracking error standard deviations (TESDs) tend to
trade at higher premiums and the liquidity benefits offered by foreign ETFs and fixed income
ETFs are revealed to be the most valuable to investors. Further tests validate that TESD has
the desirable properties of a liquidity segmentation measure.
Keywords: ETF premium, liquidity segmentation, limits to arbitrage, tracking error
JEL Classifications: G10, G11, G13
1. Introduction
During the last decade, funds invested in exchange traded funds (ETFs) have
grown significantly from $100 billion in 2000 to $1 trillion in 2010. ETFs are
Corresponding author: Department of Finance, School of Business, Universityat Albany, State University
of New York,Albany, NY 12222; Phone: (518) 956-8182; E-mail:
I thank Richard S. Warr (the Editor) and twoanonymous referees as well as Na Dai, Phil Davies, Stoyu I.
Ivanov, AndrewA. Lynch, Banikanta Mishra, Robert A. Van Ness, Bob Patrick, Pauline Shum, Hongfei
Tang,Yvonne Wang, Daniel Weaver, YangruWu, Eleanor Xu, and seminar participants at Rutgers Business
School, the 2015 Eastern Finance Association Annual Conference, the 2014 New York Accounting and
FinanceForum, the 2014 Northern Finance Association Annual Conference, and the 2014 Southern Finance
Association Annual Meeting for helpful comments and suggestions. All errors are my sole responsibility.
C2018 The Eastern Finance Association 117
118 L. R. Piccotti/The Financial Review 53 (2018) 117–152
investment vehicles that are hybrids between traditional open-end mutual funds and
closed-end funds (CEFs). ETFs invest in a basket of securities and issue shares,
representing claims to the underlying net asset value (NAV) of the fund, that trade
on stock exchanges similar to common stocks. Unlike open-end funds, shareholders
generally cannot redeem shares at NAV directly. Dissimilar to CEFs, the amount
of shares outstanding can vary over time due to the creation/redemption arbitrage
mechanism that is unique to ETFs.
In a frictionless market, the ETF share price equals its NAV. If this were not
the case, then a riskless arbitrage opportunity would exist. To the contrary, ETFs
tend to trade at a premium to NAV (Delcoure and Zhong, 2007). Using a sample of
224 iShares ETFs,1I find that the mean ETF share prices are greater than their mean
NAVs in 89% of the sample months. The fraction of ETFs trading at a premium is
routinely greater than the 1% critical value that would exist if an ETF is equally as
likely to trade at a premium as at a discount. This is a puzzling result in the presence
of capitalized management fees, expenses, and replication transaction costs. ETFs
track passive indices in which case fees cannot be linked to managerial skill as found
in the CEF market by Chay and Trzcinka (1999) and modeled by Berk and Stanton
(2007). Additionally, ETFs do not have open-ending concerns, which is in contrast
to the CEF market (Brauer, 1988; Bradley, Brav, Goldstein and Jiang, 2010).
ETFs provide a means for investors to gain indirect access to a cash flow stream
that may be inaccessible completely or only accessible at a high cost otherwise. If
markets are segmented and ETFs provide liquidity benefits to investors,then rational
investors should be willing to pay a premium to NAV as long as the cost of the
premium is less than the liquidity benefits received. My primary hypothesis is that
the liquidity benefits that ETFs offer can explain the tendency for ETFs to trade at
a premium to their NAV. The evidence that ETFs provide valuable liquidity benefits
is provided by Barnhart and Rosenstein (2010) and Agapova (2011). Barnhart and
Rosenstein (2010) show that investors viewCEFs and ETFs as substitutes by finding
that, when new ETFs are introduced to the market, competing CEFs experience a
widening in their discounts, relative to their NAV. Agapova (2011) finds that ETFs and
conventional mutual funds are partial substitutes with net inflows into conventional
mutual funds decreasing as inflows into ETFs increase.
As a measure of liquidity segmentation, I use the NAV tracking error standard
deviation (TESD), which I define as the standard deviation of the difference between
NAV returns and returns on the underlying basket of securities that it aims to replicate.
In more segmented markets, ETF managers are expected to have less precise tracking
ability due to increased barriers to entry and increased illiquidity costs. TESD is
positively related to premiums with a 100 basis point increase in TESD being related
to an increase in premiums of 13.5 basis points. Accessibility to foreign securities
and fixed income securities through ETFs are revealed to be the most valuable to
1These 224 ETFs roughly had a combined $400 billion in assets under management (AUM) in 2011.
L. R. Piccotti/The Financial Review 53 (2018) 117–152 119
investors. Whereas a 100 basis point increase in TESDs for foreign and fixed income
ETFs is related to an increase in premiums of 16.6 basis points and 48.9 basis
points, respectively, 100 basis point increases in TESDs for domestic and equities
ETFs are related to increases in premiums of 7.8 basis points and 11.8 basis points,
respectively. Increased liquidity segmentation is also associated with a slower speed
of premium correction with a 100 basis point increase in TESD being related to a
premium correction speed that is 2.6 days slower.
However, the decision to initiate an ETF is endogenous. Cherkes, Sagi and
Stanton (2009) model CEF initial public offerings (IPOs) to occur when investors are
indifferent to holding the newly issued CEF at a premium and holding an identical
seasoned CEF at a premium. As more CEFs are issued, the competition drives down
the marginal liquidity benefits offered by individual CEFs and premiums fall to
discounts. I find a similar life-cycle pattern in ETFs with premiums initially being
large and positive and converging to zero as ETFs age. Across ETF types, ETFs that
are less than one year old trade at significantly higher premiums than ETFs that are in
excess of five years old. Likewise,absolute tracking errors for ETFs that are in excess
of five years old are significantly lower than those for ETFs that are less than one
year old. The marginal liquidity benefits of ETFs are greatest for newly issued ETFs.
Finally, I show that the TESD is a robust measure of liquidity segmentation.
Trading costs that increase segmentation such as illiquidity and variancealso increase
TESD. When compared to ETF underlying index return liquidity betas, there is a
strong positive relationship between TESD and liquidity beta, which is not present
in bivaraite comparisons between TESD and other commonly used (orthogonalized)
factor loadings. Further, ETFs investing in securities that trade in foreign markets
have larger TESDs than ETFs that invest in domestic securities; small-cap equity
ETFs have larger TESDs than large-cap equity ETFs; and foreign fixedincome ETFs
have larger TESDs than domestic fixed income ETFs.
Cherkes, Sagi and Stanton (2009) and Ramadorai (2012) are the papers that are
most similar in spirit to my paper. Cherkes, Sagi and Stanton (2009) develop a model
in which CEF discounts are rationally governed by the tradeoff between liquidity
benefits that the CEF provides and capitalized management fees that reduce fund
value. Empirically, theyshow that investors are willing to pay a liquidity premium in
the CEF market, which is consistent with their model. Ramadorai finds that investors
are also willing to pay a premium to NAV for liquidity benefits in the secondary
market for hedge funds.
Similar papers showing that investors are willing to pay a premium to NAV
for liquidity benefits regarding international diversification benefits include Bonser-
Neal, Brauer, Neal and Wheatley (1990), Eun, Janakiramanan and Senbet (2002),
Lee and Hong (2002), and Nishiotis (2004). Bonser-Neal, Brauer, Neal and Wheatley
(1990) find that CEF premiums significantly decrease following a decrease in country
investmentrestrictions. Eun, Janakiramanan and Senbet (2002) also find that investors
are willing to pay a larger premium to NAV if a country has greater indirect investment
barriers. Eun, Janakiramanan and Senbet (2002) and Lee and Hong (2002) show

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