These mechanical rules tend to result, either explicitly or
implicitly, in systematic tilts toward certain factors. Chow et al.
(2011) find that the added value of popular smart beta indexes
is entirely explained by exposures toward classic factor premiums. A fundamentally weighted index, for instance, is an
implicit value strategy, and a minimum-variance index is
essentially a low-volatility strategy. ETFs on a wide number of
smart beta indexes are available nowadays.
Assets in ETFs have been growing strongly, and the end is
not yet in sight. This growth gives rise to the concern that the
practical implementation of factor strategies will become
increasingly costly. Ang et al. (2016) examine this issue in detail
and find that factor strategies remain profitable after costs even
when applied on a huge scale, which suggests that smart beta
ETFs have plenty of capacity left for further growth. Another
concern, however, is that if too many investors start chasing the
same factor premiums, the magnitude of these premiums will
come down. If ETFs, on aggregate, are systematically harvesting
factor premiums, then steadily rising assets in these ETFs
require an increasing number of other investors who are willing
to be on the other side of these factor trades. Such an imbalance
between demand and supply may cause factor premiums to
shrink or perhaps even disappear. In effect, ETF investors would
be arbitraging away factor premiums.
This paper examines whether this concern is justified. Using a
comprehensive sample of U.S. equity ETFs, it finds that many
funds offer a large positive exposure to target factors such as
size, value, momentum, and low volatility. At the same time,
however, many funds also offer a large negative exposure
toward these factors. On aggregate, the exposures toward the
size, value, momentum, and low-volatility factors turn out to be
very close to zero. The finding that ETFs are collectively neutral
on factors argues against the concern that factor premiums are
being arbitraged away rapidly by ETF investors.
DATA AND METHODOLOGY
The sample consists of all ETFs that are listed in the United
States and that invest in U.S. equities, and which have at least
thirty-six months of return history as of December 31, 2015.
This amounts to 415 distinct funds, with combined assets
Some exchange-traded funds (ETFs) are specifically designed for harvesting factor premiums, such as the size, value, momentum, and low-volatility effects. Other
ETFs, however, may implicitly go against these factors. This
paper analyzes the factor exposures of U.S. equity ETFs and
finds that, indeed, for each factor there are funds that offer a
large positive exposure and also funds that offer a large negative exposure toward that factor. On aggregate, all factor exposures turn out to be close to zero, and plain market exposure
is all that remains. This finding argues against the concern
that factor premiums are being arbitraged away rapidly by
investors in ETFs.
This paper investigates if factor premiums, such as the size,
value, momentum, and low-volatility effects, are systematically
being harvested by investors in exchange-traded funds (ETFs).
Some ETFs are clearly designed to harvest factor premiums.
For instance, the Powershares S&P Low Volatility Index ETF
(SPLV) specifically targets the low-volatility premium. Many
other ETFs, however, are not factor-based but based on a different philosophy. For example, a large number of ETFs target a
specific sector. Implicitly this also tends to bring along factor
exposures, although not necessarily in the right direction from
a factor investing perspective. This raises the question of what
kind of factor exposures ETFs exhibit individually as well as
The process of systematically harvesting the premiums offered
by factors that have been thoroughly established in the academic literature is known as “factor investing,” and it is
advocated by studies such as Ang et al. (2009), Bender et al.
(2010), Blitz (2012, 2015), and Ang (2014). In theory, long-short factor portfolios, which capture pure factor premiums and
have a low correlation with asset class risk premiums, are most
attractive for factor investing purposes; see, for example,
Ilmanen and Kizer (2012). In practice, however, factor investing
typically is implemented using long-only strategies, which offer
a combination of market exposure and factor exposure. Smart
beta indexes, which use mechanical rules to deviate from the
capitalization-weighted market index, are a popular example.
Are ExchangeTraded Funds
Harvesting Factor Premiums?
By David Blitz, PhD