KE1068 Funds and Factor Investing
Funds and Factor Investing
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: THEend8_
KE1068
Smart Beta Exchange-Traded
Funds and Factor Investing
It was early 2015 and executives in the iShares Factor Strategies Group were considering the
launch of some new exchange-traded funds (ETFs). !e new ETFs were in a class of ETFs called
smart beta funds. Most traditional ETFs’ portfolio weights were based on the market capitalization
of a stock (stock price times the number of outstanding shares), but smart beta ETFs’ weighting
schemes were based on #rms’ #nancial characteristics or properties of their stock returns.
iShares was a division of BlackRock, Inc., an international investment management company
based in New York. In 2014, BlackRock was the world’s leading asset manager, with over $4.5
trillion in assets under management.1 In 2014, iShares globally o"ered over 700 ETFs with almost
$800 billion in net asset value in the US and $1 trillion globally—a 39% market share, making
iShares the largest issuer of ETFs in the world.2
!e new smart beta multifactor ETFs being considered by iShares would provide investors
with simultaneous exposure to four fundamental factors that had shown themselves historically
to be signi#cant in driving stock returns: the stock market value of a #rm, the relative value of a
#rm’s #nancial position, the quality of a #rm’s #nancial position, and the momentum a #rm’s stock
price has had. While each of these factors existed in di"erent combinations and di"erent forms
in ETFs already in the marketplace, no #rm was currently o"ering these four as a combination
in a multifactor ETF. !e executives at iShares were unsure whether there would be demand in
the marketplace for such multifactor ETFs, since their value added from an investor’s portfolio
perspective was unknown.
Factors
Smart beta portfolios were driven by academic research that shows there is a common set of
driving forces, or factors, that consistently explains stocks’ average returns and systematic risks.
Professors Eugene Fama and Kenneth French have been at the forefront of developing factor models,
having written a series of papers examining di"erent #nancially based factor models of stock returns.
As the culmination of their decades of research, in 2013 Fama and French introduced a model that
showed a systematic relationship between the average returns of stocks and #ve underlying factors.3
Based on the capital asset pricing model (CAPM) theory, the #rst factor Fama and French
considered was the relative covariance of a #rm’s stock returns with a market portfolio. !e CAPM
theory states that stock return movements can be broken down into two components: movements
due to the returns of an underlying market portfolio (a market factor) and #rm-speci#c movements,
with a stock’s average returns determined by its co-movements with the market portfolio. !is can
be seen via the CAPM’s factor model:
( ), , ,i t f i i m t f i tr r r rD E H− = + − + (1)
where ri,t is the return at time t for some security i, rf is the return to a risk-free asset, αi and βi are
regression coe$cients, rm,t is the return to a market portfolio (the market factor) at time t, and
εi,t is the regression’s residual. A stock’s co-movement with the market portfolio is measured by its
CAPM beta (βi), the regression coe$cient above.
Figure 1 shows the average returns on #ve portfolios ranked by the size of their betas using
annual data from 1964 through 2014. !ese portfolios were created by sorting all stocks listed
on the NYSE, AMEX, and NASDAQ by their betas and then splitting the stocks into quintile
portfolios based on the level of their beta, with those stocks with betas in the bottom 20% of
the distribution put into a smallest beta quintile portfolio, then the next 20% into the next
highest quintile portfolio, and so on. Figure 1 shows no evidence of a strong upward sloping
relationship between a portfolio’s beta and its average return, contrary to the prediction of the
CAPM theory. !us, Fama and French concluded that the CAPM theory does not explain average
returns very well.*4
!e second factor Fama and French explored was the size of a #rm as measured by its market
capitalization, what is termed a size factor. Fama and French (and others) found a signi#cant
negative relationship between the size of a #rm and its average return—that is, #rms with small
market capitalizations earn higher average returns across time than #rms with large market
capitalizations. Figure 2 plots the average returns of portfolios grouped into quintiles by their
market capitalizations. What we see in Figure 2 is the small #rm e"ect; small #rms have higher
average returns than large #rms, with a consistent rise in average returns as we go from the smallest
to the largest #rm’s portfolio.
* Note that Fama and French used a longer time frame, from 1927–1990, to construct the portfolios in this study
than were used to construct the portfolios in Figure 1 (1964–2014).