Morningstar - Summer 2019 - 80

Investors

Ibbotson: I disagree with that, by the way!
I think it was new, and I think it was clearly all
about equilibrium, and from my perspective,
it really was a theory. But Paul really made a great
contribution by formalizing it; the theory
wasn't formalized enough.
Kaplan: The authors lacked any kind of a
formal, academic-style model. Instead, they had
some charts.
Ibbotson: A graphical equilibrium, essentially.

Formalizing Popularity
Kaplan: I came up with this idea: Let's start
with the assumptions of the CAPM, and let's add
one additional assumption: that investors
don't care only about risk and return. They care
about many other things as well-what Roger and
Tom had been calling the dimensions of
popularity. I went through the mathematical
development of the model, wrote it up as a paper,
and circulated it among several people, including
Roger. As I recall, Roger, your reaction was
that this was the kind of mathematical, rigorous
equilibrium theory you had in mind when
you wrote the original paper, but that there was
no one that had the math skills to put it together
at the time.
Ibbotson: Yes. The CAPM is a great framework,
because it puts everybody in a risk-averse
framework, but the PAPM adds something to that.

We set up an example with three investor
types and five securities. One of the investor types
is a CAPM investor who doesn't care about
popularity; he or she only cares about risk. That
investor tries to take advantage of the way
things are priced through popularity. From that
investor's perspective, the most popular securities
end up being overpriced, with the lowest
expected returns, and the least popular securities
end up being underpriced, with the highest
expected returns. This CAPM investor effectively

takes long positions and levers up in these
underpriced securities and then shorts overly
popular securities.
The key finding is that prices are formed by
the aggregation of the demand of all the investors.
So, popularity gets into the prices and ultimately
gets into the pricing model. Therefore, that's what
we have-the PAPM.
Kaplan: After having circulated this paper,
I was in Chicago one evening having dinner with
Larry Siegel. He's familiar with all three
of the papers we've been discussing, and he
asked us if we would want to put them
together into a CFA monograph. Of course,
we said yes.6
Idzorek: Meanwhile, throughout this time,
James was doing a lot of our empirical
heavy lifting as we searched for new and different
data sets in which to see if we can see the
popularity phenomenon at work. We had a wealth
of stuff that had been written and studied,
all in a hodgepodge. A CFA monograph seemed
like a wonderful home, a way to bring it
all together.
Rekenthaler: James, as I understand it, your role

in this collaboration was to look at some of the specific
factors of popularity. Is "factors" the right word?

James Xiong: You can call them factors.
Popularity is a kind of umbrella. It can include
a bunch of factors or characteristics. The
empirical research started back in 2014, before
the popularity and asset pricing paper for
the Journal of Investing was published. We looked
at a bunch of potential popularity-related security
factors and characteristics.

The first focus was on brands. Interbrand
publishes the Best Global Brands report every year.
We got data starting from 2000 of the best
100 brands each year. Google GOOGL was among
the lowest on the list in 2000. In 2018, it
was second, with Apple AAPL first. Back in 2000,
they were not among the top ones. We

found a very good empirical result, consistent
with the hypothesis of popularity.
The second focus was the Morningstar
Economic Moat Rating. Morningstar analysts
rate economic moats and sort companies
into one of three moat categories: wide, narrow,
or none. That's the second variable that
we tested.
The third one was the Harris Poll Reputation
Quotient, which rates the reputations of
different companies. It's similar to the brands list
and updated annually.
We added two other characteristics related to
skewness. Typically, investors prefer lotterylike
stocks, stocks with a very positive skewness. And
they don't like the kind of stocks that lose
even more than the market when the market is
down, stocks that have negative coskewness.
Those are the five characteristics for which
we looked at empirical evidence, in chapter six
of the book. Chapter seven is based on the
work of Roger and [Zebra Capital Management
research director] Daniel Kim, and looks at
traditional factors, such as size, value, liquidity,
and momentum. The results are pretty consistent
with the popularity hypothesis.7
Ibbotson: In the research that James referred to,
Daniel Kim and I were looking at premiums
that were already well known and showing that
they're consistent with the idea of popularity. James looked at new tests that are specifically
targeted toward measuring popularity,
such as brand, which obviously has a clear link
to popularity. James was able to confirm the
popularity concepts in these new areas of research.
Idzorek: I think "factors" and "characteristics" are
appropriate terms. I also like "dimensions"
of popularity. These are either company or security
characteristics. There are various premiums and
anomalies in the marketplace that are well
researched. We went on a quest for new proxies

6 Ibbotson, R.G., Idzorek, T.M., Kaplan, P.D., & Xiong, J. X. 2018. Popularity: A Bridge Between Classical and Behavioral Finance, CFA Institute Research Foundation.
7 Ibbotson, R.G. & Kim, D. Y.-J. 2017. "Risk and Return Within the Stock Market: What Works Best?" Zebra Capital Management, White Paper.

Morningstar Summer 2019

Morningstar - Summer 2019

Table of Contents for the Digital Edition of Morningstar - Summer 2019