Morningstar Magazine - June/July 2018 - 93

more broadly diversified the underlying benchmark,
and the less costly it is to replicate, the better.
Our team's highest-conviction index strategies are
those that are the most sweeping-those
that capture 98% of the market capitalization of
a particular segment of the market and have very
low turnover, because they are more often
than not market-cap-weighted and have fees that
are less than 10 basis points. These are representative of the opportunity set that's available
to active managers in a particular asset class and
almost eliminate all of the various frictions that
active managers face in that asset class.
Cremers: I think strategic-beta funds have a
potentially large role to play. They compete more
with the moderate or lower active share
mutual funds than with the very high active share
mutual funds. The market is evolving, and their
fees are going down.

Finding the Right Measure
When I started at Morningstar quite a ways back,
benchmarking wasn't very sophisticated: A U.S. equity
fund compared itself against the S&P 500. Before
Morningstar, there was no company running numbers
to see whether this fund actually looked at all like the
S&P 500. How much have we improved?

Inker: It's definitely better than it used to be.
But there is a strong desire by institutions
and consultants to have a relatively simple set of
benchmarks for people, like the nine-box grid-
you know, value, core, growth, small, mid, large.
Wait, it's our fault?

Inker: That is a step better than what went before!
You can tell if somebody has a small-cap bias,
and you can measure how well the manager did
given that bias. The most useful thing about
those buckets is that by helping you diagnose the
reasons why a portfolio performed the way
it did, they allow you to potentially be more patient.
You've got to be willing to hold on to things for
a while to have a shot of getting decent long-term
performance, and the style box helps with that.
But there are biases that don't fit very well into

those buckets, and that may drive performance
for some managers. If a manager's approach
involves a big bet on, say, information technology,
you need to know that and expect that when
IT is doing badly, the strategy will do badly.
Again, this is the key thing, whether hiring active
managers or investing in asset classes: The more
your confidence in what you're doing comes from
something other than trailing performance, the
better a shot you've got of having a good outcome.
To go back to passive portfolios, there are a
couple of issues that come in when an index has
significant turnover. Passive managers are
price-takers; they require liquidity from the market.
The more often they are in the position of having to
demand liquidity from the market, the more
it's going to cost them. A low-turnover strategy for
an index is more than just a good idea, it's an
almost essential idea, because the transaction
costs for index funds can wind up surprisingly high
based on the fact that they don't get to make a lot
of choices about when they buy.
With equities, there's a good case to be made
for indexes. For bonds, the appeal is a lot less clear.
For one thing, it is impossible to have a really
low-turnover fixed-income index, given the fact
that these are finite-life securities. The other thing
is that it's not as obvious that the size, the
market cap, of an issuer is a good guide to how
much you should own. If you own a global
government bond index, you wind up with a lot of
Japan because Japan has a lot of debt. That might
not be a good reason to own Japanese government bonds. Among other things, they don't really
yield much of anything. In the extreme case,
if they have a lot of debt relative to GDP, they may
have trouble paying back their bonds and have to
issue more-and you wind up owning more.
That's a good wrap-up. Any other final words?

Johnson: I agree that there are important
limitations to fixed-income indexes, especially to
the extent that you're designing an index
that will underpin an investable product. The
poster child is the high-yield space. The current

menu of high-yield ETFs is underpinned by indexes
that aren't fully representative of that particular
asset class, owing to liquidity constraints. We have
a pretty dim view of these products.
Conversely, we have a favorable view on a number
of different active managers who are able
to more nimbly navigate those markets. Some of
them have been pointing to opportunities that
have been created by the inclusion or exclusion
effects that are being driven by the growth of
indexed products within these same marketplaces.
Cremers: I'll close with some thoughts on
benchmarking. Research has suggested that most
fund flows can be best predicted by how
funds have performed relative to a benchmark
without adjusting for risk. A caveat, though, is that
some academic benchmarks are not really
investable, and may not be representative of the
stocks that active managers can actually buy.
For example, some of these academic benchmarks
are dominated by small value stocks that are
largely outside of the universe for most managers.

In my new research, we compare the prospectus
benchmark to what we call the minimum
active share benchmark.7 Those minimum active
share benchmarks tend to be riskier than
the prospectus benchmarks. We estimate that
about 25% to 30% of funds have a prospectus
benchmark that is meaningfully different than the
benchmark that better captures their investment
strategy. Most of these funds have high active
share. The typical fund in this group is significantly
more risky than its benchmark, so in some sense
the benchmark is relatively easy to beat.
Johnson: Martijn, I'm surprised that such a small
sample among the universe of active managers
would self-select such a low hurdle for themselves!
I'll close on that benchmarking joke because it will
probably be quite a while before our next joke comes.

K
John Rekenthaler is vice president of research with
Morningstar Research Services.

7 Cremers, M., Fulkerson, J., & Riley, T. 2018, "Benchmark Discrepancies and Mutual Fund Performance Evaluation. Working paper: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3173437.

global.morningstar.com/Morningstarmagazine

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3173437 http://global.morningstar.com/Morningstarmagazine

Table of Contents for the Digital Edition of Morningstar Magazine - June/July 2018