trepidation, and the text began: “The first time I met Antti, I
thought he was insane and I was right.”
4 At that point, I was
sweating like a pig. But then he turned the comment around,
saying I’m insane in a good way, and it all worked out fine, but
that was a serious scare.
Margaret Towle: Over the years, we have seen an evolving role for
investment consultants, as well as a distinction between consultants
who serve institutional clients and advisors who are wealth managers serving individual investors. From your vantage point, as someone with both an academic background and experience as a hedge
fund manager, what do you see as the appropriate role of investment consultants, for those serving institutional clients and those
serving individual investors?
Antti Ilmanen: Actually, my answers are pretty similar. I think
the critical role for both types of advisors is educating clients
about what’s important and what’s feasible and what’s not.
Consultants should avoid overpromising with regard to expected
returns. They should also help investors avoid the classic bad
habits of chasing multiyear returns, overconfidence, microman-agement, and hindsight bias. Another important task for consultants is explaining how slowly we learn from performance data.
This may be something that’s impossible to square, but we learn
so slowly that common performance evaluation windows of three
to five years are likely to hurt investors because at that horizon
good performance is more likely to be followed by reversals.
Cliff has said the problem isn’t that investors chase returns but
that they chase returns at reversal horizons. Following the multi-year performance is thus a bad idea, so at AQR we’ve thought
hard about what to do about that. It’s a difficult problem to solve
because suggesting that investors wait ten years, twenty years, or
more is rarely a realistic solution.
Promoting good habits—including patience, education through
various means, avoiding line-item thinking, not too frequent performance evaluation—these are all ideas I would emphasize. One
particularly bad habit is underdiversification. This problem can be
highlighted in many ways, but I would say one, perhaps controversial, way is that most investment portfolios have too much concentration in equity risk. We are in an investment world where multiple returns or risks can be exploited, and yet most portfolios today
(e.g., 60/40), have more than 90 percent of the risk coming from
equities. To adequately address this problem, portfolios need to use
some leverage to make other asset classes or market-neutral strategies matter nearly as much as equities. I would say the consultant’s
role includes challenging certain constraints, particularly if this can
help long-run returns.
One of AQR’s founding partners, David Kabiller, has long said that
investment success requires good investment strategies and good
investors.5 So we’ve tried to do something on the latter front as
well. Last year, in this spirit, we began an interview series called
“Words from the Wise” among senior thought leaders. Many of
these leaders were in your Masters of Finance compendium. It’s still
early going for us, but if the project sounds familiar to you, all I can
say is that imitation is the sincerest form of flattery.
Margaret Towle: At this point, our committee members have additional questions for you.
Geoffrey Gerber: More than twenty years ago, you argued
that bonds appear to have a positive risk premium but mainly
at the front end of the curve; beyond two years, you noted that
it was unclear whether extending duration increased expected
returns at all. Given the significant decline in the bond risk premium over the past twenty years and the historically low interest
rates since the financial crisis, how would you assess the bond
risk premium today?
Antti Ilmanen: First, my expectations from prior decades had
some micro-aspects of the shape of the expected return curve. At
very short maturities, Treasury-bill buying by the central banks
that lacked return-seeking incentives made the front end of the
curve very steep. Then at the back end we saw liability-matching
pension funds that were paying a premium for long-duration
bonds. Today, there may still be some of that return-insensitive
central bank buying at the front end but much less so. Overall,
I think the shape of the curve could have become more linear.
But you are really asking about the low level of bond risk premiums
today, and, according to many measures, they are near historic
lows. We can see this from real yields that are negative, but I prefer
to look at the gap between nominal yields and average expected
T-bill rates over the next five or ten years, and that also is negative.
They may suggest that bonds are expensive, but I think bonds may
remain sustainably expensive for several years ahead.
From a financial perspective, in a low-growth, low-inflation
environment, government bonds have been safe-haven assets—
negative-beta assets versus equities—for the past fifteen-plus years.
The capital asset pricing model then implies that bonds should be
expected to earn negative returns over cash. And that is before we
consider the exceptional demand that comes from pension funds,
notably for liability-driven investing, and, more recently, quantitative easing by the central bank. I think all of these factors can keep
bonds expensive for quite a while.
On the positive side, there still is a moderately steep yield curve, at
least in some countries. You thus get positive carry and so-called
rolldown gains, which historically have been better predictors of
near-term bond returns than value indicators related to yield level.
Overall, I can still see U.S. Treasury bond yields at 1.5–2 percent
offering a pretty symmetrical outlook. It’s much harder to tell a
positive story for the German Bunds and Japanese government
bonds at zero to slightly negative yields because you have to rely on
the roll-down and assume that yields just stay at these levels for