might expect. I think this is one good source of return, but one of
many. I think the endowment model is a bit overrated; it’s pretty
much an equity-beta, illiquidity premium, manager alpha model.
I have great faith in the first but some skepticism toward the other
two as consistent return enhancers.
Mark Anson: Do you think the endowment market is a crowded
trade because so many large institutional investors are now investing in the style of a Yale or Harvard or Norwegian pension plan?
Antti Ilmanen: Basically, I think every long-only investment is
currently expensive, and therefore the argument for a crowded
trade can be made. The claim of overcrowding is made on credits,
on private equity, on various “smart beta” factors, and surely elsewhere. At the heart of the system is the situation in which global
real yields have been pulled to near zero or negative levels. The
fact is that every long-only investment is fundamentally valued
by summing expected cash flows, divided (or discounted) by
one plus the riskless rate and myriad risk premiums. Now that
the common component in these discount rates—the riskless
rate—is historically super low, it is not surprising that all long-only assets have their real yields at historic lows. So I think this
type of crowding exists, but it’s not really associated with one particular approach, the endowment model or sovereign wealth fund
or any of the approaches I mentioned. It’s happening across the
board, and I think that is the most important challenge in today’s
Edward Baker: In Chapter 29 of your book, you discuss the
advantages long-term investors have, and you mention that insurance selling is something long-term investors should pursue.
Who do you think is the other side of those trades, and is there
a sufficiently developed market to make this a viable strategy for
a long-term investor?
Antti Ilmanen: We could talk about many types of insurance,
but let’s just talk about financial insurance through volatility-selling strategies. I think this is a good source of long-run return; it’s
the flip side of the expensiveness of buying index puts or some other
volatility-buying strategy. Historically, investors have made good
returns by selling volatility or, even more specifically, by selling puts.
The downside is that short-volatility investors lose money during
the worst possible times, so anybody who wants to pursue this
strategy should do it with eyes open. This is a situation in which
investors may be overconfident about whether they can tolerate
the resulting losses. Sadly, many investors actually may have
made both types of mistakes. They were sucked into volatility
selling after the good years of 2005–2006, and then after the horrible experience in 2008, they gave up on this strategy. Then they
may have become interested in buying volatility or hedging tail
risk, but they bled returns for a few years and have since capitulated. Either way, this is a strategy that’s difficult to manage in a
And it certainly seems that the size of “the other side” for selling
financial catastrophe insurance varies over time. I think plenty of
investors would be interested in buying tail-risk hedges, especially
after bad times, but when in 2009–2010 I asked the biggest investors in the world about being that deep-pocket investor who
wanted to sell financial catastrophe insurance, I couldn’t get any
takers at that point. This situation, of course, has subsequently
changed. I take this as further anecdotal evidence of a time-related
inconsistency in investor behavior.
Edward Baker: You cite carry as an example of a good long-term
strategy, but the returns on currency carry have been somewhat disappointing since the financial crisis. Have you changed your thinking about this, or do you still view this as a good long-term strategy?
Antti Ilmanen: The short answer is I think carry may have just had
a bad draw. Every investment can have good and bad draws, and
when we study past performance, it’s not clear whether we should
react to it by expecting continuation or reversals. Actually, we find
pretty weak evidence either way. If you find a good strategy that
has worked in many places, but then it has, let’s say, five bad years,
should you totally de-allocate? The answer is not so clear, and I
would rather err by sticking with a strategy like that.
Incidentally, the currency carry strategy has similar risk characteristics as volatility selling. You can call it picking up pennies in front
of a steamroller, and the steamroller tends to come in bad times. So
if your requirement of a good long-term strategy is that it must
have a good risk-based explanation to make you confident that a
positive long-run reward will persist, it doesn’t get better than this.
People now ask of any strategy after a few bad years, whether the
reward went away. Well, if it went away, we should at least determine whether it went away because the strategy became expensive
or because of some other reason. I don’t recall that at AQR we saw
any clear signs that carry currencies became particularly expensive.
Overall, I would stay with this strategy because of the pervasive
long-term evidence in many different places. I wouldn’t drop the
strategy after a few bad years.
Edward Baker: But outside of currencies, isn’t this carry strategy,
as you define it, really just a form of value?
Antti Ilmanen: I see your point. It’s a fair assessment in many
cases, but it’s a matter of degree. If you think of stock selection,
“If you find a good strategy that has worked in many places, but then
it has, let’s say, five bad years, should
you totally de-allocate? The answer is
not so clear, and I would rather err by
sticking with a strategy like that.”