Edward Baker: How much of this is an artificial result of current
central bank policy, which seems anomalous relative to history?
Antti Ilmanen: Central bank policy is clearly part of the picture,
but it’s not the only factor influencing bond investments. Many
claim that low bond yields are distortions exclusively caused by
ultra-loose monetary policies; but viewed purely from the macro
side, factors such as dwindling inflation premium, worries about
secular stagnation or insufficient demand, and the negative-beta, or
safe-haven, aspect—these factors are all apart from the central
bank buying. By the way, when yields eventually start to rise, I
think there will be lots of liability-driven investors who will be buying bonds. This also tells me that bond risk premiums may stay low
for quite a while.
Edward Baker: I found your book extremely interesting and useful, and one aspect that struck me had to do with the risk factors
you identified. You identified only four, and I was a bit surprised at
what you chose: growth, liquidity, inflation, and tail risks. Can you
comment on why you settled on four and whether there were others that you considered but in the end did not include?
Antti Ilmanen: The main reason for the cube illustrations in my
book was to argue that investors should use three dimensions to
evaluate their portfolios: asset class, strategy style, and underlying
risk factors [see figure 1]. Some investors think about asset class
risk premiums as major factors. Others think about smart-beta
premiums or style premiums, and others think about deeper,
non-investable fundamental factors. I think all of these perspectives
are valid and complementary, and I wanted to highlight all of them.
However, picking individual factors is harder. On the third side of
the cube, where I considered the uninvestable fundamental macro
factors, growth and inflation were pretty obvious influences to
include. Then I was thinking about liquidity, real yields, and monetary policy—all overlapping concepts—and separately about volatility, tail risk events, and financial crises—again overlapping concepts. Although purely discretionary, I chose one factor from each
of these pockets and discussed that foursome in the book. Among
the other candidates, real yield and monetary policy might also be
included on the list.
Edward Baker: Have you put together a formal definition of tail risks?
Antti Ilmanen: Ultimately, tail risks depend on what’s important in
your portfolio. Given that equity risk dominates the portfolios of
most investors, I would say that equity-related tail risks are most
important. The second most important risk could be bond-related.
There’s been a lot of work on tail risks, and the most common way
investors think about hedging tail risk may be through index put
buying. However, when you look at historical data, this is roughly a
minus-one Sharpe ratio strategy.
There are some other candidate strategies, and one of my favorites
is trend-following. Trend-following has a clear positive Sharpe
ratio, and it has done well in most of the historical bear markets
over the past one-hundred years. The way I initially explained this
to myself and others was that bear markets tended to be protracted
and gradual affairs, which allowed a trend-follower to turn from
bullish to bearish and ride the bear market. In contrast, if we see
fast bear markets and fast crashes, the only reliable protection is
through index puts. It’s interesting that financial markets offer fast-crash protection so expensively (as reflected in a quite negative
long-run Sharpe ratio) and allow another strategy, which provides
slow-crash protection, to perform quite well in the long run.
Finally, the Brexit vote reminded us that even when a fast crash
occurs, trend-followers can do pretty well as long as the big negative move doesn’t happen at the turning point just after a big rally.
If it happens during an existing downward trend, as it did with
Brexit or Lehman Brothers, the trend-followers will already be
positioned on the right side of the move. My key message to
investors when it comes to managing tail risk is to go for the cost-effective way of buying tail protection rather than the expensive way.
Edward Baker: Obviously, liquidity is important for trend-followers
if they’re going to make a dynamic shift along with the reversal
of the trend.
Antti Ilmanen: Typically, trend-following employs highly liquid
investments in futures and major currencies. Tail-risk hedging generally tries to protect against directional market falls. In these situations, trend-following has tended to increase equity short positions
and empirically related positions in other asset classes, e.g., long-duration positions in fixed income, anti-carry in currencies, and
Figure 1: The Cube: Asset Class (front), Strategy Style (top),
and Risk Factor (side) Perspectives on Investments
Source: Ilmanen (2011)