I’ve been reading a lot of Peter Bernstein lately. Bernstein wrote the book on risk, Against the Gods, along with several others, but my focus has been on his much shorter pieces. One that stuck out was a 2002 speech he gave to the Society of Actuaries.
In the speech, Bernstein makes it very clear that not only do we rely too much on averages, but we ignore the hidden lesson in the data being averaged — variety. Averages lead people to focus on “the whole rather than the doughnut,” as Bernstein puts it. So decisions revolve entirely around the averages.
One way to look at this is that the averages provide certainty where none exist. Basing decisions on averages work for most things, but investing comes with a nasty snag called uncertainty. Averages stop being perfect predictors once an unknown is involved. No matter how far back we go, the future makes asset return data incomplete.
Instead, we should view any output purely through the lens of only for the most part. The averages offer a rough guide of what’s possible but, just in case, we better understand the consequences of being wrong. Because being wrong can be devastating. And history — all that variety overlooked by averages — shows just how often results failed to live up to the averages and investors felt the consequences of being wrong.
Bernstein explains it using Darwinian evolution.
Darwin perceives the future not as a ladder — somehow over time we get better and better and humans are a big advance over the primeval ooze and so on, which is the way we like to think about it — but rather as a bush or a tree that’s growing and there are branches. We don’t know how many branches it’s going to have; we don’t know where the branches are going to have subbranches. We have a general idea of what this thing looks like, but have absolutely no sense of how it is going to develop in detail. We just don’t know, but everything that happens there happens as a result of a cause.
In this Darwinian religion, variation stands out as the fundamental reality, and calculated averages simply become abstractions. This is the key sentence of what I’m trying to tell you — that variation is the fundamental reality, not the calculated averages. Starting from this…it would be an error to view the measure of central tendency as the most likely outcome…
My thesis is that all of us, not just as actuaries and investors, but all of us in a very broad sense, depend far too much on measures of central tendency and therefore are mesmerized by the hole rather than the doughnut…
The trouble with most of us is that we focus on the 95 percent of the results that we believe are not due to chance, because it makes life a lot easier, and that the essence is in the measure of the central tendency. We live with averages, and these days the statisticians have found all kinds of new sorts of averages — arches and garches and God knows how many arches of various types — trends, R2, coefficients of correlations, normal and other species of distribution. There is the theory of probability itself, which really expresses shares of certainty. The theory of probability says we know all the outcomes; the question is just, how are they going to break down, how are they going to be distributed? But we don’t know all the outcomes. What we know is only for the most part, but only for the most part is not everything. Our errors stem from deviations from the averages of the norm, from the outliers, from outcomes never even imagined, and in the past year or so Americans have learned about outcomes never even imagined, but that is the way life develops. In other words, we can’t ignore the possibility of the outliers and indeed should focus on the outliers because only for the most part is not sufficient.
I’m in the investment world, as you know, and we live with this problem in the investment world all the time. We have wonderful long-term series of stock returns, bond returns and so on — Ibbotson, Sinquefield, and so on, and these things are like beacons that lure us into their clutches; we hang onto them. They have sigmas, and sigmas get their due, but what we really like is the averages because the averages lead us to make estimates of what’s normal, but Gould reminds us that there is no such thing as normal. Variation in the history of the capital markets, up and down and up and down, is not — as Gould used the expression — it’s not a pool of inconsequential happenstances. Variation in the capital markets is Darwinian — each episode is a result of the preceding episode. It’s not a series of accidents and therefore this is not a random series, this is history from which we can derive an average and feel some confidence in it. Each episode is a result of the preceding episode.
Goodness knows what we are going through at this moment in the capital markets is a consequence of the foolishness of 1998-2000. And that, in turn, was the climax of a bull market that grew out of the dark days of the 1970s and our ability to overcome the problems of the 1970s: high rate of inflation, the war and the consequences of war in Vietnam and so on. Those high rates of inflation developed from the sense that we were never going to have the Great Depression again and we knew how to deal with those problems and that government could create and maintain full employment no matter what. So that great period grew out of the Depression and the Depression itself was a consequence of what had happened during World War I and the aftermath of World War I and so on, going back to the beginning of time. Each episode is a result of the preceding episode and it makes for a very heterogeneous bunch of stuff…
The trick in risk management is in recognizing that normal is not a state of nature, but a state of transition and that trend is not destiny…
What would life be like if…Darwin was wrong and everything resembled its central tendency? What if we were all clones of just one human being? This would not only be very boring, (I know I’m great, but enough of me is enough) but really very risky. Variety is what makes risk management possible. Indeed, variety is not only the spice of life, but also the essence of survival. Not all disasters happen at the same moment… Diversification of risk matters not just defensively, but because it maximizes returns as well, because we expose ourselves to all of the opportunities that there may be out there. Diversification is not just a defensive step, but to my way of thinking, a strong offensive step.
