Once upon a time, we wrote that pretending present and future risk can best be measured by past volatility is akin to driving down the highway with eyes firmly planted on the rear view mirror. (Our theory is the future will not be like the past, so the windshield is a better thing on which to focus.)

The current market upset is the first one featuring the latest generation of so-called risk analytics. These high tech tools generate risk numbers for individual investments as well as entire portfolios. They are entirely based on past volatility. Incorporating frequent updates to the database of past price behavior and enabling near-instantaneous assessments expressed as an index number (say, 1 to 100), they seem quite scientific and highly analytical.

But all of this precision is predicated on the idea that past volatility is a measure of present and future investment risk. So in the latest version, it is like driving down the highway with eyes firmly fixed on an array of rear view mirrors, wide-angle and telephoto and high-definition.

The interesting thing is, when the inevitable downturn occurs, the ‘volatility equals risk’ crowd is quick to point out that their statistical methods are designed to predict what will happen, with a 95% probability. So if you run into a wall while focused on the rear view mirror, there was nothing wrong with the analytics – you just landed in the 5%.

Call us what you want, but a system designed to ferret out risk that fails exactly when you do need it to work may not be all that useful. (An elevator that does not crash to the basement 95% of the time would not be useful, either, in our opinion.)

We will continue to work with our understanding that volatility is a feature of long term investments. It necessitates a long time horizon, so the effect of temporary declines may be mitigated by time. Short term money needs to be kept out of long term investments. And we will keep our eyes focused on the future, not the past.

Clients, if you would like to talk about this or anything else, please email us or call.