Not that long ago, I wrote about the foolishness of listening to market predictions.  A recent piece by Salil Mehta has taken this idea and applied it to a trove of data about historic predictions and their accuracy.  I commend the piece in its entirety for anyone to read.

It’s not pretty. Here is what you need to know.

Over the past 18 years, market returns have been 4.5% annually.  You have not imagined it: we have been in an era of lower market returns relative to historic (back to the 1920s) averages.  Within that average, we have seen variance (as measured by standard deviation) of 19%.  Keep those numbers in focus: 4.5% average return, 19% variance.

The collective predictive powers of Wall Street have been hugely over-optimistic. Their predictions averaged 9% return and 8% variance.  These numbers matter in that an investor relying upon them would believe in a world that is much rosier than we have actually experienced in terms of expected returns (higher is better, naturally) but also in terms of volatility (lower is better; much better).

My favorite part of the piece is in the second half where Mehta reviews the ability of predictions to call the simple direction of a market move (ignoring the magnitude).  In other words, if I say the market will be ‘up’ in 2016 I get the same credit whether the market moves up 1% or 40%.  Collectively, Wall Street predictions had the market moving up 95% of the time.  In our data sample the market actually moves up in 73% of years.

Believing markets move up over long periods of time is sensible, and the basis of a sound financial plan.  What is also sensible is knowing with full confidence that the movements of a short period of time are unknowable.

Bottom line: (1) you need cash in the bank to pay your bills; (2) markets move from moment to moment in a world of their own, one which no one has ever been able to predict with any degree of reliability.