With a spectacular 2013 in the record books, stock market forecasters have been busily looking forward to what 2014 might bring.
You should avoid listening to them.
Simply put, market forecasters make weather forecasters look good. We've all been led astray by the weather man from time to time, but his short-term predictions are right more often than not.
Regrettably, the opposite holds true of stock market forecasters, who tend to be wrong more often than they're right.
My biggest beef with market forecasts is that they're rarely accompanied by any indication of the tremendous uncertainties involved.
To get a sense of just how much the market can swing from one year to the next it's useful to consider the historical record. Yale professor Robert Shiller maintains a database of information on the S&P 500, which helps to illuminate matters.
Prof. Shiller's data show that the S&P 500 gained an average of 9 per cent annually from 1872 to 2013, including dividends. The very long-term gains are phenomenally good and they represent a big reason why investors like to buy stocks.
However, the average return doesn't provide a sense of how wild the ride was on a year-by-year basis. If you've been investing for any length of time, you know that the market's swings can be jarring.
For instance, just look back to 2008 when the market lost 39 per cent. While that stretch might have been particularly bad, history is littered with other difficult periods.
On the bright side, the market has also seen some excellent times. Investors were quite pleased in 1954 with an advance of 48 per cent and 1995 saw a climb of 38 per cent. (The year just past was no slouch, as the S&P 500 shot up 32 per cent and the S&P/TSX Composite climbed a respectable 13 per cent.)
One way to try to encapsulate such a range of possibilities is to present them much like the results of a political poll.
You've seen the format: a poll might say, for instance, that one party is supported by 40 per cent of the public, plus or minus three percentage points, 19 times out of 20.
In other words, they should get between 37 per cent and 43 per cent of the vote except for that bothersome one election in 20.
Using the pollsters' method for the markets isn't a perfect one because the market is even more unpredictable than voters, which skews the statistics. More specifically, stocks tend to go to extremes more often than even the worst partisans. As a result, the following results probably overstate the level of certainty involved.
According to Prof. Shiller's data, when you knock out the one in 20 most extreme annual periods from the S&P 500's record, its returns have ranged between a loss of 26 per cent and a gain of 45 per cent.
Put in the language of the pollsters, the index generated an average annual return of 9 per cent, plus or minus roughly 35 percentage points, 19 years out of 20.
About once every two decades it exceeded these limits and – as you might imagine – such periods tend to be quite memorable. In addition, such extreme years often come in clumps, which can be a good thing because large downturns tend to be followed by big turnarounds.
The huge amount of annual return variability is a disquieting fact and it makes predicting what the market will do in 2014 a mug's game.
On the bright side, most investors don't put their money in the market for only one year. As they keep it invested longer, the big annual gains and losses should average out to provide less extreme long-term results.
In other words, it's usually best to take a much longer view and avoid short-term speculation – as well as those annoying forecasters.