The Super Bowl Stock Predictor has been correct seven years on the trot. What’s more it has correctly anticipated the future direction of US stock markets in 40 of the last 49 of them – thus sporting an impressive 82% accuracy rate.

If it gets it right again this time, some say we can expect a good year for stocks if the Carolina Panthers win, but a miserable one if the Denver Broncos come out on top.

So how does it work? Surely, it must be a complex algorithm. Well, firstly one looks to see if the event is won by a team from the National Football League (NFL) as opposed to the American Football League (AFL). The Super Bowl always features one team from each conference (this year Broncos AFL, Panthers NFL), so it has to be one or the other. You then apply this formula:

• An  ‘NFL win’ will see a stock market rise over the coming year
• An ‘AFL win’ will see the market fall

OK, so perhaps the brainchild of veteran Wall St analyst Robert H. Stovall isn’t so difficult to grasp after all.

Of course, Stovall’s real point has always been to highlight the fact that clever financial models used to inform investment decisions may not always be all they’re cracked up to be. More precisely, just because two things appear at first glance to be correlated, it doesn’t necessarily imply a causal link. Of course, no sane person would suggest the Super Bowl Stock Predictor should be taken seriously.

The more serious point behind Stovall’s bit of fun is that investors often trip up by making rather less crazy-sounding assumptions about correlations that don’t actually exist. Many people assume, for example, that stock market returns will be higher in countries where economic growth is strongest.

Growth seems to have an almost irresistible allure for investors. But historical data shows that there is actually a slight negative correlation between real per capita gross domestic product (GDP) growth and real equity returns. In other words, if anything, equity returns are actually worse in rapidly growing economies!

Between 1900 and 2009, Japan’s annualised economic growth rate was the highest of 18 major OECD countries, and more than 50% higher than Canada’s, yet Canada’s annualised equity returns were approximately 50% higher than those of Japan over the same period.

Nor are individual businesses inherently safer because they operate in an economically robust environment. It may turn out less risky to own a high-quality company with a strong balance sheet in a low growth economy than a mediocre or heavily indebted one in a high growth one.

The most important factor in determining the long-term equity returns is the relationship between the price paid for a share and its “intrinsic” value – the price a rational businessperson with all the facts would pay for the business. Even though intrinsic value does not change much over short periods of time, share prices can and do fluctuate wildly along with extremes in investor emotions.

And talking of emotions, they are bound to be running high in America this weekend. But sadly, the Super Bowl result won’t guarantee success investing on the stock market.

Dan Brocklebank is a director of Orbis Access