I’m not alone in having recently been gripped by the ideas of Yuval Harari (author of Sapiens: A Brief History of Humankind, among other books). Harari gives a fascinating account of the birth of statistics when the Scottish Widows insurance company was started more than 200 years ago.

Here it is in a nutshell.

In 1744, two Presbyterian ministers, Wallace and Webster, wanted to create a solution to safeguard the widows and families of dead clergymen. To do this, they needed to be able to make predictions around such events. Being Scottish (and the practical sort), they consulted a maths professor from Edinburgh and used a recent breakthrough in statistics — Jacob Bernoulli’s law of large numbers — to solve their problem.

While they couldn’t predict when a single individual might die, they could, almost certainly, predict, given a certain population, what to expect across a group.

So, of the 930 Scottish Presbyterian clergyman in any year, 27 were expected to die. Of these, 18 would have widows and five would leave an orphaned child. And so the scheme was launched on the back of these odds. Premiums of between £2 and £6 would secure different returns in the event of death.

They predicted that the fund, by 1765, some 20 years later, would have £58,348 and (drum roll) amazingly, after 20 years they were only £1 off at £58,347. The scheme formed the origins of the very successful insurance company, Scottish Widows (now a subsidiary of Lloyds Banking Group).

What can we take from this? In this instance, the priests were up against several unknown variables, but they were able to work with precision. In aiming to “control the ‘controllables’,” they systematically — mechanically, even — used evidence, stats and figures to address a financial problem.


What does this mean for our finances? Well, it’s very difficult, for instance, to predict how the JSE may perform relative to inflation over a single year, but, if we extend this time frame, the probabilities tighten.

  1. Probability of outperforming inflation over one year: 61%.
  2. Probability of outperforming inflation over five years: 90%.

According to the Dimson-Marsh-Staunton database, the JSE has outperformed inflation since 1901 by more than 7% a year on average — but not so much over the shorter term. Since the start of 2015, markets have only just kept up with inflation (5.26% [JSE] versus 5.07% [CPI]). So far, 2019 has seen a sharp improvement on 2018, but still well below the historical average.

Rationally, it doesn’t make sense for this trend to continue forever. Equity markets should outperform safer asset classes (such as cash) over time, because investors are rewarded for taking the risk. In other words, there’s a sensible risk-based explanation.

For the maths to work in our favour, we need one essential ingredient: time. As US businessman and investor John Bogle put it: “Time is your friend; impulse is your enemy.”

So, when your advisor encourages you to stay invested even though you might recently have been getting poor returns on some asset classes, they’re probably right. The cost of switching based on recent data, (our “recency bias”), could actually stop us from meeting our savings goals.

• Stobie is MD of CoreShares.

Would you like to comment on this article or view other readers' comments?
Register (it’s quick and free) or sign in now.

Speech Bubbles

Please read our Comment Policy before commenting.