Probabilities Versus Predictions

Last week, South Korea stunned the football world by knocking out World Cup favourites Germany. In an astonishing finish, South Korea kicked two goals within minutes of the final whistle during extra time, in one of the biggest upsets in the sport’s history. Why? Because Germany were expected to take out the 2018 Fifa World Cup.

Here’s the 2 minute wrap up of the match courtesy of SBS:

It wasn’t only the football world who expected the German’s to take the cup home, it was also the expectation of UBS’ analytical team who ran complicated statistical models to place probabilities on all nations competing in the World Cup. Here’s the report if you’re curious.

Following Germany’s loss to South Korea, UBS have been copping criticism from journalists and social media trolls, about their inability to predict or forecast the future. Individuals’ and companies’ inability to forecast the future is well documented and certainly not news to anyone that studies the market, no matter how sophisticated they are or their technology is.

Let’s get one thing clear, UBS nor any of the other investment banks “predicted” Germany would win the World Cup. They simply applied a 24% probability of winning, in other words, a 76% probability of not winning – there is a huge difference.

“We are humble enough not to outright claim that Germany will win the tournament again, but our simulations indicate there is no other team with higher odds to lift the trophy than the defending champion.” – UBS (emphasis mine)

As nerdy and as absurd as this analysis may seem, what else do you have to rely on? Your gut feel? The tip your taxi driver gave you? Your “expert” football mate? I’ll take the odds thank you very much.

This is exactly how casino’s work. Their gaming systems are all designed to ensure the odds are firmly in their favour. Sure, you may win, and you may even win big, which is why you keep playing – but the odds are slim. And if you keep playing for long enough, you will eventually lose.

And when it comes to investing, investors seem to throw the odds out the window and prefer to play a very different game. One that is akin to the gambler at the roulette table. One where investment professionals try to outguess prices established by the collective wisdom of millions of different buyers and sellers each and every day.

Investors may be surprised by:

1) The number of investment funds that become obsolete over time, and

2) The low percentage of funds that are able to outperform their benchmark.

The chart below shows the sample number of funds that existed as at 31 December 2017, the number of funds that survived, and the number of funds that outperformed their benchmark. For example, 5 years ending 31 December 2017 (from 31 December 2012), there were 2,867 sample funds, of which 82% survived the 5 years, and only 26% were able to outperform their benchmark.

Source: Dimensional Fund Advisers (DFA)

Both survival and out-performance rates fall as the time horizon expands. For 15 years ending 31 December 2017, only 14% of funds survived and outperformed their benchmark. The odds of this game don’t seem very compelling if you ask me.

Let’s say you’ve found a manager who’s been able to outperform their benchmark for the last 3 years, and you’ve decided to hire them. Most investors and advisers use this method of manager selection, reasoning that a fund manager’s past success is likely to continue into the future – sack the poor performers, and hire the strong performers is how the narrative goes. The evidence suggests the contrary.

The chart below shows that among funds ranked in the top quartile (25%) based on previous three-year returns, most of them did not repeat their top-quartile ranking over the following. Over the periods studied, top-quartile persistence of three-year performers averaged 26%.

Source: DFA

The assumption that strong past performance will continue often proves faulty, leaving many investors disappointed. And despite all the evidence, investors continue to search for the winning investment – taking far greater risks than they ever expected.

Imagine for one second you could invest like the casinos. Putting the odds of success firmly in your favour the longer you play the game. As investors, we need to consider more than just a compelling story, and more than just good past performance. You may choose to ignore the evidence. You may choose to take on the odds. You may choose to ignore probabilities and make decisions based on predictions. Now that Paul the octopus is no longer with us, you may as well ask Achilles the cat for stock tips.