How Warren Buffet invests in terms of probabilities

“There’s a lot of different scenarios that can play out and under some scenarios, we’ll spend a lot of money and other scenarios, we won’t.” So said Warren Buffett at almost exactly the halfway mark of the 2020 annual meeting of his Berkshire Hathaway investment vehicle at the start of this month – and in one sentence drove to the heart of why thinking probabilistically – that is, in terms of probabilities – is so important for investors.

How investors deal with unpredictable events has been on our minds a lot, here on The Value Perspective – and not just because the Covid-19 virus has made all our lives a great deal less sure. As regular visitors will know, we have just completed a series of podcasts on the theme of decision-making at times of uncertainty – talking to, among others, explorers, members of the military, a space-tech entrepreneur and a film producer.

That probably explains why, when we read the transcript of this year’s Berkshire Hathaway annual meeting, we were especially on the look-out for indications of how much Buffett weights his investment decisions probabilistically. It seems, well, probable that others will have picked up on this before but we were genuinely surprised at the degree to which this approach appears to be embedded into his investment thinking

‘Range of possibilities’

Be it in relation to the economy as a whole, what markets are pricing in or the future of individual companies, it turns out Buffett is constantly weighting scenarios in his head and deciding where the odds stand. Over the course of the four and a half hours that, due to the coronavirus, the ‘Sage of Omaha’ spent in the unusual position of addressing an almost empty room, he specifically referred to probabilities a dozen times.

We will not list them all but there are four instances in the first 15 minutes alone – all referencing the pandemic and its investment implications. At one point, Buffett says: “I don’t know anything you don’t know about health matters, but I do think the range of possibilities has narrowed down somewhat in that respect. We know we’re not getting a best case, and we know we’re not getting a worst case.”

He also notes: “When we started on this journey, which we didn’t ask for, it seemed to me there was an extraordinarily wide variety of possibilities on both the health side and on the economic side. There was DEFCON 5 on one side and DEFCON 1 on the other side, and nobody really knows, of course, all the possibilities that there are, and they don’t know what probability factor to stick on them.”

‘Quite an experiment’

A little later, he adds: “In any event, the range of probabilities on health narrowed down somewhat – I would say the range of probabilities, or possibilities, on the economic side are still extraordinarily wide. We do not know exactly what happens when you voluntarily shut down a substantial portion of your society ... this time we just pulled the train off the tracks and put it on a siding, and I don’t really know of any parallel.”

His final observation in this opening part of his presentation runs: “This is quite an experiment and we may know the answer to most of the questions reasonably soon, but we may not know the answers to some very important questions for many years. So it still has this enormous range of possibilities.” From this opening salvo, then, comes the first of two key points we want to highlight.

Clearly, when Buffett is thinking about the future, he is constantly considering multiple scenarios weighted by probabilities and acknowledging these could translate into multiple possible outcomes. His underlying message is that, given how wide that distribution is in the case of Covid-19, he does not find the odds in his favour – in effect, the future is just too uncertain and can still play out in many different ways.

‘An understandable mistake’

Our other main point relates to Berkshire Hathaway taking positions in four of the largest airlines in the US. For obvious reasons, this has not turned out well but Buffett describes it as “an understandable mistake”, adding: “It was a probability-weighted decision when we bought – that we were getting an attractive amount for our money when investing across the airlines business.”

Unfortunately, as he goes on to explain, “a low-probability event” occurred “and it happened to hurt particularly the travel business – the hotel business, cruise business, the theme park business, but the airline business in particular. And of course the airline business has the problem that, if the business comes back 70% or 80%, the aircraft don’t disappear”.

The thing is, because Buffett adheres to a repeatable system and thinks in terms of probabilities, he knows not to be too hard on himself – as author and retired poker player Annie Duke, another of our podcast guests, would attest, sometimes the cards just do not fall your way. “It’s always a problem if there are things on the lower levels of probabilities that happen sometimes,” says Buffett. “And it happened to the airlines.”

And he concludes: “When I look at worst-case possibilities, I would say that there are things that I think are quite improbable and I hope they don’t happen – but that doesn’t mean they won’t happen.” To quote one final podcast guest of The Value Perspective, market strategist and author Michael Mauboussin: “If your process is well-calibrated, meaning it recognises the event happens the ‘right’ percentage of the time, then you just dust yourself off and do it again tomorrow, right?”