Beware the illusion of large financial data sets – with Jake Taylor

Why do so many investors feel so confident they can predict the future? As we continually point out, here on The Value Perspective, in pieces such as Our 2021 anti-forecast – Vic Reeves on a roll, by its very nature, the future is uncertain and thus impossible to predict on a consistent basis – and yet people cannot resist trying. One possible explanation, according to our most recent podcast guest, is “the illusion of large data sets”.

The versatile Jake Taylor is not only chief executive officer of value manager Farnam Street Investments but also the host of podcasts, such as ‘Five Good Questions’ and ‘Value: After Hours’ – not to mention the author of the unlikely sounding novel The Rebel Allocator, whose plot he likens to “The Karate Kid but where Mr Miyagi is Warren Buffett”.

One aspect of Taylor’s work that really strikes a chord is the ‘Veggies’ feature of ‘Value: After Hours’ – largely because it is something we also seek to do, here on The Value Perspective. So called because ‘you save the veggies for last’, these segments use ideas from biology, history, psychology and so on to offer lessons on finance and investment.

Deepest pools

When trying to drill to the truth on any subject, Taylor argues, we need to seek out “the deepest pools of data”. “When I do my veggie segments, I have three primary buckets where I am looking for an ‘n’,” he explains. “Number one is the inorganic universe, which is about 13.7 billion years old. That is a lot of interactions of matter and energy running into each other – gravity and so on – so that is a very large ‘n’.

“The next bucket is biology – which is roughly 3.7 billion years of life on earth so is also a lot of interactions of things eating other things and evolution playing out – and then there is human history. Sumerian cuneiform was roughly 5,000 years ago so we have written history dating from then but we probably have about 200,000 years’ worth of human history while humans in some form arguably go back about two million years.”

In contrast to these very large data sets, argues Taylor, finance really only offers the illusion of large data sets. “It is not always as solid as it looks,” he continues. “With something like market prices, say, we could take every single squiggle per second and have millions and millions of data points on pricing but what is the true data we are looking for? What are the really big moves?

“According to Fidelity, there have only been 16 bear markets since 1926 so, when it comes to investing, our real ‘n’ is often much smaller than we realise. Because we are surrounded by so many numbers, we can feel the data is so rich that we have more ‘standing room’ than we do. Yet, compared with the inorganic universe, human history and biology, we are not even close. We should probably be a lot more humble than we are.”

If anyone failed to find 2020 even a little bit humbling, they were not paying attention but, as the economist John Kenneth Galbraith pointedly observed: “There can be few fields of human endeavour in which history counts for so little as in the world of finance.” As the new year progresses, here on The Value Perspective, we will continue to work hard – as Taylor no doubt will with his ‘veggies’ – to keep our readers’ memory green.