Lessons from the Future

January, 2021

In investing, there is always a metaphorical tug of war between the past and the future. Is history always doomed to repeat itself, or will the future be vastly different from the past? Is looking into the past the best way to predict the future or is such an approach as flawed as looking through the rear-view mirror while driving?

Like most questions involving extremes, the rational answer is usually somewhere in the middle.

In many ways, history repeating itself is inevitable. The chemistry of the human brain remains largely unchanged through the aeons. People as far back as Plato understood that emotions can trump reason. Emotions are hard to control. They are a survival mechanism that doesn’t necessarily have the individual in mind. They work across large number of trials to keep the species alive and over the span of millions of years, more mammals lived than died by letting emotion override reason, so emotion was selected for.

Naturally, we are left open to the same biases and errors of judgement as people in the past. Extremes of fear and greed and the corresponding effect this has on market prices is here to stay.

In many ways, value investing is a bet on the past. It is a bet that things will not deviate too far from historical norms. On average, this strategy is poised to work well because people’s reaction to events often exceeds their actual importance.

The key word here is often. On a very few occasions throughout history, the reverse happens and the actual importance of a new piece of technology is greatly understated. This is the domain of the growth investor. A well-known example of this is the internet. A lesser-known example of this is GPUs.

GPUs (graphics processing units) were originally added to computers and video game consoles to accelerate the rendering of images with millions of pixels. By 2005, GPUs were produced in such large quantities that they became a commodity. Commodities do not have pricing power, so the price of GPUs naturally fell. In 2009, cheap GPUs were for the first time linked to run neural networks (artificial brains of sorts). A cluster of GPUs could accomplish in a single day what it took traditional processors several weeks to complete. Today, these neural networks are being used to train the self-driving cars of the future. At the end of last year, engineers at Google used neural networks to solve a 50-year-old problem in biology that has significant implications for drug discovery.

In the early days of the internet and the world wide web, no one could have predicted the impact it would have on society. Similarly, GPUs were intended for improving the rendering of pixels, not solving autonomous driving or drug discovery. The point of this is that new technological developments often have second order effects that can destroy entire industries while creating new ones.

As stewards of your capital, having one foot planted in the past and the other in the future is the rational approach. Being too enamoured with the future inevitably leads to investing in growth traps – companies that promise growth which never materialises. At the other extreme, being too drawn by the past blinds you from changes in the economy that can both create and destroy wealth.

Related posts