The AI Investment Opportunity | Weekly Market Update

One question I am asked occasionally is what is the best way to invest to make the most of the current Artificial Intelligence (AI) revolution. As we don’t do single stock equities, I don’t have a strong view at the individual company level. But we have ensured we have a decent (passive) allocation to the US market and we have added to this with a growth-focussed global manager. As the returns to AI have (so far) accrued to large cap US technology companies, I feel that this approach has meant we have owned enough of the beneficiaries of the AI revolution to keep up.

The large cap winners here have been semiconductor manufacturers (like NVIDIA) that effectively make the shovels for the AI gold diggers. And the AI gold diggers themselves which are most obviously the US technology giants: Microsoft, Google, Meta, Amazon and Apple. Are these stocks still the right way to play AI today? Goldman Sachs was out with an interesting note this week that attempted to answer this question using a historical perspective. Their research focussed on investment returns during previous technological revolutions. These were books (16^th century), canals (18^th century), railways and the telegraph (19^th century), the telephone, the radio and the PC (20^th century) and then finally the internet in the 2000s.

The conclusion of their research is that if AI behaves like previous technological revolutions then the current crop of AI winners might not continue to be such great buy-and-hold investments going forward. Take the railways for example. There was an initial phase of exuberance as people came to realise the transformative effect railways would have. This meant there was plenty of capital for new railway projects. In the UK alone, there were 1,240 projects seeking capital in 1845 and the number of railway miles went from 100 in 1830 to 6,123 in 1845. Meanwhile railway stocks were fantastic at first but eventually the optimism bubble burst and by 1850 most stocks were down an average of -85% down from their peak. This pattern: initial exuberance leading to over-investment leading to poor longer term returns is a theme of the historical record.

But who were the winners of these revolutions? Generally, it was the consumers and companies that benefitted from these new innovations (be it miles of new railway lines, telephones or the internet). The initial builders of them generally did not fare so well. Goldman note that the initial pioneers consistently under-estimated the impact of competition on their returns and over-estimated the money they thought they would make in the future. The parallel today (if there is indeed one) is that we will all benefit from the investments being made today by AI giants. But the AI giants themselves will see fewer of these benefits than they think.

Instead, it will be existing companies who can leverage the new technology, as well as new entrants (who will likely start life today as venture capital backed private companies) who will be the winners. If this is the case then we should expect the turnover we have seen in the largest US companies to continue for a while yet:

So, today, my focus would be on the companies that will ultimately benefit from the AI investments being made (but, crucially, who are not paying for them). As for new entrants, it still feels quite early to know where all this will end up. I think some patience might be rewarded.

What is the case that today’s large cap technology AI winners will in fact continue to be fantastic investments? My first note of caution is on taking all lessons from the past at face value. There is an investment cliché that the four most dangerous words in finance are “this time is different”. The older I get the more I think a better way to think is “every time is different”. Covid and its impact on investment markets and inflation (as well as the lives of us all) has sparked an economic cycle (and central bank policy response) that we haven’t seen before. Most historical parallels (e.g. to the 1970s) have so far proved to be wrong (and money-losing). More generally, I think history provides a good guide to the behavioural side of markets (and patterns of boom and bust). But AI (especially if it continues to develop at the pace it has been recently) may prove to be so fundamentally different from previous technological advance that historical analogies prove to be more or less meaningless.

So, it may be that the next generation manage this technological revolution differently to previous generations. For reasons I am still not sure, the Goldman note also highlighted the most popular music artists over time. Aside from being pleased to make this list in 2010, I feel have a pretty good handle on who these people are up to 2015. But 2023? Bad Bunny? Feid? Karol G? If this is the generation that will be living through the AI generation you might want to ask someone younger than me what they think as well.

 

 

Chris Brown, CIO

cbrown@ipscap.com

The value of investments may fall as well as rise and you may not get back all capital invested. Past Performance is not a guide to future performance and should not be relied upon. Nothing in this market commentary should be read as or constitutes investment advice.