Dotcom bust offers lessons for those in search of the next AI winner

IT SEEMS like every other article recently has been on artificial intelligence (AI). Even non-investment magazines are featuring front-cover articles on AI. It is enough to make every contrarian investor’s hair stand on end.

The potential of AI is being touted as a game changer, primed to revamp our ways of doing business and drastically increase productivity, revenue and profits. Yet, when it comes to solid advice for investors, the conversation is still relatively muted. Perhaps this is because AI is a greenfield sector. Predicting its trajectory is akin to weather forecasting in its infancy.

My inaugural encounter with AI investment research harks back to a 2018 Morgan Stanley recommendation. Their line-up included tech heavyweights Alphabet (Google), Microsoft, Nvidia, Amazon, Alibaba and Baidu. Quite a prescient list, and while the two China companies have lost 30 per cent to 50 per cent of their value, the rest have thrived, with Nvidia stealing the spotlight with an impressive gain of almost 700 per cent (65 per cent annualised).

If you invested in all six stocks equally, you would have made 169 per cent, a figure that certainly is not insignificant. However, an investment in the reliable Nasdaq exchange-traded fund (ETF) would have yielded a slightly less, yet still substantial, 125 per cent return.

What if you had invested instead into the handful of larger AI ETFs available over the last five years? They all underperformed the Nasdaq index:

• Global Robotics & Artificial Intelligence ETF (BOTZ): +35%
• Global Robotics & Automation ETF (ROBO) +48%
• iShares Robotics and Artificial Intelligence Multisector ETF: +46%
• First Trust Indxx NextG ETF (NXTG): +59%
• Artificial Intelligence & Technology ETF (AIQ) +91%

Except AIQ, they even underperformed the S&P 500 ETF, which gained 78 per cent in the same period. If one was looking to invest specifically in the AI domain, existing ETFs may not be the golden ticket, at least not based on the previous half-decade. Many of the AI beneficiaries happen to be the largest technology companies that are part of the S&P 500 – Apple, Microsoft, Alphabet, Amazon, Tesla and Nvidia.

What lessons can be gleaned from this current AI investment landscape? A look back at the dawn of the Internet era and the subsequent dotcom bust is instructive.

Back then, Cisco stock was touted as the “must own” company to benefit from the Internet. You could “buy and forget” it, letting the company compound its earnings at double-digits forever. Two decades later, we can see that the Internet’s forecasted growth has been met, or even exceeded. I still remember the days of faxing 30-page presentations that came out in fuzzy black and white, and could hardly be read. Now, we can send 30-page PDF presentations in full colour instantly, and for free, to hundreds of people at the click of a button.

And yet, two decades later, Cisco stock still trades 30 per cent below its peak at the start of the millennium. At the peak, the stock had a lofty valuation of 29 times its sales, and a price-earnings ratio of 201. Cisco’s edge was not that significant, in that many other companies started making Internet connectivity equipment which did not require particularly advanced technology.

Cisco’s historical example can be a warning for fans of Nvidia. While Nvidia’s AI chips are much more advanced and difficult to replicate compared to Cisco’s equipment, there is the possibility that future technological advances will reduce the cost of such equipment, and even come up with alternatives.

Two decades ago, Intel’s dominance of the CPU market looked unassailable. And yet Intel has been another casualty of technological advances, with its stock 50 per cent below its peak in the year 2000. Meanwhile, Advanced Micro Devices, Intel’s perennial second-place competitor, has left Intel in the dust, with a gain of 4,300 per cent in eight years. In the same period, Intel gained a meagre 1 per cent.

Current valuations of Nvidia assume that it would maintain its market-leading position. History shows that companies are rarely able to do so.

Another takeaway is that many Internet success stories were still in their infancy during the dotcom boom. A parallel can be drawn with today’s AI investment landscape. Apart from OpenAI, valued at an impressive US$29 billion, most other AI-centric private entities have far more modest valuations in the US$1 billion to US$5 billion bracket.

Investors should bear in mind that while pioneering technology companies can often falter, the next big player in the AI space could be among nascent entities. In the realm of AI investments, it might be prudent to consider opportunities beyond publicly-listed companies.

The writer is head of investments Singapore, AlTi Tiedemann Global