Whenever the market experiences a rapid and dramatic period of growth, the discussion of a bubble usually arises. Investors have experienced them often enough in history to make it a legitimate concern.
Since we benefited from high growth in the markets over the past few years because of the AI boom, the dialogue around a market bubble has become common among stakeholders.
Nvidia has reached a market capitalization of US$5-trillion, and SpaceX is trading near 100 times revenue. Comparisons are being made between this boom and the growth in the tech market during the “dot-com” boom and crash that occurred a quarter of a century ago.
Opinion: Canada’s new AI data centres could be obsolete before they open
Although the two events share some speculative characteristics, there are some differences in the underlying economics of the two.
The internet bubble of the late 1990s was driven by a set of unique factors that fuelled the growth. The dawn of the internet produced a new technology that was accompanied by a great deal of hype from both the industry and the media.
The hysteria at the time caused company valuations to skyrocket to levels out of sync with the revenues they generated. The prices of these companies continued to grow because of a cheap-funding environment, as venture-capital companies were chasing any company with a “dot-com” in its name.
However, the mismatch between valuations and revenue, combined with unproven business models, eventually resulted in the financial correction. The Nasdaq composite would ultimately fall 78 per cent from its peak in March, 2000, to the bottom in October, 2002.
In contrast, the artificial-intelligence boom we are experiencing has been driven by hardware developments such as GPUs and related computing power breakthroughs. The target market for AI companies is enterprise adoption of their technology in areas such as cloud, automation, and other productivity improvements.
Because of the nature of this technology, capital investment is being made by a concentrated group of players, such as chip manufacturers and AI labs. Also, these players have experienced real revenue growth over this period. While the leading companies are profitable, their valuations have continued to grow at a much higher rate than the broader market.
Opinion: China is AI-maxxing, and it has a lot to teach us
Although the nature of the internet and artificial-intelligence booms may differ, both have some similarities. Venture-capital and private-equity companies chased investments in these industries during both growth periods. One of the driving factors was the media and analyst hype that both technologies received during their introduction and early growth period.
During the internet era, valuations of existing companies would literally increase if they added “.com” to their name. Today, every CEO doing an earnings call will mention AI and how their company has positioned themselves to capitalize on growth of the industry. Anecdotal expectations of what the new technologies could provide to the world were used as a basis for growing valuations, and investors were seeking to get on the ground floor of these industries.
From the perspective of investors, both booms also resulted in concentration in a few of the larger names of the time. During the internet boom, the names included Cisco, AOL and Yahoo, while prominent public companies benefiting during the AI boom include Nvidia, Microsoft and Apple.
However, there are some significant differences between the two periods as well. The most meaningful is arguably the fact that the major players in the AI boom are generating billions in revenue and profits, while most dot-com companies were unprofitable and many had no revenue.
Another key distinction that internet-based companies targeted the retail marketplace by creating websites that offered products and services to individual consumers. But a large portion of the AI boom has focused on corporate consumers, driven by the growth of GPU-powered computing and enterprise software.
Opinion: The real danger of AI isn’t what it creates – it’s the human agency we are giving away
Another difference, from an investor’s perspective, is the pool of target companies. The internet boom featured thousands of unprofitable startups, so investors tended to bet on one (or a few) in hopes of picking a long-term winner. This was famously demonstrated when 19 early-stage internet companies ran expensive ads during the Super Bowl in 2000 in hopes of gaining national exposure.
The AI boom is the opposite since there are a small number of profitable companies that investors have chosen to focus on.
The common saying “history doesn’t repeat itself, but it often rhymes” applies today: Market booms and bubbles in the financial markets may have similarities but are never the same.
Can we conclude that the AI growth we are experiencing will result in the same crash that internet companies experienced? Probably not, since the differences are significant.
However, this doesn’t imply that the AI industry can avoid a price correction. Current valuations of AI companies are built on unproven optimism and may not be sustainable. Investors are pricing in instant gains from productivity that historically occurs over decades.
History has shown that stock valuations always return to realistic levels and AI will be no exception.
Anwar Husain is an award-winning finance professor at the University of Toronto and a senior investment adviser and wealth adviser with Richardson Wealth. He can be reached at Anwar.Husain@RichardsonWealth.com.
Editor’s note: The byline has been updated to correct the spelling of Anwar Husain's name.