Haunting of the Dot-Com Era: Comparing Yesterday’s Tech Bubble with Today’s AI Craze

Self-driving cars whisk us to our destinations, AI assistants manage our schedule and streamline tasks at work, and these intelligent systems optimize everything from renewable energy production to healthcare. As we have witnessed artificial intelligence’s (AI) exponential rise in power and capability in recent years, it becomes increasingly conceivable that in due time it will become ingrained in almost every part of daily life. This is the world many investors are betting on, with markets looking to capitalize on every angle of the AI boom. Especially coming after the publicization of large language models/chat bots and resulting observable differences in quality even month over month, AI is no doubt the current technological craze. Companies that develop AI software, semiconductor chips, and energy infrastructure have seen exorbitant increases in the last two years. Most prominently, Nvidia, a semiconductor chip manufacturer and data centre operator, has seen a 210% increase year over year. Other companies like Advanced Micro Devices, Microsoft, and IBM are similar stocks rising on the back of artificial intelligence.

Look Back to the Dot Com Era

To best analyze the validity of the AI boom, one must revisit the past. The dot-com bubble of the early-2000’s was defined by similar exuberance as various startups entered the public equity markets with innovative Internet-based business operations. Investors clamored to try and ride the wave of the incoming technological revolution resulting in the price of Internet companies to soar in the late nineties. However, when the bubble burst in early 2000 and the NASDAQ lost 34 percent of its value just a month after peaking, it left a trail of dot-com bankruptcies and economic fallout, visualized by the US Share Price Index chart below. Companies with weak fundamentals and high leverage prioritized scaling an unsustainable Internet business plan, rather than creating a feasible path to profitability. While the current winners in AI do not have the same level of profitability issues, there are even a fair share of industry leaders who are not overly concerned about making money. Most notably, OpenAI CEO Sam Altman once said “Whether we burn $500 million a year or $5 billion—or $50 billion a year—I don’t care, I genuinely don’t,” he continued, “As long as we can figure out a way to pay the bills, we’re making AGI. It’s going to be expensive.”

Figure 1: S&P 500 and NASDAQ during the Dot-Com Era

Despite industry experts' spot-on predictions of the internet’s ability to revolutionize our lives, the stock prices of internet companies themselves were not as smooth of a ride, providing a cautionary tale for investors in AI companies. Investors could not get enough of internet companies who conducted “e-commerce” employing “straight to consumer” all operating on the “world wide web.”. The AI boom has its own set of buzzwords, with companies across industries looking to implement “machine learning,” “neural networks,” and look ahead to “artificial general intelligence.” AI also seems to have similar, if not much greater, potential to innovate the way we live, giving many people dot-com-like fear of missing out on profiting off companies building AI infrastructure. Given these apparent situational similarities, it begs the question: Is AI the world’s next technological revolution, or will it sputter out like dot-com? Or could both be true?

Is AI a Bubble?

Although the AI industry’s abnormally high returns, the nature of the technological revolution, and snowballing investor sentiment offer parallels to the Dot-Com Crisis, ultimately, the two are fundamentally different. While investors should be wary about continued price growth causing stretched valuations, at this moment, markets appear to have learned their lesson from the 2000’s crash. It’s not unreasonable to expect slight drawbacks in these stocks in the short term, but the thesis that the AI sector is artificially inflated enough to cause a bubble pop is largely unfounded. As opposed to the dot com crisis, the biggest stock risers in the AI boom have earnings to back up their rapid price growth. Using Nvidia as an example again, their one-year forward price to earnings (P/E) is 47x, a high ratio compared to the 35.9x of an index with the 30 biggest semiconductor companies (NASDAQ: SOX), but nowhere near the absurdity of where internet companies were trading at the turn of the century. At the peak of the crisis, the NASDAQ 100, traded at a trailing P/E of 200x, and Cisco, the darling of dot com like Nvidia to AI, traded at 150x. Although AI sector leaders like Nvidia, Microsoft (33.78x), Apple (29.15x), and AMD (46.73x) trade at relatively high multiples, they have demonstrated strong profitability from AI projects, and expected growth opportunities in the sector. Microsoft experienced 17% revenue growth year over year, largely due to a 31% increase in their Azure platform, with 7 percentage points coming from the AI development aspect of the platform. Amazon also reported a multibillion-dollar revenue run rate on the AI front. These examples among others show much stronger company and macro fundamentals behind the AI boom than the dot com crisis.

Further, not only do these major players stand on stronger profitability, but they have largely established their businesses across hardware, software, and infrastructure before foraying into AI. This contrasts with the vast number of companies who have entered the public market with their business solely built off of the innovation of the internet. Pets.com and eToys.com are prime examples of this, prioritizing marketing spend to gain users and website views and employing unique but unsustainable business models. While the companies at the forefront of the AI industry are undoubtedly pouring capital into the AI bucket, they are not solely relying on their technology for their business, as the dot-com companies did.

These are large companies like Microsoft, Apple, Oracle, and IBM, which not only have profitability beyond AI but are integrating and adding AI to one of their pre-established businesses or products. Although they are investing in AI through a period of high rates and borrowing costs, they are still able to produce profit, with all Magnificent Seven stocks, except Tesla, producing a double-digit percentage increase in earnings year over year in Q1 of 2024 (Figure 2). There is also a notion in the market after the dot com crisis that any major technological revolution creates inflated stocks, but according to Goldman Sachs research, that’s not always the case. Goldman Sachs looked at 51 major tech innovations across several years, from 1825 to 2000, and found investment bubbles were evident in 73% of the cases. While there may be a few stocks stretched slightly past standard valuations, the industry is not yet in this 73%, due to stable and growing profitability, AI’s proven ability to produce value, and diversified business across AI leaders.

Figure 2: Earnings gains of the Magnificent Seven Stocks

Conclusion

To conclude, while the AI boom shares some superficial similarities with the dot-com era, the foundations of the current surge are significantly more robust. Established tech giants with diverse revenue streams and proven profitability anchor the AI sector, contrasting sharply with the speculative nature of dot-com companies. Although valuations are high, they are supported by growth prospects and tangible earnings. Investors should remain vigilant about possible market corrections in the short term, but the AI industry and broader market seem to have learned lessons from dot-com, and solid financial underpinnings suggest it is more than just a fleeting bubble.