Zapata Corp was originally an oil driller founded in 1953 by President George H. W. Bush. It ended up with a Rochester connection in the mid-1990s, then majority owned by Malcolm Glazer (who also owned the Tampa Bay Buccaneers and Manchester United soccer team). By 1998 its primary business was fish meal, but it decided to try to cash in on the internet boom by changing its name to zap.com and announcing its intention to acquire several web sites.
As a fish meal company, the stock traded in a rough range of $3 to $5 per share. However, when it announced its internet strategy, the shares rose into the teens and ultimately peaked out north of $24 per share. In short order, an investor could have made more than five times his money. However, what goes up often must come down, and mere months later when the internet strategy was abandoned, the stock dropped to around $6. So those jumping on the bandwagon during the stock’s rise, likely saw significant losses not too long after.
Ridiculous as the Zapata example sounds, that same dynamic has been playing out in today’s market. Allbirds was a maker of shoes favored by the hedge fund crowd. However, sales slowed and the company decided to pivot by getting into the artificial intelligence infrastructure business. Its stock had been declining for some time and was trading around $2.50 this past April. It then announced its strategic pivot and the shares jumped to over $24. Alas, that too was short lived as the stock has fallen to $3.70 as of this writing.
While these may be extreme examples, it wasn’t uncommon for stocks to rise 30% or more in a single day in the 1990s when companies announced they were putting up a website. Similarly, Cracker Barrel stock recently went up over 30% in a single session when it was announced it was implementing AI in its operations. I’m not sure that will make the biscuits and gravy taste any better, but apparently investors expect it to boost long-term profitability.
These types of anecdotes show some of the crazy similarities between the 1990s internet stock mania and what may ultimately prove to be AI mania in our current age. Of course, there are some key differences as well. Many commentators point out that today’s tech companies are real businesses that are massively profitable, unlike the 1990s when companies like Pets.com could go public with only a sock puppet and a Super Bowl ad.
That is undoubtedly true. I worked at a large Wall Street investment bank during the 1990s and saw firsthand companies going public with no earnings, no prospects of earnings, and in some cases no sales. They made the SpaceX IPO look like a stock for widows and orphans.
However, there were some good companies during the 1990s too. Cisco Systems was a darling of the internet age. It produced networking equipment that essentially provided the backbone of the internet. For a while it was the largest public company in the world. It was well run, massively profitable, and there seemed to be endless demand for its products; sort of like NVIDIA today. However, the stock’s high in early 2000 wasn’t achieved again until October of 2025. That means you had to suffer through 25 years of losses if you bought the stock at the peak when its future looked brightest.
With the stock market today trading almost as high as it did in 2000, will today’s darlings be tomorrow’s losers? Even those with strong balance sheets, robust earnings, and solid management? Can Tesla justify a price/earnings ratio of 379? Maybe if the robot army and true self-driving cars come to fruition, but there is a lot of uncertainty before those become reality. Other of the “Magnificent 7” stocks trade at more reasonable valuation levels, but there is still room for 30% or greater declines if the economics of AI fail to live up to expectations.
Could that possibly happen? It has in the past. Another Rochester company, Global Crossing, laid undersea fiber optic cables to accommodate the meteoric rise in demand for internet bandwidth. It turned out that the demand was real, but the profitability wasn’t. A glut of fiber ultimately depressed pricing and while internet traffic outstripped all estimates, it became far harder to make money on it. Global Crossing did its IPO in 1998 at $19 per share. It rose as high as $64 per share in early 1999 and declared bankruptcy in 2000 after the stock had fallen 99.5% from its peak.
Data centers today may be the fiber optic cable of the 1990s. Insatiable demand is leading to a rapid buildout of data centers that appear wildly profitable on paper, drawing in companies like Allbirds (now called Smartbird). I have no doubt that AI will be part of our future. However, that doesn’t mean the stock market boom for all things AI will continue. In fact, I’m quite sure it won’t. Eventually there will be a slowdown in growth either because we’ve hit “peak AI,” or because it will be clear that projected profitability won’t materialize. At that point, today’s high stock market valuations will become difficult to justify and great companies along with terrible companies will see their stocks suffer, just like Cisco’s 25 years ago.
None of this is meant to be a near-term stock market forecast. There is plenty of room for doom and gloom, but the stock market could continue to make new highs for a while longer. However, nothing goes on forever in the financial world, and even a new era of AI usefulness won’t spare those concentrated in today’s tech stocks from suffering the same fate as those of the internet era.
Disclaimer: Armbruster Capital Management’s views as portrayed in this post are subject to change based on market conditions and other factors. These views should not be construed as a recommendation for any specific security or sector. Investing involves risks, and the value of your investment will fluctuate over time, and you may gain or lose money. Past performance is no guarantee of future results.