Wednesday Morning Rant

Despite most of the Horde not yet reaching the age of 29, many nonetheless have memories that go back much further. This is a temporal oddity, but a useful one. As a result, many will remember the period 30 years ago. A lot was going on the interregnum between the fall of the USSR and the rise of globalization and Islamic extremism. Economically and financially, this era was defined by high finance (though not as much was the mid-aughts would be) and high technology. The latter, fueled by the promise of the Internet and manic irrationality brought us what is now usually referred to as the "dot-com bubble."

Companies with no profits, no real plan to become profitable and in some cases even without meaningful revenues went public and raked in credulous investor capital. Other companies saw staggering growth in valuations as already-public companies that became major players in internet infrastructure were handsomely rewarded by the mania. Not all of the latter companies failed, but many never recovered and have never achieved their dot-com heights in the thirty years since. The narrative of invincibility became a narrative of dread as it the mania collapsed. The aftermath of the dot-com bubble took years to shake out, and set up some of the environment that led directly to the subsequent 2008 "Global Financial Crisis" and the following Eurozone debt crises.

There are a couple of companies that help summarize the dot-com bubble. One was an outfit called "Internet Capital Group," which was a business-to-business eCommerce holding company and venture capital firm. It had about $70 million in revenues when it went public in August of 1999 at $12 per share. The price doubled on the first day and gave the company a $3 billion market cap. By the end of the month, it had increased nearly five-fold to $58 per share. By the end of the year, the stock price had gained over 2700% and was the most successful IPO of the year. It reached a peak market valuation of over $50 billion. The stock split in its first six months. By the end of 2001, the company was a penny stock. By the end of 2002, the company had a market cap of around $200 million and it finally liquidated in 2018 after shambling along as a zombie for more than a decade, shedding most of its properties along the way. This dot-com high-flier was a pure bubble stock, and billions of credulous investor dollars evaporated in its cash furnace.

On the other side, there were the real dot-com companies like Cisco Systems. Cisco was - and is - a real company. A real company with real products and real customers and real revenues. As the dominant manufacturer of network switching equipment, the explosive growth of the internet resulted in a lot of sales for Cisco. At the peak of the dot-com excesses in March of 2000, this company that made the tools other bubble companies needed had a stock price of around $80 share and a market cap of over half a trillion dollars - a hell of a run in its decade as a public company, given that it started with a quarter-billion-dollar market cap when it went public in 1990. This year marks the first year that Cisco has managed to regain its prior value of around half a trillion dollars. Nominal. Inflation-adjusted, it's still nowhere close to where it peaked.

Beyond just the financials of the bubble players, there was the wider mania that helped fuel it. The number of estimated users for internet companies in the late 90s totaled to more people than existed on the planet - by a decently wide margin. The sky wasn't the limit, there was simply no limit at all. Anything was possible, and the internet was going to do it all, however absurd it sounded. Everyone believed, and everyone got rich. Right up until the whole thing collapsed like a dying star. The internet wasn't fake, and it wasn't unimportant. Its value, however, was totally disconnected from reality and the promises everyone made had no way to be fulfilled.

It ended the only way it could, with the stark reminder that no player is safe during that kind of mania. Whether a long-shot startup or a major, established and very real company or even the narrative itself, the collapse of a mania takes out everybody. Even the survivors walk away extremely bloody.

Flash forward thirty years, and it's happening again. AI is driven by companies with massive valuations, lots of promises, and an assumption that trees - this time - will grow to the sky. Take Cerebras, an AI chip maker and services player. It went public this year with half a billion dollars in revenues and, if you account for unusual one-time accounting, losses of around $75 million and negative free cash flow. It has potentially valuable future contracts with some big players. It has a market cap of around $52 billion. AI majors Anthropic and OpenAI will probably IPO this year. OpenAI is a loss-making enterprise that, according to the company's own estimates, will rack up in excess of $100 in losses by the end of the decade, with profitability not expected before sometime in 2030s. Anthropic also loses money, but expects to achieve profitability this decade rather than next. Each company will likely end up being valued in the neighborhood of a trillion dollars.

The infrastructure players are no different. Nvidia, which makes the processors preferred by AI companies, has seen its market cap surge from around $325 billion in 2020 to around $5.4 trillion today. Not bad for a few years. Broadcom, which makes a variety of computer chips and software and whose products are found in most computers, mobile devices and networking equipment, is not directly an AI player but is benefiting from the boom. It has seen its market cap move from around $180 billion five years ago to around $2.2 trillion today. Like Cisco, these are real players with real products and real customers and real revenues and real profits and who have seen unprecedented surges in sales and enterprise value on the back of a new technology sector.

And the narrative is helping drive it all. You can't swing a cat without seeing stories about how AI is changing everything, how everyone will be living in the AI world, how AI demand is unlimited and so on. Preposterous claims like those from Microsoft - which said that all white-collar work will be replaced by AI in just 18 months - are commonplace. Don't worry about negative stories, just believe. And buy. Buy and believe.

AI and dot-com are not the same song, but it increasingly sounds like they rhyme.

posted by Joe Mannix at 11:00 AM

| Access Comments