DeepSeek: The Pin “No One” Saw Coming

“The internet bubble circa 2000 is the most extreme in modern capitalism.” [That is, until AI took its place].

— Charlie Munger

January 28, 2025 — We began our financial writing career in 1997, investigating the root cause of the dotcom bubble… and the subsequent bust that began in 2000.

If you’ve been following along for a decade or more, from the very founding of the email letter The Daily Reckoning in the late 1990s, you know we’ve written bestselling books on the subject. We try to share our insights daily; sometimes, they seem larger than life and hard to grapple with.

Booms thrive on the spark of technological innovation and the fuel of financial speculation, igniting visions of endless progress. In Financial Reckoning Day, we warned that cheap credit and wild bets inflate these dreams, convincing everyone “this time is different.”

Later, in Empire of Debt, we continued the work. We observed that the very innovations that inspire confidence — railroads, dot-coms, AI — become the fault lines when speculation cracks. Every boom and bust shows how greed, cloaked as progress, sets the stage for disaster.

The winds of political change get whipped up over the roiling tumult of the economy. Individual investors are subject to paradigm shifts in the economy and in politics. Those who aren’t aware or prepared become the unintended victims of change… the collateral damage.

It’s our mission at Grey Swan to identify the tectonic shifts that distinguish fortune from despair. The episode that began with DeepSeek a few short days ago seems like a moment we ought to pay attention to.

The dotcom era had its heroes — Amazon, eBay, and even Pets.com (briefly). The AI boom of today has Nvidia, Microsoft, and a host of startups promising to reshape the world.

But let me take you back to a spring morning in March 2000. The Nasdaq hit 5,132.52, and investors popped corks, certain the party would never end.

By April 17, the Nasdaq was down to 3,227. Over the next 30 months, it crashed 78% from its peak, bottoming out near 1,100. The pin? A sudden realization: growth projections were built on hype, not profits.

Fast forward to today, and you can’t help but feel the eerie parallels. AI mania has pushed tech stocks into the stratosphere. Nvidia, the darling of the moment, has seen valuations soar on the promise of chips that drive machine learning. And Nvidia’s valuation is now far in excess of dotcom-era darling Cisco:

Turn Your Images On

We’ve been tracking this move since the inception of Grey Swan, mindful of what happens when this bubble bursts.

But here’s the question keeping me up at night: Could DeepSeek, the AI-powered trading app, be the pin this time?

Back in 2000, investors weren’t just burned — they were blindsided.

Many had believed the hype until the very end. As someone who lived through it, I saw how quickly optimism turned to panic. The dotcom crash wiped out $5 trillion in market value, and the psychological toll was even greater. Investors didn’t just lose money; they lost faith.

DeepSeek’s arrival smacks of the fuzzy promises of those days.

The app is billed as a revolutionary tool capable of analyzing markets and predicting moves in real-time. In theory, it could give retail investors an edge. But history teaches us that the more powerful the disruptor, the more fragile the market it disrupts.

We’ve been writing about Nvidia’s vulnerability for months, even as it gathered accolades for driving the tech sector to new record closes… and “hoovering” up capital from investment funds around the global financial system.

Now comes the pin “no one saw coming” – another phrase popular when Cisco crashed early in the dotcom bust or Lehman Bros. followed suit less than a decade later.

“No one,” experience tells us includes the financial media pundits whose job it is to create soundbites so they can deliver eyeballs to advertisers. “No one” does not include anyone who is actually paying attention to how markets, manias and mayhem actually happen.

Imagine this scenario: DeepSeek users pile into AI stocks en masse, driving valuations even higher. Then, in a moment of market turbulence, the algorithms pivot, triggering automated sell-offs. Retail investors, caught off guard, panic and dump shares. The result? A domino effect reminiscent of 2000.

The fallout would be catastrophic. Pension funds, sovereign wealth funds, IRAs, and 401(k)s — many of which are heavily invested in tech, even if in passive vehicles like market index funds — could take devastating hits.

Consider pension funds, already teetering under unfunded liabilities. A collapse in AI valuations could send them into insolvency. Retirees relying on annuities might find their payouts slashed.

Sovereign wealth funds, particularly those in the Middle East, are betting heavily on AI as part of their diversification strategies. If the AI bubble bursts, these nations could face financial upheaval just as oil revenues continue to dwindle.

And let’s not forget individual investors. Many have poured their savings into tech ETFs, riding the AI wave with little thought to diversification. If the bubble bursts, it could take years — or decades — for them to recover.

The sideways movement in the markets right now feels like a warning sign. Markets are trying to shrug off the DeepSeek news as quickly as it moved in. But the AI rollout is here to stay – and now it’s clear there’s global competition. It won’t just be the big-cap U.S.-based tech giants that benefit.

DeepSeek, while innovative, adds a psychological twist: it amplifies market sentiment. If it works as advertised, it could drive prices to unsustainable highs. But if it falters, the ensuing panic could set off a chain reaction no algorithm can stop.

During the dotcom bust, the lesson was clear: bubbles don’t just destroy wealth; they redistribute it. Those who saw the crash coming and positioned themselves wisely emerged stronger. The dollar – the currency you calculate your own wealth in – is equally in jeopardy.

So what do you do now? Today, the same principle applies as it did in the early part of the millennium: Diversify your portfolio. Add hard assets like gold or real estate. Consider crypto. Focus on defensive sectors like dividend-paying consumer staples and energy utilities.

Most importantly, don’t chase the crowd. DeepSeek may be a marvel, but it could also be the pin that pops this AI bubble. And if it isn’t, today’s valuations – and historic concentration in a few companies at the top – suggest another pin is on its way.

In times like these, staying grounded is the best way to protect — and grow — your wealth.

Regards,

Addison Wiggin,
Grey Swan

P.S. If you’ve ever been to Times Square on New Year’s Eve, you know it’s all lights and confetti until the clock strikes midnight. The,n the cleanup crew arrives. That’s where we are in the market. The AI and tech bubbles have been the party, but the hangover is inevitable.

Investors love a good story, AI is (still) the hottest one in town. But as history teaches us, no story lasts forever. From the dot-com bust to the 2008 housing crisis, bubbles don’t pop quietly.

They burst.

The detritus doesn’t just land on Wall Street. The slime gets slathered all over Main Street, too.

As always, your cheerful reader feedback is welcome: addison@greyswanfraternity.com. (We read all emails. Thanks in advance for your contribution.)

“The internet bubble circa 2000 is the most extreme in modern capitalism.” [That is, until AI took its place].

— Charlie Munger

January 28, 2025 — We began our financial writing career in 1997, investigating the root cause of the dotcom bubble… and the subsequent bust that began in 2000.

The dotcom era had its heroes — Amazon, eBay, and even Pets.com (briefly). The AI boom of today has Nvidia, Microsoft, and a host of startups promising to reshape the world.

But let me take you back to a spring morning in March 2000. The Nasdaq hit 5,132.52, and investors popped corks, certain the party would never end.

Turn Your Images On

We’ve been tracking this move since the inception of Grey Swan, mindful of what happens when this bubble bursts.

But here’s the question keeping me up at night: Could DeepSeek, the AI-powered trading app, be the pin this time?

Back in 2000, investors weren’t just burned — they were blindsided.

DeepSeek’s arrival smacks of the fuzzy promises of those days.

Now comes the pin “no one saw coming” – another phrase popular when Cisco crashed early in the dotcom bust or Lehman Bros. followed suit less than a decade later.

Consider pension funds, already teetering under unfunded liabilities. A collapse in AI valuations could send them into insolvency. Retirees relying on annuities might find their payouts slashed.

In times like these, staying grounded is the best way to protect — and grow — your wealth.

Regards,

Addison Wiggin,
Grey Swan

Investors love a good story, AI is (still) the hottest one in town. But as history teaches us, no story lasts forever. From the dot-com bust to the 2008 housing crisis, bubbles don’t pop quietly.

They burst.

The detritus doesn’t just land on Wall Street. The slime gets slathered all over Main Street, too.

As always, your cheerful reader feedback is welcome: addison@greyswanfraternity.com. (We read all emails. Thanks in advance for your contribution.)

The narrative for BTC and other cryptocurrencies is that every government around the world has high debt-to-GDP ratios. It means they are going to print more currency. It means there is a need for alternative currency. In the past, this alternative currency was gold.

Gold is not very portable. It’s a good store of value. It’s not as great of a store of value as BTC in terms of actually storing it. BTC, you can store it on a hard drive or at Coinbase. Gold, if you have bars you have to keep them in a bank or you have to dig a hole in your backyard. And you can’t send gold around the world as easily as you can send BTC.

I still think this rally has legs. If you go back to where the breakout happened, we were really in November of 2024 that was the beginning of this bull market in my mind because that was the first time we hit an all-time high in a couple years. Then we rallied. We pulled back. We tested that level again.

The uptrend, in my mind and with what I’m seeing, is still intact. We’re just in an oversold condition right now.

Nvidia’s $900 billion round-trip this week wasn’t about some revelation in Jensen Huang’s chip factory. The business is firing on all cylinders – and may yet be one more reason for the market to soar higher into 2026.

The culprit was the macro — one gust of wind from the labor market and trillions in valuation shifted like sand dunes.

Nvidia’s earnings lifted the market at the open, but the jobs report’s undertow snapped sentiment like a dry twig. As we pointed out this morning, the S&P notched its biggest intraday reversal since April.

The first half of the move was classic Wall Street choreography: blowout earnings, analysts breathless with adjectives, and every fund manager terrified of underweighting the patron saint of AI.

In April, following the “Liberation Day” low, the indexes took off in the morning only to crash later in the day. The first and only other time in history we have seen a strong bullish opening followed by a sharp bearish close was during the 2020 recovery from the Covid shock.

In both cases, the markets were rebounding from exogenous shocks.

That’s not where we are today. The index-level charts may look composed, but underneath plenty of individual stocks are trading as if they’ve already slipped into a private bear market of their own.

We’ll see how the day unfolds. It’s options-expiration Friday — the monthly opex ritual when traders roll positions forward, unwind old bets, and generally yank prices around like terriers with a chew toy.

Every major tech shift has followed a similar pattern. As information moves faster, the money follows.

The telegraph made news global and opened up a world of investment opportunities. Radio, and then television, ignited a new wave of prosperity for investors. And the internet made communication instant, creating fortunes for those who saw what was coming.

Now standards like x402 are doing the same for AI and digital payments, potentially putting Jamie Dimon’s empire in jeopardy.

If you have Coinbase building the payment rails, Circle handling settlement and projects like Worldcoin and Particle Network solving for identity and wallets — do you really need a bank to validate transactions and keep track of who owns what?

All of these companies are helping to build a new layer of fintech infrastructure. And they’re all working toward an economy that runs continuously, without the need for corporate scaffolding.