The Numbers Are Genuinely Alarming

The SOX semiconductor index — which tracks chipmakers like Nvidia, Micron, AMD, and Intel — is currently trading at a peak price 62% above its 200-day moving average.

According to CNBC, Bank of America strategist Michael Hartnett dropped that figure in a Thursday note, and the comparisons he reached for are not comforting. The Dow Jones Industrial Average was only 27% above its 200-day moving average heading into Black Monday in 1987. The spread before Black Tuesday in 1929 was similarly modest.

The Nasdaq's spread before the dot-com crash in 2000 hit 55%. AI chips just passed that mark.

The only comparable episode Hartnett could find was the French CAC All Tradable index in 1720, which ran 73% above its moving average before the Mississippi Bubble collapsed — an event where banknotes backed by a failing colonial company were literally used as legal tender, the French money supply doubled, and the whole thing detonated spectacularly.

Hartnett's own words, per CNBC: "Exponential price action, market concentration, collapsing vol, stocks bossing bond yields higher, why melt-up everyone's new base case… Here we go."

A strategist at a major Wall Street bank is watching a train accelerate toward a cliff.

The Bull Case Isn't Nothing

Not everyone is panicking.

Benzinga reported Tuesday that Bank of America analysts are making the opposite argument: that Nvidia and Micron still look cheap relative to their earnings trajectories, even after the historic rally. The SOX index surged on that note, with chip stocks across the board posting gains.

The bull argument is straightforward. These companies are printing money. Nvidia's data center revenue is real. Micron is selling every chip it can manufacture. If the earnings justify the valuations, it isn't a bubble — it's a re-rating.

Bloomberg flagged that the chip stock rally has now turned historic by multiple measures, framing the bubble debate as genuinely unresolved rather than settled.

This is NOT a clear-cut case either way.

What the Mainstream Coverage Is Getting Wrong

CNBC and Bloomberg are both center-left outlets with enormous financial industry advertising relationships. They present "both sides" — the bubble camp and the bull camp — without ever forcing readers to sit with the actual historical data Hartnett assembled.

The Mississippi Bubble comparison carries weight when spelled out. That collapse didn't just hurt investors. It destabilized the French government, wiped out fortunes across Europe, and contributed to the economic conditions that eventually fed into the French Revolution. Mentioning it as a data point in a listicle about chip stocks buries the lead.

On the other side, financial media that leans bullish on tech is cherry-picking the "railroads were a bubble too and they still changed the world" argument without acknowledging that railroad investors lost fortunes even as the technology succeeded. The technology can succeed while traders lose everything.

Economist Ann Pettifor, founder and director of Policy Research in Macroeconomics (PRIME), told CNBC: "Having to amass more than a trillion dollars in cash to support the investment has led to what everybody talks about as a bubble."

The Historical Context That Should Scare You

Multiple Wall Street banks — per CNBC — believe AI investment will exceed $1 trillion next year.

Robin Wigglesworth of the Financial Times tried to put that in perspective on the podcast "Unhedged," citing a JPMorgan analysis. He argued the AI build-out is actually small compared to the 1860s railroad boom, which — adjusted for inflation and scaled to GDP — represented the equivalent of $10 trillion in today's dollars.

Wigglesworth's point: railroads were a bigger capital commitment, relative to the economy, than AI is today.

Railroad investors still got obliterated. The technology won. The shareholders lost. Those two outcomes are NOT mutually exclusive.

What Regular People Need to Know

If you own a 401(k), an index fund, or any tech-heavy investment account, you are already exposed to this. The SOX index, Nvidia's weighting in the S&P 500, and the concentration of AI capex in a handful of megacap stocks mean ordinary American savers are riding this wave whether they chose to or not.

The comparison to 1929 and 1720 does NOT mean a crash is imminent. Markets can stay irrational longer than most people expect. Hartnett himself has been cautious on tech for a while — and being early on a warning is the same as being wrong, until suddenly it isn't.

When the most comparable historical episode is a French colonial company that broke an entire national economy 300 years ago, dismissing the concern because "AI is real" is not a risk management strategy.

The technology may be transformational. The valuation still has to make sense. By Hartnett's data, it's stretched beyond almost every precedent in market history.