The debate over whether AI has inflated a stock market bubble may be missing a simpler question: what are investors measuring prices against?
Most market charts are quoted in US dollars. But the dollar itself has changed materially since 2020. The US M2 money supply has risen from about $15.4-trillion in early 2020 to about $23-trillion in 2026. Federal debt has pushed higher, annual interest costs now exceed $1-trillion and the government is still running large deficits even with employment relatively strong.
That matters because a rising stock market can reflect two things at once: companies may be more valuable but the money used to price them may also be worth less. This is the denominator problem.
The AI boom is often compared with the dot-com bubble. The comparison is useful, but only up to a point. The Nasdaq 100 is expensive, yet it is not priced like 1999. Its trailing price to earnings ratio is about 33, compared with more than 100 at the end of the dot-com era. Today’s largest technology companies are also profitable and funding much of their AI spending from cash flow. In 1999, many market leaders had weaker earnings, thinner balance sheets and far more speculative business models.
That does not mean there is no excess. Some AI-linked stocks trade at valuations that require years of near-perfect execution. The S&P 500 is also expensive by long-term historical measures. But the bubble label becomes less obvious when prices are adjusted for monetary dilution.
Gold helps show the problem. Measured in dollars, US equities look as if they have surged since 2020. Measured in gold, the gains look far less dramatic. That is not because gold is a perfect benchmark. It is because gold is a monetary asset that central banks cannot print, making it a useful check on whether nominal gains are really gains in purchasing power.
Crypto presents the more surprising picture.
Bitcoin traded near $59,000 in early July 2026, down about 53% from its October 2025 high of $126,272. Total crypto market value was about $2.2-trillion, still below its 2021 peak, even though the money supply has expanded. Put differently, crypto’s share of the global liquidity pool has shrunk.
That is unusual for an asset class built around scarcity and monetary debasement.
There are two ways to read it. The bearish interpretation is that investors have permanently lowered crypto’s expected role in the financial system. The bullish interpretation is that crypto is one of the few major asset classes still priced as though the denominator has not changed. Both views deserve attention.
The strongest case for bitcoin is not that it must rise because dollars are being printed. Markets are rarely that clean. Bitcoin can fall during liquidity shocks, regulatory pressure or forced deleveraging. The broader crypto market also faces its own dilution problem, with millions of tokens competing for capital and attention.
But the monetary backdrop is hard to ignore. If investors believe the real rate of dollar dilution is closer to a high single-digit annual rate than the official inflation rate suggests, the hurdle for every asset changes. Cash becomes less neutral. Bonds become more vulnerable. Equities need to be measured more carefully. Bitcoin and crypto need to be judged against their share of global money, not just their dollar price.
That is why the AI bubble debate may be too narrow. The real story is not only whether technology stocks have gone too far but whether investors are still using a stable ruler to measure them.
Before calling something a bubble, first measure the money.
• Muchena is founder of Proudly Associated and author of ‘Tokenized Trillions’ and ‘Artificial Intelligence Applied’.
Business Day
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.