This time’s different.

Don’t get testy with me. I didn’t say it. Rather, it was erstwhile bear Mike Wilson who alluded to the most dangerous phrase in all of finance in the course of explaining why the AI frenzy isn’t analogous to the dot-com boom.

Wilson probably isn’t wrong to downplay the shrill crescendo of bubble warnings market participants are bombarded with upon perusing the financial pages each day. After all, this time certainly is different in a lot of respects, not least of which is that the hyper-scalers — the locus of concern — are among the greatest companies in the history of capitalism and operate de facto monopolies on the so-called “attention economy.”

If you’re determined on the AI bubble point and enjoy couching your crash narratives in hackneyed old saws, it’d be more apt to caution that “history doesn’t repeat, but it often rhymes.”

For his part, Wilson doesn’t see it. Or hear it. And he won’t speak it, either. “Given the AI capex boom… some market participants are focused on whether we’re toward the ending stages of a valuation bubble, like 1999-2000,” he wrote, 16 pages into his bullish year-ahead 2026 outlook, profiled here on Tuesday. “We see some important differences on this front.”

The figures above are self-explanatory, but just in case, Wilson elaborated.

Today’s big-cap benchmarks are much higher quality than they were in the lead-up to the TMT crash, Wilson said, noting that the free cash flow yield for the median large-cap stock is nearly triple what it was in 2000 (figure on the left, above). The Republican tax bill, he mused, could mean free cash flow “ends up being higher than expected.” (It better be, because at least for the hyper-scalers, spending’s going to be “higher than expected” too and there has to be some left over for buybacks.)

In addition, if you adjust the S&P 500’s forward multiple for margins, as Wilson does in the chart shown on the right, above, the index trades some 35% cheap to the dot-come peak. That at a time when, if you buy Wilson’s macro-policy narrative as expounded in his outlook (and summarized in the above-linked article), the operating environment’s set to become more favorable.

As for the contention that the market leadership trades dangerously rich, that concentration’s uncomfortably extreme and so on, Wilson largely dismissed such concerns.

The most heavily-weighted stocks don’t trade anywhere near TMT bubble levels, he said, pointing to the figure on the left, below, which compares today’s top index weights to the valuations seen during the lead-up to the dot-com collapse.

Basically, the top 10 names are trading somewhere between 1997 and 1998 levels in terms of their forward multiple, and as Wilson emphasized, they’re 13 turns cheap to the 1999 blow-off top.

The figure on the right, above, is straightforward: Margins for the top 10 names are 22ppt better than they were when it all went left 25 years ago.

Mike goes on (the S&P’s 70% cheaper in gold terms today versus 2000, never mind that gold’s almost surely in a bubble too) and on (the ERP was lower in 2000, never mind that 2000 was an anomaly on that metric and ~30bps today counts as the lowest in 25 years) and on (92% of the S&P 1500 is free cash flow positive today versus 72% in 2000).

As always, Wilson’s analysis doesn’t suffer from being slapdash. Haphazard Wilson’s bullish view isn’t. But to this observer anyway, it reads in places like a determined effort to support a conclusion he’d already come to.

And look: As a “long only” investor, I’m certainly pulling for Mike’s thesis. To quote Ellen Ripley, “I hope you’re right. I really do.”