“This time is different.”

So said Goldman’s David Kostin in his latest. Don’t laugh. Or if you do, just know that Kostin gets the joke. He’s fully aware that those four words, arranged in that order, are dangerous in the market context, and in a lot of other contexts besides.

He was addressing the perception that 2024 is beginning to look a bit like 2021, something I’ve been on about at length recently, including in the latest Weekly.

Part and parcel of the ostensible analogue is the outperformance of richly-valued growth shares. Kostin pointed out that the high growth leg of a Goldman equal-weighted, sector-neutral, long/short growth factor trades at an 89% P/E premium to the low valuation leg of the bank’s value factor.

As the figure shows, that’s 1.3 standard deviations expensive on a four-decade lookback, an extreme only exceeded during the dot-com bubble and — drumroll — 2021.

Before explaining why this isn’t necessarily alarming, Kostin briefly addressed the somewhat odd combination of extreme growth outperformance and higher bond yields. Yields are obviously much lower versus the October cycle highs, but they’ve drifted higher in 2024, in part due to the perception that Fed cuts aren’t imminent.

Of course, the reason for that perception (the reason traders have trimmed rate-cut bets) is the resilience of the US economy, and economic resilience is good news for stocks. Risk premium compression, Kostin wrote, can “soften the impact of a higher risk-free rate.”

So, why isn’t “the prevalence of extreme valuations” a concern? Because in fact, extreme valuations aren’t as prevalent as you might be inclined to think.

The figure above is simple, but it’s pretty useful. I’m not sure I’d call it exculpatory vis-à-vis allegations that the current melt-up is vulnerable to a pullback or otherwise presents investors with a chancy proposition, but it is what it is: Evidence that investors’ inclination to pay up is confined almost entirely to big-cap growth.

The perceptive among you already penned the punchline: Although the chart exonerates the current rally from charges that we’re reliving the 2021 “everything bubble,” the sharp run-up in the market-cap share of extreme valuations compared to the unadjusted share of richly-valued companies mirrors the dot-com boom, an even more ominous analogue than the 2021 comp.

That’s not lost on Kostin. “This dynamic more closely resembles the Tech Bubble than 2021,” he wrote, before quickly suggesting that by “contrast with the late ’90s,” Goldman’s view is that valuations for the Magnificent 7 are “currently supported by their fundamentals.”

So, don’t panic. And always carry a towel.