If a reversion to the mean for cheap stocks seems like it’s taking forever, it is. In 2020, when things were heating up in the stock prices of “COVID-19 beneficiaries,” Value investors were left watching in futility as companies with thin or non-existent profits melted higher. Whether it was app-based food delivery, messenger RNA technology or exercise bikes, if the company benefitted from a ton of people suddenly staying in their houses, times were good.

Using the data library compiled by Dartmouth’s Ken French, we found that the quintile of stocks with the highest price-earnings ratio beat the quintile with the lowest P/E by more in the 10 years to 2020 than it did in the 10 years to the peak of the 2000 Tech bubble.

Clocking in at an annualized outperformance of 6.2% per year, the high P/E quintile’s outperformance in the decade through “early COVID-19” wasn’t even close to any other era; the aforementioned 1990–2000 timeframe only witnessed a 2.4% annualized performance gap. In fact, before dot-com, in no decade-long window in French’s library back to 1951 did the high P/E cohort ever even beat the low P/E group.

After 2020, things got temporarily better for low P/E value stocks, as it appeared the reversion to the mean was finally kicking in during 2022’s bear market. Though many speculative stocks were sent reeling by 80%, even 90% last year, the vibe in Value was markedly better; some of our mandates managed to eke out black ink amid the carnage. If you didn’t blink too long, the 10-year performance differential swung back in favor of low P/E for a short spell. Visions of a resurgent Value-oriented stock market were dashed when 2023’s artificial intelligence frenzy kicked in.

This year’s high P/E rally has been so impressive that the gap from mid-2013 to mid-2023 hit 2.7% per year, higher than the highest differential in any window before COVID-19—including the 1990–2000 experience.

The so-called “Magnificent 7,” the handful of mega-cap giants that have come to dominate the S&P 500 amid the artificial intelligence zoom-a-thon, have relentlessly outperformed in 2023. How long that excitement will persist is anyone’s guess.

Some posit that things have changed in the market, that quasi-monopoly status has been granted to the Tech giants. If that is taking it too far, at least we can say that many of them are money machines.

The stock market had money machines in 1999 too. According to Visual Capitalist, the 10 largest global companies by market capitalization at the time were Microsoft, GE, Cisco, Exxon, Wal-Mart, Intel, NTT, Lucent, Nokia and BP. Like today, many were money machines. A bunch of those companies were at the top of their game, accorded quasi-monopoly status, reputationally tip top. Today, only Microsoft makes the top 10 list.

The market-cap giants today are Apple, Microsoft, Alphabet, Amazon, NVIDIA, Tesla, Meta, UnitedHealth, JP Morgan Chase and Berkshire Hathaway. Most have big P/E ratios—with a very large “E” part of the ratio. Five, ten, twenty-five years from now, how many will have witnessed a fall from grace? Maybe we should ask the shareholders of Nokia or Intel their opinion on the matter.