🧵 charts: CAPE and traditional P/E measures
Using a method that accounts for differences in dividend payout ratios, I have the current S&P 500 CAPE (end of May 2026) at 42.3 versus a March 2000 value of 45.6.
Many will immediately dismiss this result as invalid because CAPE is a distorted measure. Yes it is, like all valuation measures, a point that was particularly relevant roughly 10-15 years ago. But the problems with CAPE back then were mostly tied to the fact that there were two huge writedown-heavy recessions in the trailing 10-yr EPS window that were depressing denominator of the measure. Also, there had been a lot of recent earnings growth, which pushes up on the CAPE mathematically (assuming ttm P/E is held constant).
But if you're making a comparison between today and March 2000, you don't have to worry about those distortions. They don't apply. Yes, there's recently been a lot of earnings growth, but that was equally the case back then, so any related distortions drop out. And if you look back 10 years from March 2000, there was only a small, brief recession in the trailing window, with minimal distortive impact--much like today.
The truth is, if you want to claim that this market is approaching or exceeding Tech Bubble excesses, CAPE is one of the WORST measures you could be citing. Simple ttm measures of P/E are all showing the current S&P 500 value closer to the March 2000 value than CAPE! (See the chart below, where a value of 1.00 equals the March 2000 value for each measure. The red line, CAPE, is currently the lowest of them all, showing the biggest discount relative to the 2000 valuation peak, albeit only a small one.)
Additional measures in subsequent tweets.