My (imperfect) memory causes me to think that discussion then sounded pretty similar. But since then the market has risen ~50%. Does believing in the current bubble mean believing that that entire gain will be given back? If it is, will we not simply be back to where we were two years ago, when the market was already widely judged to be overvalued?

Perhaps you're reading too much into the reasoning. A high valuation level is a great example of being "approximately right", not a market forecast.

The conclusion from high valuation levels a couple of years ago was that medium-to-long term forward returns would probably be a lot lower than average, with reasonably high confidence. That still seems right, and nothing has shown it to be wrong.
The conclusion from high valuation levels now is also that medium-to-long term forward returns would probably be a lot lower than average, just with higher confidence and consequently somewhat lower central expectation.

If you don't read any more precision than that into the observations, there is no contradiction. The market can be exuberant for a long time. Forward returns will be quite poor starting now and ending on any date in the future that corresponds to the average valuation level in the next 20 years.

Valuation alone is a truly terrible predictor of market direction in the short to medium term.* But for the long run it's as reliable as the laws of thermodynamics, just with less precision.

I have a very good idea of how high current valuations are among various categories of US stocks compared to history and the factors that seem to affect valuations, and literally hundreds of market prediction models, but I have absolutely no clue what the next year will be like. The short term will be somewhat more likely to be good than bad, and the medium-to-long term will definitely be poorer than average, but that's about it.

Jim


* The exception being moments of clearly cheap prices. Statistically, that tends to turn out well quite reliably and quite promptly.