It also seems that about 95% of rolling 7 year periods show a positive return in stocks, so you have to have big brass spheres to bet against that with authority.

The interesting thing is that the logic is unassailable: IF (a very gigantic if) valuation levels are at some chosen definition of "historically normal" on the end date, and given that the growth of any reasonable metric is pretty slow and pretty easy to forecast within a percent or two, THEN it's clear that the market will be at a lower level in seven years. The hole is merely trying to get a handle on whether you've chosen a good definition of "historically normal". You get a very different answer if your baseline era is 2020-2026 versus 1980-2026.

And in this era, unusually, there is also some modest extra doubt due to divergence between revenue metrics and sales metrics...US corporate profits as a share of GDP in the last few years have been at levels absolutely unimaginable a few decades ago, primarily as the macro flip side of giant public sector deficits. Things look bad if you use something like CAPE which includes recent amazing net earnings, but they look very much worse if you use something like aggregate revenues. Real S&P 500 earnings have risen under 1.8%/year in the last 20 years, and the index has risen rather a lot faster than that.

Yes, a prediction of a big fall sounds extreme, but it doesn't sound so wild if you phrase the same thing as merely expecting things to go maybe half way back to what used to be typical.


Even from 1929-1936 stock losses on an annualized basis were about -3.5% to -5%. And given that was a deflationary environment, it wasn't quite as "bad" as GMO's prediction.

Maybe a little worst than that. Total return before taxes in the 7 years starting in the middle of 1931 was -15.8%/year compounded for the cap-weight predecessor of the S&P 500. As you note, there was deflation rather than inflation, so in purchasing power that would be "merely" -14.6%/year. So the idea that one might see -7%/year starting now is maybe not so outlandish, given that in almost any sense the market is more richly valued today than it was in 1931.

Some things are more likely than others, but anything that has already happened is definitely something that's possible.

Jim