Jim, do you still have confidence in GMO's approach despite its fairly dismal track record over the last ~15 years? Is it your opinion that the underlying basis for their estimate methodology is sound, and that it's likely to return to being a decent predictor of medium-long term returns again if valuation multiples someday stop their seemingly endless expansion?

Absolutely. Provided you remember what it is. It's not a market prediction. Rather, it's the return you should reasonably expect, if valuations are pretty typical at the end date. And nothing super unusual happens*.

As mentioned, I think perhaps they use some history that's old enough that the valuation multiples of that time are perhaps no longer entirely relevant, for a variety of fairly subtle reasons. Corporate profitability truly is higher, and the drag of net share issuance had a step change a while back. So I think the valuations they assume are "normal" might be just a hair conservative.

But yes, I think the methodology is extremely sound. Absent better information (and who has that?) one should expect a typical result. The extent that you expect a different result for a given security should be proportional to the strength and rigour of the evidence you have that the specific security in question is exceptional. If you hear hoof beats, expect horses, not zebras...unless you have a photo out the window that contained a lot of stripes.

Jim

* As an aside, the TCJA was a good example of something pretty unusual and unexpected. The 10 year average of US corporate tax to GDP was a little over 1.9% in the decade to 2000. It's around 1.6% recently. That sounds like a small shift, but it means every dollar of sales is worth 19% more than it was before.