1. I never said allocation according to risk theory was what most people do. Many retail investors still use high cost funds, stock picking advisors etc. and we don't view that as a good argument against low costs or indexing. Other BH precepts aren't a popularity contest, so I don't see why this one would have to be.1. I agree with this. I was noting that the average market participant has not shifted their asset allocation significantly between early 2021 and today, despite what the Merton Fraction might imply about what shifts would be rational (and, in fact, has gone slightly in the "wrong" direction).Allocating at the average is 'invest at a risk appropriate to somebody else', unless I'm average.
2. I was not trying to imply that the "starting point" of the global market portfolio needed to be the "ending point" for every investor, and I think it is reasonable for investors to tilt towards or away from stocks relative to the market portfolio based on their personal risk tolerance.
3. As an aside, I would appreciate you editing your comment as you appear to have accidentally snipped the quote box in a way that made it look like I said the words "Looks like the world's holding good ol' 60/40!" when that was said by a different poster.
2. But again even using 'market portfolio' as a benchmark, bond/stock allocation of 'the world' might be more like 35/65 stock/'bond' if you count *all* fixed income, not just 'investable' from US retail POV, keeping in mind that institutional and foreign investors whose holding of 'investable' securities are included in those indexes also have a lot of the FI not included in it. I'm not saying to use 35/65 as a benchmark, just that nailing down the 'global portfolio' is difficult. Fortunately I don't think it matters much what the global portfolio really is. I do accept that 60/40 is a convention for 'middle of road' risk taking in US retail investor context. I just don't think it's obviously related to the complete market portfolio.
And again the (unsettling) point is 'tilt based on personal risk preference' only theoretically translates to constant % stock for constant personal risk preference if the expected ERP and variance are constant. Which isn't just true w/ MF's assumption of CRRA utility function. That makes it end up a simple formula. Other assumptions of utility function result in differential equations that have to be solved numerically but also tend to give lower stock allocation for lower ERP/var with the same personal risk preference parameter(s). Which has some intuitive validity IMO. Assuming ERP/var constant didn't always seem far off, but recently w/ real bond yield up 3%, ~unchanged stock valuation, it kinda does.
3. I fixed it.
Statistics: Posted by JackoC — Sun Aug 11, 2024 12:10 pm — Replies 3587 — Views 870585