^^^^^ I dunno. I’m not equipped to go toe to toe with his charts, at least until I have more coffee!
️. ....
I'm also on my first cup, and am unlikely to dig as deep as would be needed to evaluate this anyhow. And w/o that, I am willing to accept that he did this correctly, it would not be that hard (but not so easy either), to simulate all this in a spreadsheet, and he probably already has the framework for doing it.
... But with apologies for filling the page with nitpicks, his strategy just doesn’t mesh with my remain-fully-invested thinking:
- He doesn’t account for the investment opportunity cost of cash-drag. I feel my Portfolio Visualizer models above did a better job at that. He is not working with an actual 80/20 portfolio but with something less aggressive, like 60/20/20 or 70/20/10. ...
I'm not sure he shows his work, but why wouldn't the cash drag be included? I'd assume he only calculates gains/losses on the invested amount.
And Portfolio Visualizer isn't doing the "withdraw the cash only in a downturn" thing is it? Though as I think I figured, it would be drawing the cash/bond side and buying stocks in a rebalance, which should be close.
Hmmmm, on my first cup still, but I guess you could do a sequence of 1966 out a few years, starting with that cash instead of bonds, take the value/year when the cash is depleted, then start another series going forward from that date with the reduced portfolio amount and the 'normal' AA?
...
- Does he account for dividends post-crash? Dividends never stop in a stock index fund and 2 or 3% dividends over ten years is real money.
- The one-time cash injection/emergency parachute is novel. However, spending the one-time cash in a downturn leaves an investor with a more aggressive portfolio as one ages. OK, but we don’t often see that happen in real life here, do we? Folks tend to get more conservative as they age. ....
Why would you think he isn't accounting for dividends? That is straightforward, I don't get the impression that this guy would make that kind of a newbie error (but I haven't really read much of his work).
Isn't he rebalancing? Yes, if the downturn happened and you spent the cash, you'd go from a more conservative overall AA to a less conservative one, but... since I believe he compared it to a larger buffer with the same AA, you'd be going from more conservative to your 'regular, planned AA' - so you're not getting less conservative than your original plan, right? You just slide in (up) to it over time.
...
- I’d prefer to see a comparison model between his (honestly) approx. 60/20/20 initial portfolio with a 60/40 fully-invested one, in which one simply spends down the bond funds side instead (which are likely growing in value due to Fed interest rate cuts to get out of any corresponding recession and stock price swoon.) ....
Isn't that what he did? Maybe with different Eq/Fixed/Cash ratios, but that concept?
...
Regardless, if one’s plan is to rebalance back to their initial allocation, I’m convinced for myself that it doesn’t matter when one does it. The portfolio value ends up about the same in the long run, whether the rebalancing happens daily, yearly or randomly in slow motion, e.g spending down cash and bonds while stocks crash and recover, then selling stocks to rebalance.
However, ERN isn’t doing that. He’s going from less aggressive prior-crash to permanently more aggressive post-crash. And what if the crash doesn’t happen during his retirement? His pile of cash earns virtually nothing for several decades, in fact, losing ground due to inflation.
I'm also rather agnostic about rebalancing, based on the research I've seen/done.
For the bold part, I assume you mean
early in the retirement (the usual SORR worry), and I share that concern. Crashes happen after run ups. If I hold a lot of cash early in my retirement, and that crash comes along after, say 5 years, I am less prepared as I did not share so much in the run up.
I dunno, but I wonder if using his approach just makes a different year the worst year in the database? Maybe it just shifts it from 1966 to some other starting year? Are we getting into data mining? This approach would need to be run on every cycle to determine that.
So as is usual, I will do nothing but think about it, and spend too much time typing!
Actually, the deltas aren't really all that great, and history won't repeat exactly, so who knows? But it is kind of fun to think about, and I want to keep an open mind to alternate ideas.
-ERD50