Bill Bernstein wrote: ↑Wed Mar 10, 2021 8:24 pm
The assumption that some are making here is that the market price of the ETF is a more accurate measure of the NAV than the NAV determined by the third-party service used by the fund company to determine the NAV of the open-end fund.
That, as they say, is a strong assumption; one has to ask oneself: which the real price in a bad state of the world: the estimate made by the TPS who has access to thousands of transactions of the same and similar bonds made during that day, or that of panicked shareholders trading the ETFs?
The question answers itself.
So, yes, there is an an arbitrage opportunity here; as pointed out, it's the person who wants to buy the ETF and sell the closed end fund.
But that's not the person I'm talking about. There's no arbitrage available to the person I'm concerned with, which is the investor, who, in the ordinary course of rebalancing their stock allocation back to policy in a bad state of the world, is selling an open-end bond fund to buy a stock fund.
As the OP I wanted to see if I could sum up this thread into something instructive to the average individual's IPS. I've replied to Dr Bernstein, as I don't want to say/infer something he never intended, but am open to all feedback.
First, Dr Bernstein's position is that the purpose of bonds is to be as risk-less as possible consisting of 1) plain treasuries/CDs less than 5 years in maturity followed by maybe 2)a total bond market mutual fund. So there are two distinctions here, the removal of corporate bonds and then how to own them, ie individual bonds/mutual fund/ETF.
Secondly, he doesn't have a singular
purpose for the risk-less portion. For example, in The Four Pillars of Investing he advocates including one's six month emergency fund as a component. In fact there could be various length duration requirements within the bond side requiring different instruments. While opposite to the simplicity of the 3 fund portfolio, it's not better or worse, just distinct to the needs and consistency of the investor yet important in framing the discussion.
It's apparent that there are times that ETF and fund prices diverge causing in a non-optimum "hair cut" if the ETF owner has a need
to sell. The extent of this affects all bond ETFs and yet a little more so to corporates and thinly traded bond ETFs. Right?
This market may last for days/weeks (like last year) or theoretically much longer. Therefore, it is obviously prudent that to the extent one has their short term/emergency fund cash needs as part of their risk-less allocation, that those monies should absolutely NOT be in an ETF bond product. Mr. Bernstein's "bad state" world is NOT 2020. It's a hypothetical risk but easy to avoid.
The opportunistic rebalancing portion is a bit tougher. First, let's assume a SET 80/20 bond allocation with the bond side short duration, ie constant value, for simplicity. If the stocks slide 50% and I rebalance at the most opportunistic time, only 40% (8% of the 20%) is needed for rebalancing and should therefore NOT be in a bond ETF.
So what I'm left with is a risk-less allocation that looks like this:
- 40% non-ETF (preferable individual bonds if no credit risk) for rebalancing+
- x% of non-ETF funds for short term emergency needs (to the extent one includes their EF in their portfolio..please no thread drift +
- 60-x% of anything else (ETFs are okay)
I realize this is very generalized and the added complexity in implementation may not be worth it, however, I DO think it is a valid compromise between the opinion of Dr Bernstein and an investor who may have 100% of their fixed income in BND.
Only thing I'm not done with looking at is:
Just how much more liquid are T-Bills than Notes during a crisis (as in 2020)?