No free lunch – CASH ETFs revisited

On October 31, 2023, OSFI gave out a huge “trick or treat” to high-interest savings ETF owners (CASH, PSA, etc.) in the following technical bulletin.

There is a transitional period until January 31, 2024 whereby banks and HISA ETF operators will come to a separate rate structure that will most likely involve another haircut of gross yield – the media quoted one analyst from TD that claimed it would be about 50bps, but my suspicion it will be about half of that.

This will still make cash ETFs competitive, but not the no-brainer compared to alternatives, which include high-credit short-duration corporate bond funds.

In other words – there is no free lunch.

CASH.to right now yields a net 5.18%, while the nearest corporate bond alternative (looking at ZST.to as a reasonable proxy for this – they survived March 2020 fairly intact) is netting 5.47% for half a year duration risk. Government of Canada 6 month debt is at 5.11%. CBIL.to is an average 1.8 month duration government bond ETF that yields a net 4.92%. Another relatively innovative product is target date maturity ETFs (looking at RBC’s RQL.to), which has an average duration of 0.63 years and a 5.16% net yield and by all accounts looks inferior to ZST.

What other proxies for Canadian short-term investment-grade corporate debt are out there? Most of them have duration of 2-4 years which involves a significant interest rate component exposure. The other question is whether half a percentage point is adequate compensation for (albeit very low) credit risk.

The yield curve continues to remain heavily inverted – 1 year to 30 year is roughly a -146bps differential, while the more quoted 2-10yr spread is -80bps.

How not to sell an ETF

If you ever wanted to liquidate $2 million dollars of the CASH.to ETF in the last four seconds of trading, look what happens!

I could not imagine went on in the mind of somebody punching this order into the computer. Reading the tape, those sitting at around $49.75 on the limit would have received a reasonable fill and this is the financial equivalent of picking up the freest half a basis point on the planet.

Cash ETFs – Revisited

With the increase in Bank of Canada interest rates:

Cash ETFs:
(TSX: CASH) – gross yield 479bps – MER 13bps – net yield 466bps
(TSX: CSAV) – gross yield 396bps (November 24 distribution * 12, note not adjusted for recent rate increase(s)) – MER 16bps – this should be roughly in-line with the others
(TSX: HSAV) – gross yield 475bps – MER 12bps – net yield 463bps – CAUTION: trading above NAV, do your calculations accordingly
(TSX: PSA) – net yield 459bps

IBKR will give 368bps above CAD$13k.
HISA rates (one example): Home Trust gives 365bps.

Low duration, low risk, liquid:
(TSX: ZST) – 6 month effective duration – YTM 490bps – MER 16bps – net yield 474bps (November 30, 2022 numbers, note roughly 2/3rds investment-grade corporate debt here)
(TSX: XSB) – 2.7yr effective duration – YTM 410bps – MER 10bps – net yield 400bps

Sacrifice liquidity for yield:
GIC Direct is reporting 550bps rates (1 to 5 years).

US Dollar:
(TSX: HISU.U) – Cash ETF – MER 15bps
(TSX: PSU.U) – Cash ETF – MER 15bps
(NYSE: IBTD) – 0.58yr effective duration – YTM 454bps – MER 7bps – net yield 447bps – matures end of 2023
(NYSE: VGSH) – 1.9yr effective duration – YTM 460bps – MER 4bps – net yield 456bps

Cash ETFs

I’m sitting on some relatively heavy losses (0.7%) in my investment in the Canadian Short Term Bond Index ETF (TSX: XSB). The loss was a result of speculating that an average 3-year yield would perform better than a short-term cash account. Needless to say, it is annoying that had I just kept the cash in the account (even earning zero-yield, as Questrade generously offers on cash) I would have done a lot better.

Now that the Bank of Canada has cranked up short-term interest rates even further, the yield curve is flatter and spot yields are about 70bps lower than 1-year yields. The compensation for duration is quite low.

So while I contemplate flip-flopping to even shorter maturities than 3.11 years in exchange for a boost-up of about 100bps on spot rates, we look at the cash ETF options (alphabetical order by ticker symbol):

(TSX: CASH) – info – gross yield 378bps – MER 13bps
(TSX: CSAV) – info – gross yield 296bps (pre-BoC 75bps increase on Sept 7) – MER 16bps
(TSX: HSAV) – info – gross yield 375bps – MER 12bps
(TSX: PSA) – info – net yield 359bps – MER 17bps

All four of the above trade with penny spreads on the TSX in sufficient liquidity volumes.

For comparison, Interactive Brokers offers 276bps on their CAD cash yields. The question is whether it makes sense to flip cash for near-cash, and for each $10,000 transaction, it would cost approximately $2 in commissions to make a trade for approximately 90bps of yield, good for a $90/year difference, or $7.50 monthly.

The question then becomes whether the structural risk of the cash ETF (What happens if the market makers decide to quit? What happens in a true market crisis where everything blows up? What happens if the “cash” investments the ETF invests in goes belly up? What happens if the custodial arrangements the ETF management has turns out to be fraudulent or defective, etc.?) becomes worth it for a $7.50/$10k monthly difference. Is it worth a 360bps difference? Likely. Is it worth a 90bps difference? Probably not.