As was widely expected, there was no change in interest rates on either side of the border. The Bank of Canada stood pat for a third month with a policy rate of 2.75% and the Federal Reserve bank stayed the course for the fifth straight month at a rate between 4.25% and 4.5%.
These rates underpin things like the HISA table I update monthly; as a result, I wouldn’t expect much in the way of change for my upcoming August update.
The next announcements from the two are scheduled for September 17.
I have a dedicated non-registered account in my retirement portfolio that is the cash cushion for VPW’s decumulation strategy. You can read about the details of how I currently get paid in retirement here.
That non-registered account holds about 85% Canadian dollars, invested in ZMMK, with the remaining 15% invested in ICSH. Both of these ETFs are very short-term bond funds and give me a slight advantage over investing in zero-risk HISAs. ZMMK and ICSH are part of my ETF all-stars lineup, and I track HISA rates on a monthly basis.
The fact is that US interest rates are a lot higher than Canadian interest rates, almost 2% higher as of July 2025. It seems to me that I should take advantage of that fact. Taking advantage of this situation would mean selling some ZMMK, performing Norbert’s Gambit with the resultant cash, and then buying ICSH. There are costs involved at every step of the way1:
So at what point is it worth it? Let’s do a bit of math using the following assumptions:
So for various amounts, the time to profitability6 of doing the Gambit looks like this.
| $ CAD changed | Journaling Fee7 | DLR Spread Fee8 | ZMMK/ICSH spread | Total cost | TTP9 |
|---|---|---|---|---|---|
| $1k | $12 | $1.70 | $0.40 | $14.10 | ~10 months |
| $10k | $12 | $17.10 | $4 | $33.10 | ~10 weeks |
| $100k | $12 | $170 | $40 | $222 | ~6 weeks |
So clearly, for amounts around $1k this isn’t such an attractive proposition as the costs will take a fair bit of time to be negated by the bump in interest rates. For larger amounts, I’d say it’s worth it. Given ZMMK hasn’t yet paid out its dividend for the month, I guess I’ll wait until I’m ex-dividend (July 30, 2025, per the fact sheet) before doing this transaction.
This is a (hopefully monthly) look at what’s in my retirement portfolio. The original post is here. Last month’s is here.
The retirement portfolio is spread across a bunch of accounts:
The target for the overall portfolio is unchanged:
You can read about my asset-allocation approach to investing over here.
As of this morning, this is what the overall portfolio looks like:
The portfolio is dominated by my ETF all-stars; anything not on that page is held in a non-registered account and won’t be fiddled with unless it’s part of my monthly decumulation. Otherwise I’ll rack up capital gains for no real benefit.
There weren’t big changes this month. My monthly decumulation from my RRIF accounts involves selling enough XGRO to meet RRIF-minimum payments, and the rest of my retirement paycheque is funded by my non-registered accounts. This month, given the run in the US stock market of late, that involved a sale of some shares of HXS1.
The asset-class split looks like this
The bond portion of the portfolio is a little smaller than I would like. The targets for my portfolio are unchanged:
The change in the bond portion of the portfolio was amplified because I hadn’t updated the asset split of AOA in my multi-asset tracker in a while. AOA has drifted quite a bit since it only rebalances twice a year (next time in October). More on drifting in multi-asset ETFs here.
The retirement savings look quite healthy; even though I’ve been drawing a monthly salary for 7 months, I’m now ahead of where I was when I started my retirement journey. This is aligned with what my retirement planner told me to expect, but as you can see, the journey has had some interesting ups and downs already.
My VPW-calculated salary has hit a new high this year, a dizzying 0.77% higher than my first draw in January. This stability is thanks to the built-in shock-absorber of the VPW model (a 6-month cash cushion which smooths out the market gyrations considerably). I also think it’s an endorsement of my choice to take retirement payments monthly; my exposure to short-term market hiccups is greatly reduced since I’m not making big sales of ETFs to fund a year of spending all at once.
I rely a lot on asset-allocation ETFs in my retirement portfolio, mostly XGRO in the CAD side of the portfolio and AOA on the USD side1. These ETFs (about 70% of my overall retirement portfolio, as you can see here), like all asset-allocation ETFs, rebalance their holdings periodically in order to stick to their asset allocation targets. This aligns perfectly with my way of investing; I’ve always tried to stick to my asset allocation targets portfolio-wide, assisted by tools like my multi-asset tracker spreadsheet. (If you aren’t familiar with asset-allocation as an investment strategy, you could give this article a read.)
XGRO’s asset allocation targets are written right in the prospectus2:
Anyway, XGRO’s approach to making changes to the portfolio in order to maintain this target percentage is written in the prospectus too:
XGRO’s portfolio will be monitored relative to the asset class target weights and will be rebalanced back to asset class target weights from time to time …XGRO’s portfolio is not expected to deviate from the asset class target weights by more than one-tenth of the target weight for a given asset class
Page 419 of the iShares Prospectus (June 2025)
Now “from time to time” isn’t terribly precise. I thought I’d take a closer look at the history of XGRO’s asset allocations. So I dug through annual and semi-annual reports as well as the website. I focused on the Fixed Income (aka Bond) proportion of XGRO over time because that’s the asset class that’s most likely to drift lower3…equities typically outperform fixed income historically. So this is what I found:
So there is a bit of drift in the fixed income portion of XGRO, but in the past year I haven’t seen it off by more than 1.2%, meaning that the promise made in XGRO’s prospectus is being adhered to.4
Turning now to AOA, the fixed income proportion is clearly stated to be 20%, and rebalancing is stated to happen twice annually, in April and October. After that, things become a bit harder to work out56. The various equity contributions are determined by the target index, namely the S&P Target Risk Aggressive Index, which are constructed by using market capitalization of the various indices used7.
Anyway, like XGRO, what I’m most concerned about is the fixed income portion of AOA, and digging through the various reports, I came up with this:
Of late, the fixed income portion AOA has become small, almost 2% lower than it should be. And given that AOA is about 50% of my holdings, it means that my equity exposure is quite a bit higher than I would let it drift myself.
I suppose the next rebalancing in October 2025 will correct this, but I admit it makes me a little uneasy to see that sort of volatility in the asset allocation8. I could of course just sell some AOA and reinvest it in some bond fund (AOA uses IUSB and IAGG, which seem like fine choices) but then I’m just working around the asset allocation strategy I’m paying for in AOA’s management fees, which seems dumb. Not to mention that anything I do now will almost certainly have to be undone come October.
So I guess this all means I should just let sleeping dogs lie. I have minor bits of money to reinvest every month (I still contribute to my TFSAs) so using those funds to buy bonds are probably what I’ll do. It’s a tiny pre-correction that should be addressed come October…or by the next equity meltdown.
One nagging concern I’ve heard about index investing1 is that you can end up overly concentrated in a handful of stocks. For example, as of today, per XGRO’s fact sheet, for every $100 I have invested in XGRO, I’m actually investing almost $5 between two companies, NVIDIA and Microsoft. And for every $100 I have invested in AOA, it’s $7 in these two companies. Although that sounds like it might be a significant concentration, I found a way to test the correlation using https://www.portfoliovisualizer.com/asset-correlations. As you can see, XGRO and NVIDIA are not terribly well-correlated:
…even though perhaps I perhaps wish they were (sigh):
Anyway, if you’re still not comfortable with having too much concentration in your holdings, there are ETFs that limit exposure to any one stock in an index. iShares launched an ETF last year that tracks the S&P 500 while limiting the contribution of any member of that index to 3%. That product is called XUSC, which serves as a complement to its XUS uncapped S&P 500 ETF2. So when you compare XUS’s top holdings to those of XUSC, you can quickly spot the difference:
| Top Holdings | XUS (S&P 500) | XUSC (capped S&P 500)3 |
|---|---|---|
| NVIDIA | 7.86% | 3.49% |
| Microsoft | 7.09% | 3.12% |
| Apple | 5.90% | 3.07% |
| Amazon | 4% | 3.07% |
| Meta | 2.86% | 2.93% |
| Broadcom | 2.49% | 2.96% |
| Alphabet class A | 2% | 1.66% |
| Tesla | 1.73% | 2.05% |
| Alphabet class C | 1.63% | 1.35% |
| Berkshire Hathaway | 1.62% | 1.93% |
For the TSX, there’s not a full equivalent. The closest pairing for the TSX is XIC4 and XIU:
| Ticker | Index | Capped? | # of stocks |
|---|---|---|---|
| XIC | S&P/TSX Capped Composite | Yes | 214 |
| XIU | S&P/TSX 60 | No | 615 |
XIC holds a lot more companies, and in addition to capping any single company, this has the effect of shifting the emphasis to smaller companies. So does this make a big difference in performance? Let’s check using this calculator.
So not much difference; over the past 5.5 years, we see a very slight edge to XIU, the uncapped TSX60 fund.
Whether the relationship between XIU and XIC will be the same as XUS/XUSC is anybody’s guess. In my own case, I don’t own the capped version of the S&P 500, but I do own XIC over XIU, mostly because of its significantly lower MER (0.06% versus 0.18% for XIU).
The Money Engineer 