A guide to Canadian Depositary Receipts by account type, and why the RRSP is the worst place for one.
A CDR is not a share. It is a receipt issued by a bank, representing a fractional claim on shares that the bank holds in custody. CIBC launched the product in Canada in 2021 and lists on the TSX. BMO followed with a Cboe Canada listing and has expanded aggressively, adding roughly eighty names across the US, Europe and Japan through 2025.
Three mechanics matter.
The ratio. Each CDR represents some fraction of one underlying share. A ratio of 0.5 means two CDRs give you the economic exposure of one share. This is what makes a 700 dollar US stock accessible at 20 dollars.
The notional hedge. This is the product’s entire pitch. Short-dated FX forwards strip out the currency movement between the underlying and the Canadian dollar. Critically, the hedge does not adjust the CDR’s price. It adjusts the ratio. When the loonie strengthens, your CDR comes to represent more underlying shares. When it weakens, fewer.
The depositary relationship. You do not own the share. The bank does, in custody, and you hold a contractual claim against it under a deposit agreement. Voting is done by instruction, and the depositary uses commercially reasonable best efforts to vote as directed.
That third point gets glossed over in most comparisons and it is the source of every ambiguity that follows.
The Cost That Is Certain
Neither CIBC nor BMO charges a management expense ratio, and both say so prominently. That is technically true and practically misleading.
BMO’s language: the administrator earns compensation for the notional currency hedge which will on average not include a spread greater than 60 basis points annualised. CIBC’s arrangement is the same in substance, capped around 0.60%.
Because the fee is embedded in the FX forward rate and deducted through the ratio, it never appears on your statement as a fee. Your CDR simply comes to represent a slightly smaller fraction of the underlying share each month. You cannot see it, you cannot avoid it, and it compounds.
| Holding period | Cost of 0.60% annually, as a share of terminal capital |
|---|---|
| 5 years | 3.0% |
| 10 years | 6.2% |
| 16 years | 9.1% |
| 25 years | 13.9% |
Now put that against what the dividend actually delivers. On a name yielding 0.26%, the hedge fee is more than twice the entire dividend. On Alphabet, Nvidia, Apple or Amex, you are paying the bank several times what the company pays you, every year, to remove a currency exposure you may not have wanted removed.
For a dividend-focused investor this inverts in an unpleasant way. The higher the yield, the more the whole package costs relative to holding the share, because you pay 0.60% on the position regardless and the withholding question sits on top of it.
The Hedge Is a Cost Even When It Works
Set the fee aside. The hedge itself is not free in an economic sense, because it removes something rather than adding it.
A Canadian who owns Coca-Cola directly holds two exposures: the business and the US dollar. Over a sixteen-year horizon those are two distinct sources of return, and they are not perfectly correlated. When Canada has a bad decade, the loonie usually weakens, and the unhedged US position gains in Canadian dollar terms exactly when the domestic portfolio is struggling. That is a genuine diversification benefit and it is precisely what the hedge deletes.
CIBC’s own material acknowledges the trade, noting that while the hedge reduces exchange rate impact, investors generally do not participate in currency upside either.
If your portfolio already leans heavily Canadian, and most Canadian retail portfolios do, then buying US businesses through a currency-hedged wrapper gets you the earnings diversification and throws away the currency diversification. You are paying 60 basis points a year to make your portfolio more correlated with itself.
The hedge is not wrong in every case. It is wrong when your liabilities are far in the future and your existing assets are concentrated at home, which describes most RRSPs.
Four Things That Do Not Show Up in the Comparison
T1135 still applies. BMO states that its CDRs are specified foreign property and count toward the 100,000 dollar threshold, alongside any directly held underlying shares and other foreign property. The reporting convenience many articles cite does not exist here. Confirm the position for CIBC CDRs separately before relying on it either way.
US estate tax situs is unresolved. BMO says CDRs represent a Canadian issued derivative security but that it cannot confirm whether they would be treated as US situs property for US estate tax purposes. Directly held US shares are unambiguously US situs, which is a known problem with known planning solutions. An unresolved question is worse than a known problem, because you cannot plan around it.
There is no CRA direction. Edward Jones notes that the Canada Revenue Agency has issued no specific guidance on the taxation of CDRs, and that treatment is anticipated rather than confirmed. CIBC’s own wording is that it is the expectation that Canadian tax consequences will match direct ownership. Both issuers are careful with their verbs, and you should notice that.
Withholding reclaims exist, which tells you something. BMO’s site directs CDR holders to a third-party withholding tax reclaim specialist for circumstances where reclaims may be available, while disclaiming any warranty that reclaims will succeed. If the treaty benefit always flowed automatically, that section would not need to be there. The likely reality is that the treaty position is sound but its application at source depends on your broker having the paperwork right, and that when it goes wrong the remedy is a specialist and a process rather than a phone call.
Why the RRSP Is the Worst Fit
Every argument above lands hardest in a retirement account, for three compounding reasons.
The horizon is longest. An RRSP is the account you hold for thirty years. A 0.60% annual drag that costs 3% of capital over five years costs nearly 14% over twenty-five. There is no account where a small recurring fee does more damage.
The treaty benefit is largest and clearest when held directly. The RRSP, RRIF, LIRA and LIF are the only Canadian accounts where US dividends escape withholding entirely under the treaty. That exemption is well established, mechanical and requires nothing of you beyond a W-8BEN with your broker. Through a CDR the issuers say it should apply, no tax authority has confirmed it, and a reclaim industry exists. Direct ownership converts an ambiguity into a certainty for no cost.
The currency hedge is least appropriate. Your RRSP funds retirement decades away. You have no fixed near-term Canadian dollar liability to protect. What you have is a portfolio probably overweight Canadian banks, energy and utilities, which is exactly the portfolio that benefits most from unhedged US exposure.
Put together: the account with the longest compounding period, the strongest tax reason to hold directly, and the weakest case for hedging is also the account where a CDR costs you the most. If you own CDRs in one place, this is the place not to own them.
The Practical Objection, and Why It Is Smaller Than It Looks
The argument for CDRs is convenience. You avoid converting currency, you avoid a US dollar account, you can buy a fraction of an expensive share, and everything is in Canadian dollars on one statement.
The conversion problem has a well-known solution. Norbert’s Gambit involves buying a dual-listed instrument in Canadian dollars, journalling it to its US dollar counterpart, and selling it for US dollars, converting at close to the interbank rate rather than the 1.5% or so a retail desk charges. The cost is two commissions and a settlement wait, typically a fixed cost of twenty dollars or less regardless of the amount converted.
Against 0.60% a year, the arithmetic is not close:
| Position size | Annual hedge fee at 0.60% | Months for a ~20 dollar gambit to pay back |
|---|---|---|
| 3,000 | 18 | 13 |
| 5,000 | 30 | 8 |
| 10,000 | 60 | 4 |
| 25,000 | 150 | under 2 |
On anything above a few thousand dollars held for more than a year, the one-time conversion cost is trivially cheaper than the recurring hedge. And the gambit is a fixed cost, so the larger the position, the more absurd the comparison becomes.
The fractional share argument has also weakened. Several Canadian brokers now offer fractional US shares directly, which removes the accessibility case for expensive names.
What remains genuinely useful: you keep one currency on your statement, and you avoid tracking a US dollar cash balance and its own adjusted cost base. For some people that simplicity is worth 60 basis points. It should be a deliberate choice, not an accident.
Recommendations by Account
| Account | US withholding on direct shares | Recommendation | Reasoning |
|---|---|---|---|
| RRSP, RRIF, LIRA, LIF | 0% under the treaty | Hold direct US shares | Longest horizon, clearest treaty benefit, least need for a hedge. The strongest case against CDRs. |
| Non-registered | 15%, recoverable via the foreign tax credit | Hold direct US shares | Withholding is recoverable either way and T1135 applies either way, so the 0.60% buys you almost nothing. |
| TFSA | 15%, not recoverable | Direct preferred, CDR defensible for small or short holdings | The withholding is a wash. The only difference is the 0.60% and the hedge. Direct wins on any position you will hold for years. |
| RDSP | 15%, not recoverable | Direct preferred | Same logic as the TFSA, with an even longer horizon. |
| RESP | 15%, not recoverable | CDRs are genuinely defensible | Short, defined horizon and a Canadian dollar liability at the end. See below. |
| FHSA | 15%, not recoverable | CDRs are genuinely defensible | Same reasoning as the RESP. |
Where CDRs Actually Win
The framework that makes this coherent is not about registered versus non-registered. It is about horizon and liability currency.
A currency hedge protects you when you know what you will owe, in what currency, and roughly when. An RESP is the clearest case in Canadian investing. Your child will attend a Canadian university in a known number of years and pay tuition in Canadian dollars. If you hold unhedged US equities and the loonie strengthens 15% in the two years before first tuition, you have taken a real loss against a real liability. Hedging that is not a fee, it is insurance with a purpose.
The FHSA works identically. You are saving toward a Canadian house, priced in Canadian dollars, on a horizon measured in a few years rather than decades. Currency volatility is a risk you are being paid to remove rather than a diversifier you are throwing away.
Note also that the 0.60% costs far less over a short horizon. Three percent of capital over five years in an RESP is a real cost but a proportionate one for removing a currency mismatch against a defined liability. Nine percent over sixteen years in an RRSP, to remove a diversifier you wanted, is not.
So the rule of thumb:
- Long horizon, no defined Canadian dollar liability, treaty benefit available: hold the share directly. This is the RRSP, and it is most of your portfolio.
- Short horizon, defined Canadian dollar liability, no treaty benefit either way: the CDR earns its fee. This is the RESP and the FHSA.
- Everything in between: direct ownership unless the convenience is worth 60 basis points a year to you, which is a legitimate answer as long as you have priced it.
What to Verify Before Acting
The tax positions in this article come from issuer documentation and brokerage guidance, not from a tax authority. There is no CRA ruling on CDRs. Both CIBC and BMO frame their tax commentary as expectation rather than fact, and both direct investors to their own advisors.
Three things worth confirming for your own situation:
- Whether your broker applies the RRSP treaty exemption at source on CDR dividends, or withholds and leaves you to reclaim. Ask them directly and get the answer in writing.
- Whether CIBC CDRs carry the same specified foreign property treatment BMO confirms for its own. Assume they do until told otherwise.
- Whether financial transaction taxes apply to any non-US CDR you are considering. BMO discloses 0.20% on purchases of CDRs linked to Spanish companies and 0.40% on French ones, charged on the buy side.
And one operational note. Because the hedge works by adjusting the ratio daily rather than the price, the number of underlying shares your position represents drifts continuously. That makes it harder to reconcile what you own against the company’s per-share reporting, which for anyone doing their own valuation work is a small but persistent friction.
The Bottom Line
The case against CDRs in an RRSP does not rest on the withholding tax argument that circulates online, which the issuers dispute and which is more likely wrong than right. It rests on three things that are not in dispute. You pay roughly 0.60% a year, invisibly, forever. Over a sixteen-year holding period that is about 9% of your terminal capital, and on a low-yielding stock it exceeds the dividend several times over. You are paying that fee to delete a diversifier. A Canadian portfolio overweight domestic banks, energy and utilities benefits from unhedged US dollar exposure, and the RRSP is the account with the longest horizon over which that benefit can accrue. And you are converting a clean, well-established tax position into an ambiguous one. The RRSP treaty exemption for directly held US shares is settled law with a simple form attached. Through a receipt issued by a bank, it becomes a position the issuer expects to hold, that no tax authority has confirmed, and for which a reclaim industry has grown up. None of that makes CDRs a bad product. It makes them a product with a specific job. That job is hedging a short-horizon portfolio against a Canadian dollar liability you can name, and the RESP and FHSA are where that job exists. In a retirement account it is a 60 basis point solution to a problem you do not have.
Disclaimer: This analysis is for informational purposes only and does not constitute investment advice. Always perform your own due diligence or consult with a financial advisor before making investment decisions.
