CDRs versus U.S. stocks: Which is better for Canadian investors?

Take Microsoft as an example. If you search the ticker MSFT, several options may appear. One is the actual Microsoft stock listed on the NASDAQ which, as of May 8, 2026, traded around US$416 per share. But you may also notice alternatives sitting right below it, such as the CIBC Microsoft CDR CAD Hedged under the same ticker, and the BMO Microsoft CDR CAD Hedged under ticker ZMSF.

These are Canadian depositary receipts, or CDRs. MoneySense previously covered the mechanics of CDRs in detail in a 2025 article, but the short version is that they allow Canadians to gain exposure to U.S. stocks (and now international stocks) in Canadian dollars with a built-in currency hedge. 

That currency hedge is designed to offset fluctuations between the Canadian and U.S. dollars. Investors still receive exposure to the underlying stock’s dividends, though those remain subject to the standard 15% U.S. withholding tax.

At first glance, they can look very appealing. One reason is accessibility. A single share of Microsoft costs more than US$400, while the CIBC Microsoft CDR traded around $29.35 Canadian and the BMO version around $8.32 CAD. For investors without access to fractional shares, or those who simply prefer not to transact in U.S. dollars, that lower entry price can make building a portfolio feel much easier. And if your brokerage charges a higher commission on U.S.-listed stocks than Canadian-listed issues, you could save on your transaction costs as well.

But CDRs are not a free lunch. The built-in currency hedge comes with a cost. Depending on the provider, that fee can range from roughly 0.6% to 0.8% annually. Over time, those costs can add up, particularly when compared to simply holding the underlying U.S. stock directly.

That raises an important question: historically, how would a Canadian investor have fared owning the CDR version of a U.S. stock instead of the stock itself? More specifically, after accounting for currency hedging costs and foreign withholding taxes, how different would the longer-term returns have been?

Answering that helps clarify when CDRs make sense, versus when owning the underlying U.S. stock directly may be the better option for Canadian investors.

Quantifying CDR tracking error versus U.S. stocks

To test this, I back-tested two widely traded and long-standing blue-chip CDRs against their underlying U.S. stocks. I specifically wanted to look at two different situations. 

The first involved a dividend-paying company, where the CDR investor would face both currency hedging costs and the drag from the 15% U.S. withholding tax on dividends. 

The second involved a company that does not pay dividends, to isolate how much of the difference could simply be attributed to the currency hedge and other structural frictions.

All data was sourced from PortfolioVisualizer.com using the longest common return period available for both the stock and its corresponding CDR. Returns are presented net of management costs, but before taxes, brokerage commissions, or implicit trading frictions such as bid-ask spreads.

The first comparison was The Coca-Cola Company (KO) versus the CIBC Coca-Cola CDR (COLA). From January 2023 to April 2026, shares of Coca-Cola compounded at an annualized 9.76% with dividends reinvested. The CDR lagged behind at 8.14% annualized. That is a noticeable 1.62% difference.

Source: Portfolio Visualizer

To break down where some of that drag likely came from, we can start with the currency hedge. Assuming the higher end of the provider’s estimated currency hedging cost, around 0.6%, adding that back to the CDR’s return brings it to 8.74%. Then there is dividend withholding tax. 

As of May 8, 2026, Coca-Cola’s trailing five-year average dividend yield was 2.89%. Applying a 15% withholding tax to that yield results in another 0.43% drag. Even after accounting for both the hedge cost and dividend withholding (a total of 1.03%), the CDR still trails the underlying stock by a 0.59% variance.

Of course, these calculations are somewhat back-of-the-napkin in nature, but the broader point still stands: there appears to be some additional drag for CDRs beyond just the headline currency hedging spread and foreign withholding tax on dividends.

For my second example, I used a stock that historically paid little to no dividends (Amazon) and over a shorter period from January 2026 through April 2026. The results were still weaker for the CDR, though the gap narrowed to 0.99%. 

Source: Portfolio Visualizer