If you hold investments at more than one broker, at some point you will want a single number: how did my money actually do this year? It is a surprisingly hard number to get right, and almost every shortcut people reach for is wrong. You cannot add the returns your brokers report. You cannot average them. And the "gain divided by what I put in" figure most people compute in their head is not a rate of return at all.
The good news is that there is a correct answer, it is not complicated once you see it, and it only requires the cash flows you already have. This guide explains why the shortcuts fail, walks through the two returns that matter with worked Canadian examples, and shows how to fold a US-dollar account into a Canadian-dollar figure without distorting it.
Why you cannot average broker returns
Each broker computes a return for the money inside its own walls, over its own period, using its own method. Two of those numbers do not combine by averaging, because averaging throws away the two things that actually matter: how many dollars were in each account, and for how long.
A concrete case makes it obvious. Two accounts, one calendar year:
- Broker A: $10,000 invested on January 1, worth $11,500 on December 31. No deposits or withdrawals. That is a 15% return.
- Broker B: $5,000 deposited on July 1, worth $5,000 on December 31. Flat. That is a 0% return.
Average the two percentages and you get 7.5%. But 7.5% is meaningless here. Broker A's 15% was earned on twice the money, held for twice as long, as Broker B's 0%. The accurate blended figure has to give Broker A far more weight. Weight each account's gain by the capital and time actually at work and the combined return lands near 12%, not 7.5%. The simple average is not a little off; it is answering a different question than the one you asked.
The number that is actually right: money-weighted return
The figure that correctly blends dollars and time is the money-weighted return, also called the internal rate of return or, in a spreadsheet, XIRR. It is the single annual rate that, applied to every dollar for exactly as long as that dollar was invested, reproduces your ending value.
You feed it a dated list of cash flows: money going in is negative, money and value coming out are positive. Take the two accounts above as one combined series:
- January 1, 2026: deposit $10,000 (into A) → -10,000
- July 1, 2026: deposit $5,000 (into B) → -5,000
- December 31, 2026: total value across both accounts $16,500 → +16,500
The money-weighted return is the rate r that solves:
10,000 x (1 + r) + 5,000 x (1 + r)^0.5 = 16,500Solving gives r of about 12.1%. That is your real, dollar-weighted return across both brokers for the year. Notice it landed near the 12% we reasoned to above, and nowhere near the 7.5% simple average.
Now compare it to the shortcut most people use in their heads: total gain over total contributions. You put in $15,000 and ended with $16,500, a $1,500 gain, which reads as 10%. But 10% understates you, because the second $5,000 was only invested for half the year. The money-weighted return credits you for the fact that most of your capital worked the full year. 10% is a fact about your dollars; 12.1% is your rate of return. They are not the same thing, and only one annualizes correctly.
The other number: time-weighted return
There is a second return, and it answers a different question. The time-weighted return deliberately removes the effect of your contribution timing. It breaks the year into sub-periods at each cash flow, measures the pure investment return in each, and links them together geometrically.
Why would you want to erase your own deposit timing? Because it isolates the investments. If you want to know whether your fund selection beat the S&P 500 or a Canadian aggregate bond index, you need a number that is not flattered or punished by the lucky or unlucky timing of when you added cash. An index has no deposits, so to compare fairly, neither should your return. That is the time-weighted return, and it is the one fund managers are required to report for exactly this reason.
The rule of thumb:
- Use the money-weighted (XIRR) return to answer how did my actual dollars do?
- Use the time-weighted return to answer did my investments beat the market?
For a household with money at several brokers, the money-weighted return is usually the one you feel in your account balance, and the time-weighted return is the one that fairly grades your choices. A big deposit right before a rally will lift your money-weighted return above your time-weighted return; the same deposit right before a drop will do the opposite.
Folding in a US-dollar account
Canadians running multiple brokers almost always have at least one US-dollar account, and currency quietly changes the answer. Your return in Canadian dollars is the investment's own return combined with the move in the exchange rate over the same window.
Worked example. A US holding rises 10% measured in US dollars over the year. But the loonie strengthens, so one US dollar buys about 5% fewer Canadian dollars at year-end than at the start. Your return in Canadian dollars is not 10% and it is not 5%; it is the two multiplied:
1.10 x 0.95 - 1 = 0.045, about 4.5%Currency erased more than half the gain. If the loonie had weakened 5% instead, your CAD return would have been about 15.5%. This is why a US-dollar account's own reported return is not your return, and why you cannot drop its raw percentage into a blended figure.
The fix is mechanical: to combine a USD account with CAD accounts, convert every deposit, every withdrawal, and the ending value to Canadian dollars at the exchange rate on each transaction's own date, then run one money-weighted calculation on the fully-CAD series. Convert first, blend second. Do it in the other order and the currency move contaminates every account's weight.
Getting the cash flows in one place
The calculation is not the hard part; assembling the inputs is. A correct money-weighted return needs every dated deposit and withdrawal across every account, plus one ending value, all in the same currency. When those transactions are spread across three brokers, two of which export CSVs in different date formats and one of which reports a US-dollar account, the reconciliation is where the hours go and where the errors creep in. This is a different job from simply seeing your balances side by side, which the companion guide on tracking multiple brokerage accounts covers; here the point is the math on top of those balances.
How Mozaic helps
The Mozaic investment tracker reads the transactions and balances from your accounts across the major Canadian brokers (read-only, via SnapTrade and Plaid), converts every cash flow to Canadian dollars at the rate on its date, and computes one return across all of them, so the dated deposit-and-withdrawal series a money-weighted return depends on is assembled for you instead of stitched together from three CSV exports. Because it holds every account's flows in one place, the blended figure weights each broker by the dollars and time actually at work, the correction the simple average throws away.
The connection is read-only; Mozaic cannot trade or move money, only read what settled. It does not give tax advice, and a personal rate of return is not a tax figure. It gives you the one accurate number across brokers that no single broker's statement can. Data is stored in Google Cloud's Montreal region under PIPEDA and Quebec Law 25 (/security).
The bottom line
You cannot add or average the returns your brokers report, because those numbers discard how many dollars sat in each account and for how long. The money-weighted return (XIRR) blends them correctly and tells you how your actual dollars did; the time-weighted return removes your deposit timing and tells you whether your investments beat an index. For any US-dollar account, convert every flow to Canadian dollars on its own date before you blend, because the currency move is part of your real return.
If you would like your real return computed across every broker in Canadian dollars without exporting a single CSV, the 14-day free trial needs no card, and the guide on how USD accounts affect your Canadian net worth covers the currency side in more depth.