What the 2022–2023 rate cycle did to borrowed dividend strategies

If you want a single case study in why a “spread” between a borrowing rate and a distribution yield is not a spread, this is it. The numbers below are from the Bank of Canada’s published series, and they are worth sitting with.

The setup

By April 2020, prime had fallen to 2.45% — the lowest it had been in the modern era, arrived at through three emergency cuts in a single month. It stayed there for almost two years.

That was a genuinely attractive environment for borrowing to invest. A secured line of credit at prime, or prime plus a small spread, against a Canadian income fund yielding 6% left several points of daylight. The arithmetic was easy, the cushion looked thick, and a lot of people did it.

What happened next

Effective	Prime
1 April 2020	2.45%
9 March 2022	2.70%
20 April 2022	3.20%
8 June 2022	3.70%
20 July 2022	4.70%
14 September 2022	5.45%
2 November 2022	5.95%
14 December 2022	6.45%
1 February 2023	6.70%
14 June 2023	6.95%
19 July 2023	7.20%

From 2.45% to 7.20%. Nearly three times the cost, in sixteen months.

Look at July 2022 in particular: a single move of one full percentage point. Anyone whose plan assumed rates move in gentle quarter-point increments had that assumption tested in one day.

What it did to the position

Take the simple case. $100,000 borrowed at prime, holding a fund yielding 6%.

• April 2020: cost $2,450, income $6,000. Spread $3,550.
• July 2023: cost $7,200, income $6,000. Spread negative $1,200.

The position went from generating $3,550 a year to costing $1,200 a year, and the investor did nothing wrong and made no decision. The rate simply moved.

And that understates it, because the second variable moved too.

Both sides were floating

The arithmetic above holds the distribution constant at 6%. In practice, rising rates put pressure on exactly the sectors Canadian income investors hold — REITs, utilities, anything with debt to refinance. Unit prices fell, and some distributions were reduced.

So the realistic version is worse than the table: borrowing cost tripled while the income side softened and the collateral value fell. Three things moving against the position at once, which is the general character of leverage — the risks are correlated, and they arrive together.

This is the part the spread calculation misses entirely. It treats two independent numbers. They are not independent, and their correlation is unfavourable.

What it did to margin

For anyone borrowing against the portfolio itself rather than against a house, there was a fourth problem. Falling unit prices reduce collateral value, and a margin call arrives when prices are lowest — forcing sales at the worst possible moment, crystallising losses that would otherwise have been paper.

A HELOC secured against a house does not do this, which is a real structural advantage of that route and one of the few genuine arguments for it.

The recovery, and what it does not prove

Rates came back down. Prime fell through 2024 and 2025 and reached 4.45% in November 2025, where the series ends. Positions that survived became viable again.

The word doing the work there is survived. The investor who held on has been fine. The investor who was forced to sell in 2022 crystallised a loss and did not participate in the recovery. Leverage converts a temporary decline into a permanent loss whenever it forces a sale, and the whole question is whether you can hold through the bad part.

What this is actually evidence for

Not that borrowing to invest never works. It is evidence for a narrower and more useful claim: a plan that only works at today’s rate is not a plan.

The tests worth applying before starting:

• Does the spread survive a 3-point rise in your borrowing rate? That is not a stress test — it is smaller than what actually happened between 2022 and 2023.
• Can you service the loan from employment income, with the distribution stopped entirely?
• Does the loan force a sale if the portfolio falls 30%?
• Is the interest deductible? In a non-registered account, generally yes. For a TFSA or RRSP, no — which removes a large part of the advantage before rates enter the picture at all.

The loan spread calculator runs both sides at a rate you choose, deductibility included. Set it to 7.20% and see whether the plan still stands up.

Rate figures transcribed from the Bank of Canada’s published series.