What "sales up" actually means

The +5.5% national sales print is real, but it is a level number, not a trend number.

Actual (not seasonally adjusted) monthly activity in May 2026 was still 5.1% below May 2025.

So the strongest month in two years is still a smaller market than the same month a year ago.

Read the seasonally adjusted number and the market is recovering. Read the year-over-year number and the market is still smaller.

The sales-to-new-listings ratio in May tightened to 49.2% (up from 46.2% in April). The long-term average is 54.8%.

A ratio between 45% and 65% is generally considered balanced market territory. So 49.2% is mid-balanced, not "hot." Months of inventory fell to 4.8 from 5.1 — still close to the long-term average of 5.0, and nowhere near a seller's market reading (under 3.6 months).

In other words: the market is moving. It is not on fire. The headline told you the first half. The footnote told you the second.

What "$702,079" actually means

The national average sale price is the number most readers see, and it is the number most readers get wrong.

The $702,079 figure is not the price of a typical home.

It is the average of every home that sold through MLS® in May — which means it is heavily influenced by the mix of what sold, not just by the price of any individual home.

This is where the contradiction sharpens.

While the actual average broke $700K, the MLS® Home Price Index (HPI) — a benchmark that controls for the type and size of home sold — inched down 0.1% month-over-month in May.

The HPI was down 4.1% year-over-year, the smallest YoY decline so far in 2026 but still a decline.

Regionally, prices remain down year-over-year in British Columbia, Alberta, and Ontario — the three most expensive provinces — while gains in other provinces are pulling the headline average up.

In plain terms: the average price is up because the composition of what sold shifted.

More higher-end Ontario sales returned in May after a slow April (CREA's own read: "the HST rebate on new builds may have only briefly drawn the attention of buyers away from the existing home market").

The actual benchmark price of a typical home is still slightly softer than it was a year ago.

Both can be true.

The headline reads the first. The forecast reads the second.

Why CREA cut the 2026 forecast anyway

This is the move that does not compute on first glance. If May is the strongest month in two years and the average is back above $700K, why would the trade body cut its 2026 forecast in April?

Because the forecast is built on the HPI, not the average.

And the HPI in April 2026 was still showing a year-over-year decline steeper than the long-run norm.

The April downgrade was CREA reading the benchmark data and concluding: yes, the year-over-year decline is shrinking, but the recovery to flat is going to take longer than the January forecast assumed.

The May print — with the +5.5% sales spike, the +1.5% YoY average price rise, and the smallest YoY HPI decline of 2026 — is consistent with the new (lower) forecast.

It is not a reason to revise the forecast back up yet, because the May print is one month, and one month does not make a year.

This is also why CREA's own economist language was calibrated, not triumphant.

Cathcart: "While it was just the first month in 2026 to see any meaningful upward momentum in headline demand, under the surface conditions have been improving for some time." First month.

Under the surface.

Note the hedge.

The April downgrade is the institutional view.

The May print is the early evidence that the institutional view is, perhaps, too pessimistic — but one print does not flip an annual forecast.

What this means for a mortgage decision in 2026

If you are renewing or buying right now, the headline/forecast gap is actionable.

If you were waiting for "the crash"

It is not coming in the form most people expected.

The HPI is still slightly down YoY but the smallest decline of the year.

The average is up because the mix is up.

A 25% drop requires a credit shock the data does not currently show.

The 2.25% BoC policy rate — held at the June 12 decision for the fifth consecutive meeting — is the closest thing to a soft landing the Bank has engineered in 25 years. If you are still anchoring on a 2022-style correction, the data has moved past you.

If you were waiting for "the recovery"

It is also not here in the form most people expected.

The HPI is still down year-over-year.

Months of inventory is still in balanced-market territory, not seller-market territory.

The forecast was cut, not raised.

A 5% YoY rebound requires the rate cuts the bond market is pricing in to actually arrive.

They have not.

The Bank of Canada is on hold. A first cut is now a Q3-or-later story, not a Q2 story.

The live specials are already telling you what to do with that gap. As of late June 2026, the best publicly-quoted 5-year fixed is sitting around 4.19% (Coast Capital, standard uninsured), the best 5-year variable is at 3.75% (Neo Financial, insurable 75.01–80% LTV), and insured 5-year fixed purchases are quoted around 3.94% (Prospera, standard).

Those are the specials you can actually lock today, not the specials you hope to lock if the BoC cuts twice by year-end.

With the BoC on hold and the 5-year Government of Canada bond yield drifting lower through June, lenders have been quietly widening the spread between their specials and the contract rate, not narrowing it. That is a brief window.

The math for a typical $700K mortgage right now

• At today's best 5-year fixed of 4.19% (Coast Capital, 25-year amortization): roughly $3,725/month
• At today's best 5-year variable of 3.75% (Neo Financial, insured 75.01–80% LTV, prime − 0.85% in the current spread): roughly $3,422/month
• A 44 bps spread, on a $700K balance, is roughly $300/month in cash flow over the next 12 months, before any BoC move

That is not a number to anchor a 5-year decision on. It is a number large enough to make the next 60 days worth a serious conversation with a broker.

The bigger story: the headline and the forecast have already split

The most important thing the May CREA print tells you is not that the market is recovering.

It is that the institution publishing the data is no longer fully aligned with the headline the data generates. The HPI-based forecast says "still soft, just less soft." The actual-average-based headline says "up, strongest in two years." Both are derived from the same MLS® system. They tell different stories because they are measuring different things.

For anyone making a mortgage, renewal, or purchase decision in summer 2026, the right read is: the headline is real but small. The forecast is institutional and pessimistic.

The spread between today's 5-year fixed specials and the variable alternatives is wider than it has been in two years. And the next BoC decision is July 9, 2026.

If you are still waiting for a single, clean signal that the bottom is in, the May print was not it. If you are still waiting for the crash the data said was coming in 2024, the May print was not that either.

It was the print that made the gap between the two views — for the first time in this cycle — wide enough to be the story.

---

The pragmatic take

A locked 4.19% five-year fixed in June 2026, with the BoC on hold and the variable alternative at prime − 85 bps, is a defensible bet that the recovery will be slow and uneven.

A variable at prime − 85 bps is a defensible bet that the BoC cuts in the second half.

The two bets are different.

The right one depends on how you read the gap between the headline and the forecast — and now, finally, there is data on both sides of that question.

---

Sources: CREA National Statistics, June 16 2026 (https://stats.crea.ca); CREA, "CREA Downgrades Resale Housing Market Forecast," April 16 2026 (https://www.crea.ca); Bank of Canada policy rate decisions, January–June 2026 (https://www.bankofcanada.ca); Coast Capital, Neo Financial, Prospera posted public rates as of June 23 2026 via Pragmatic Mortgage Lending's Rate Explorer (https://pragmatic.mortgage/rates).