Real reports.
Clear limitations.
A useful comparison tells you where its numbers come from — and when they aren’t enough.
What counts as a real rate?
A borrower may report a verbal offer, a written quote, a rate hold, an accepted agreement, or a funded mortgage. We label each stage. Our outcome comparisons use accepted and funded reports; offer comparisons use quotes and rate holds separately.
Reports describe what contributors say they received. “Source reviewed” means a reviewer checked the supporting text. “Document reviewed” means relevant document fields were checked. Neither means a bank independently confirmed the mortgage.
Comparing like with like
We keep rate type, term, transaction purpose, province, occupancy, mortgage structure and insurance status separate. Balance, loan-to-value and remaining amortization help refine the match. If we broaden these latter details to find a useful sample, we tell you.
Unknown insurance stays unknown. A form default does not become evidence until the contributor confirms it.
Dates matter
Current comparisons normally use reports agreed within 30 days. Newer reports receive more weight, with a 14-day half-life. If only older evidence is available, we may show up to 60 days as historical context. A mortgage from last year never becomes a current report because someone submits it today.
What the range means
The range covers the middle 50% of eligible, weighted reports. It is not a guaranteed offer or a 95% confidence interval. We count each mortgage journey once, rather than treating every counteroffer as a different borrower. We exclude your own journeys when checking your offer.
- Insufficient: fewer than five comparable episodes, missing critical details, or stale evidence. No personalized numeric estimate.
- Limited: a small, incomplete, older, or concentrated sample. Context only; we do not grade your offer.
- Moderate: at least 12 episodes and an effective sample of eight, with fresh evidence and the essential profile details known.
- Strong: at least 30 episodes and an effective sample of 20, with sufficient recent reports, diverse sources, and reviewed supporting evidence.
These are product thresholds under evaluation. A large sample does not remove self-selection, missing information, or the possibility of misleading reports.
Variable rates need extra care
Lenders can use different prime benchmarks. We only recalculate a spread when the lender’s benchmark is known and recently checked. A recalculated equivalent is not a currently available offer.
What makes an offer worth accepting?
Rate is one part of the decision. Compare fees, prepayment privileges, cash-back conditions, qualification, switching time and your priorities. Our preparation tools help you ask informed questions; they cannot approve financing or choose a mortgage for you.