The Signal Beneath the Noise: What Equifax’s Regional Fraud Data Tells Us About Mortgage Risk
A national average can hide more than it reveals. That is the lesson sitting inside Equifax Canada’s latest fraud and credit data for the second quarter of 2026. On the surface, the headline number looks encouraging: mortgage application fraud across the country fell to 0.20%. Dig one layer deeper, by province, and a very different story emerges, one that should matter to anyone building a model, a dashboard, or a lending policy on top of national figures alone.
Ontario’s mortgage fraud rate, while down from 0.40% a year earlier, still sits at 0.28%, the highest in the country. Alberta follows at 0.19%, improved from 0.24%. Quebec eased to 0.18%. These are not dramatic swings in isolation, but when layered against delinquency data, the pattern sharpens considerably. Ontario mortgage balances at least 90 days past due reached roughly 0.40% in July, nearly double the 0.21% recorded across the rest of Canada. That is the kind of divergence that a single national metric will always smooth over and quietly misrepresent.
What makes the Ontario data especially interesting from a signal standpoint is where the fraud is concentrated. More than 80% of flagged fraudulent mortgage applications in the province came from applicants with Equifax Risk Score 2 values above 700, meaning prime and super-prime borrowers, not the higher risk segment a simple risk model might flag first. Add in the age distribution, with borrowers aged 36 to 45 accounting for roughly 36% of fraudulent applications, and a clearer behavioural pattern starts to take shape: financially established households under renewal pressure, misrepresenting income or employment to manage payment shock, rather than opportunistic fraud from applicants with weak credit histories to begin with.
Alberta adds a second, distinct signal worth tracking: migration-linked credit risk. Equifax found that more than 70% of people under 35 who relocated to Alberta from another province last year came from Ontario and British Columbia, and that this group carries a 30-plus-day delinquency rate of 7.24%, well above the province’s general rate of 5.64%. That is not a fraud signal in the traditional sense. It is a geographic risk signal, one that only becomes visible when migration data, age cohorts, and delinquency tracking are cross-referenced rather than viewed separately.
Any national underwriting policies that fail to account for those provincial differences are going to lead to some very real challenges and increasing losses in this space.
That quote, from Carl Davies, Equifax Canada’s head of fraud, identity and compliance, is really a data governance argument as much as a fraud warning. With $1.97 trillion in outstanding mortgage balances nationally, a rate that looks reassuring in aggregate can still mask meaningful pockets of elevated risk once you segment by province, borrower age, credit tier, and migration pattern. For anyone working with property or credit data, this is the recurring lesson: the aggregate tells you what happened, but the segmented view tells you why, and where to look next.
As renewal cycles continue to push borrowers from pandemic era low rates into today’s higher payments, the provinces showing stress now, Ontario and Alberta in particular, are worth watching closely in the datasets that follow. Source: Canadian Mortgage Trends.
