Foreclosures, Resets and the 2026 Rate Path: What the Data Actually Shows

Foreclosure activity in most developed mortgage markets spent the first half of this decade at levels well below any historical norm, and it did so for three specific reasons rather than because households had become unusually prudent.

The first was policy. Pandemic era forbearance programmes formally paused payment obligations for millions of borrowers, and servicers were subject to moratoria on filing. Loans that would ordinarily have progressed through delinquency into filing simply did not, which pushed a cohort of stress out of the data entirely.

The second was underwriting. The rules written after 2008 removed the loan products that produce rapid default. Verified income became mandatory, negative amortisation effectively disappeared from mainstream lending, and borrowers were assessed against a stress tested qualifying rate rather than the introductory one. The resulting loan book is genuinely more resilient than its predecessor, and that is a durable improvement rather than a cyclical one.

The third was the rate lock. An extraordinary share of outstanding mortgages was written or refinanced during the low rate window, at fixed rates far below current market levels. Those borrowers have no reset to face and every incentive to stay where they are, which both suppresses arrears and freezes transaction volumes.

Against that base, a large percentage increase is arithmetically easy to produce and analytically almost meaningless in isolation. The correct question is never how fast filings are rising. It is where the level sits relative to the long run average, and which cohorts the increase is concentrated in.

The reference frame most readers carry is 2008, and it is the wrong instrument for reading 2026.

The 2008 crisis was a product failure before it was a price event. The loans at its centre were constructed so that they could only perform if house prices continued to rise. Payments were designed to increase sharply after an introductory period, borrowers were qualified against the introductory payment rather than the eventual one, income was frequently not verified, and the securitisation chain distributed the consequence far enough from the originator that nobody in it had a reason to object. When prices stopped rising the refinancing escape route closed and the loans did what they were built to do.

None of those features describe the dominant loan today. A fully amortising, income verified mortgage stress tested at a higher qualifying rate behaves differently under pressure. It does not blow up. It grinds. The household reduces spending, defers maintenance, extends amortisation where the lender permits it, delays moving, and in many cases delays forming a second household at all. That is a real economic cost and it shows up in consumption data long before it shows up in credit data, but it is not a default wave.

The practical implication is that anyone waiting for a 2008 style signal to confirm that housing stress exists will conclude, wrongly, that there is no stress. The signal is in a different series.

National averages obscure the picture because the risk is not distributed evenly. It sits in three identifiable cohorts, and a reader can place themselves or a market in or out of them with reasonable confidence.

The first cohort is recent high loan to value purchasers. A household that bought near the price peak with a minimal deposit carries the highest monthly payment in the market and holds the smallest equity buffer. Equity matters more than most coverage acknowledges, because it determines whether a household under pressure has an exit. A borrower with meaningful equity who cannot sustain the payment can sell, repay the loan and preserve their credit. A borrower with none cannot, and the only remaining paths are modification or the filing sequence.

The second cohort is borrowers reaching reset. This includes variable rate holders whose payment adjusts directly and short term fixed holders whose term expires and must be renewed at whatever the prevailing rate is. The magnitude of the shock is simply the gap between the rate they hold and the rate now available, applied to their outstanding balance. Where that gap is three or four percentage points, monthly payments can rise by forty per cent or more without anything happening to the household's income. The calculator below applies that arithmetic to figures the reader supplies, including the resulting debt service ratio.

The third cohort is small scale investor landlords. Their exposure differs in kind because their income is rental rather than employment, their financing is often shorter dated and repriced more frequently, and they sit outside most owner occupier hardship protections. Where rents have not kept pace with financing costs, the arithmetic of holding the property stops working, and their response is to sell rather than to default. That is a supply effect rather than a credit event, which is why investor heavy markets show inventory increases before they show arrears increases.

The mechanism connecting a central bank decision to a foreclosure filing is long, sequential and remarkably consistent across cycles, and understanding it removes most of the confusion in the public conversation.

A household facing a higher payment does not miss it immediately. It reduces discretionary spending first. Then it draws down savings. Then it substitutes revolving credit for cash flow, which is why credit card balances and utilisation rates are a leading indicator for mortgage arrears rather than an unrelated series. Only when those buffers are exhausted does a mortgage payment go unpaid, and that sequence typically consumes two to four quarters.

From the first missed payment the process becomes legal rather than economic. Delinquency is reported to credit bureaus in the second month. The ninety day mark, in the third, is when the loan enters serious delinquency statistics and becomes visible in published data. A loss mitigation window follows, in which modification or forbearance is decided. Only where that fails does a notice of default issue, and only after that does a filing occur. Depending on whether the jurisdiction requires judicial foreclosure, the interval from first missed payment to filing runs eight to eighteen months.

Adding the two stages together gives a total transmission lag of roughly twelve to twenty four months. This has a consequence that is easy to state and frequently ignored. Arrears data published this quarter is a report on interest rate decisions taken one to two years ago. The 2026 rate path has already determined a substantial part of the 2027 arrears picture, and no policy change made after the fact can retrieve the households already inside the sequence.

Four observable series would change the reading, and none of them require a forecast.

The first is the ninety day delinquency rate rather than the filing count, because filings are a lagging legal artefact while serious delinquency is the earliest reliable credit signal. The second is credit card utilisation among mortgage holders, which reveals buffer exhaustion before the mortgage itself is missed. The third is the volume of loan modifications and term extensions, which measures how much stress is being absorbed rather than resolved and which suppresses reported arrears while the underlying pressure persists. The fourth is metropolitan level inventory in investor heavy markets, which shows landlord exit before it shows credit distress.

A material and sustained rise in the first two, occurring together, would indicate that the cohort pressure described above is broadening into something more general. Their absence, alongside rising filings, confirms the base effect reading.

The defensible position is narrower than either of the loud ones.

There is no national foreclosure crisis in the 2008 sense, and the loan construction that made that crisis possible has largely been legislated out of the market. There is genuine, concentrated and worsening pressure inside three identifiable cohorts, and that pressure is being expressed as reduced consumption, delayed household formation and extended amortisation rather than as mass default. Because the transmission lag runs one to two years, the pressure already in the pipeline will continue to surface in the data regardless of the next rate decision.

For an individual household the useful action is not to track the national statistic. It is to calculate the personal one: the debt service ratio after the next reset, and whether equity exists to provide an exit if that ratio proves unmanageable. The model in this piece produces both figures from the reader's own numbers. Readers examining the financing side of the same rate structure should also read the accompanying reference on how leverage converted an accurate market thesis into a forced liquidation this summer.