2Y 4.610 ▼5Y 4.780 ▼10Y 4.980 ▼30Y 5.320 ▼VIX 17.20 ▲MOVE 93.54 ▼DXY 101.78 ▼USD/JPY 157.42 ▼GOLD 4,003 ▼COPPER 6.55 ▼WTI 103.28 ▲BTC 78,069 ▼ETH 2,238 ▼S&P 7,720 ▼NDX 29,403 ▼MSTR 141.62 ▼
Rates · Liquidity · Scarcity · Credit · Power · 2026-10-29

Housing Is Adjusting Through Volume, Not Defaults

The usual way a housing cycle reaches the real economy is through defaults: payments stretch, arrears rise, lenders tighten, and the damage shows up in bank earnings. That is not what is happening. The 30-year fixed mortgage rate sits at 7.40%, roughly four and three-quarter percentage points above its 2021 low, and housing starts are 7.4% below where they were a year ago. Yet consumer credit outstanding has reached a record $5.20tn and the bank card delinquency rate has fallen to 2.85% from 3.04%. Prices are high, activity is falling, and credit quality is improving. Those three facts look contradictory until you see that the market is clearing by volume rather than by price — and that the people who cannot transact are simply absent from the credit data. This is the transmission channel that matters for the midterms, and it is the one that never shows up as a crisis.

The four numbers

The 30-year fixed mortgage rate was 7.40% on 15 October, unchanged on the week. It peaked at 7.79% in October 2023 and bottomed at 2.65% in January 2021, so the current level is close to the top of the range and nowhere near the bottom.

Housing starts ran at 1,221,000 annualised in September, down 7.4% from 1,319,000 a year earlier and about a third below the 1,807,000 peak of April 2022.

Consumer credit outstanding reached $5.20tn in August, a record for the series and 2.7% higher than a year earlier.

Bank card delinquency was 2.85% in the second quarter, down from 3.04% a year ago and below the 3.22% peak of 2024.

Why the defaults are not coming

The American mortgage market was rebuilt after 2008 around the fixed-rate, long-dated, fully amortising loan. Most outstanding mortgages were written when rates were far lower, which means the household sector's interest burden barely moves when the policy rate does. A borrower on a 3% mortgage is not exposed to a 7.40% market rate; they are exposed to nothing at all until they move.

That is the lock-in. It protects the balance sheet and paralyse the transaction market at the same time. Existing owners do not list, because selling means trading a cheap mortgage for an expensive one. First-time buyers face the expensive one with no equity to roll. Volume collapses, prices stay sticky because supply collapses with demand, and the credit data stays clean because the only people completing transactions are the ones who can comfortably afford them.

This is why delinquency is falling while affordability is at its worst in a generation. It is a composition effect, not an improvement in circumstances: the pool of borrowers shrank to the strongest names.

The credit data is a survivor's sample

A record $5.20tn of consumer credit growing 2.7% year on year, against card delinquency falling to 2.85%, reads like a healthy consumer. Read it as a denominator instead. The stock keeps rising because the price level keeps rising and because borrowing is how the excluded participate — the renter putting a deposit on a card, the buyer financing a used car instead of a house.

What the series cannot show is the household that never applied: the one that looked at a 7.40% payment, did the arithmetic, and stayed put. That household is invisible in delinquency, invisible in credit outstanding, and fully present in sentiment.

The University of Michigan index was 51.7 in August, down 11.2% on the year and within reach of its May 2026 low of 44.8. That is not a credit signal. It is the signal of people who have been priced out of the largest purchase of their lives and have stopped expecting that to change.

Why this is the political channel

A credit event is legible. Losses appear, banks provision, a facility gets announced, and the political system has something to respond to. An access event is diffuse: no single loss, no single villain, just a cohort of renters and young buyers whose entry point drifted out of reach while everyone told them the economy was strong.

That cohort votes on cost of living, and it does not experience the official statistics as true. When a central bank says inflation is near target and a 30-year mortgage costs 7.40%, both statements can be correct and only one is believed. This is the mechanism by which a strong labour market and a politically angry electorate coexist.

30-year fixed mortgage rate, % (weekly)2.503.885.256.628.007.402021202320252026Bank card delinquency rate, % (quarterly)1.401.902.402.903.402.8520192021202320252026

What would falsify this

The claim is not that housing is fine, and not that a crisis is coming. It is that the adjustment is running through quantity and access rather than through defaults, and that this keeps the damage invisible to credit metrics while concentrating it politically.

Falsifier one: delinquency turns up while starts keep falling. That would mean the composition effect has run its course and the marginal borrower is now missing payments rather than merely absent. Watch card and auto delinquencies together — card alone can rise with borrowing for necessities.

Falsifier two: the mortgage rate falls below roughly 6.5% and starts do not recover within two quarters. That would prove the binding constraint is price level and income, not the cost of money, and would move this from a rate story to a valuation story.

Until one of those happens, the correct read is a market clearing by exclusion, which is slower, quieter, and more politically expensive than a default cycle.

CALL & INVALIDATION
CallUS housing is clearing by quantity, not by price or default: a 7.40% mortgage has cut starts 7.4% year on year while record consumer credit and falling card delinquency show a survivor's sample, not a healthy consumer. The damage is concentrated in access and therefore in votes, not in bank loss rates.
EvidenceFRED: MORTGAGE30US 7.40% (15 Oct 2026), peak 7.79% (Oct 2023), low 2.65% (Jan 2021); HOUST 1,221k SAAR (Sep 2026), -7.4% y/y, peak 1,807k (Apr 2022); TOTALSL $5.20tn (Aug 2026), +2.7% y/y, series high; DRCCLACBS 2.85% (Q2 2026) vs 3.04% a year earlier, peak 3.22% (2024); UMCSENT 51.7 (Aug 2026), -11.2% y/y, low 44.8 (May 2026). Delinquency is a quarterly series and lags; do not read it as September data. Macro context: US 10yr 5.15%, 2s10s 38.0bp on the 19 October par close.
InvalidationCard and auto delinquency rising together while housing starts keep falling — the marginal borrower is defaulting, not just absent. Secondary: mortgage rate below 6.5% with no starts recovery within two quarters, which makes this a price-level story rather than a rate story.
Levels2Y 4.77 / 10Y 5.15 / 30Y 5.48 / 2s10s 38.0bp
RELATED
Credit Dispersion: The Average Is Lying to You — same method — the aggregate looks calm because the weak borrowers left the pool
Capex Rose 110%. Profit Fell 56%. That Is the Whole AI Cycle. — the other half of the same week: capital spending without the revenue to service it
DISCLOSURE — NOT INVESTMENT ADVICE
NoticeMacroTape publishes macroeconomic analysis and market commentary for information and education only. Nothing published here is financial, investment, legal or tax advice, or a personal recommendation, and no author is a registered investment adviser. Calls, Levels and invalidation points are analytical views about how a market mechanism works, not instructions to trade. Readers act on their own responsibility and carry their own losses.