Housing's affordability strain excludes buyers before it forces owners to sell
The affordability measure has stayed above 40% longer than it did before 2008, yet few owners are underwater, foreclosure inventory is low, and forced supply remains scarce.
See the accompanying note for definitions, calculations and data provenance.
Part 1 showed why narrower mortgage spreads failed to restore affordability. It left open a second path: a credit event that lowers prices and rates together. This article asks whether that channel is active.
The affordability reading looks familiar. The Atlanta Fed considers a median home affordable when its annual cost stays below 30% of median income. Yet its Home Ownership Affordability Monitor has now exceeded 40% for 37 consecutive months, 15 months longer than during the 2000s housing bubble. The current run also reached higher: 45.3% in October 2023 against 43.9% in July 2006.
The owner-credit picture looks nothing like the last cycle. After 2006, falling prices drove owners underwater, defaults rose, and foreclosures pushed distressed homes onto the market. Today, home prices have not collapsed, foreclosures remain scarce, and forced sales have not overwhelmed the market.
The two measures describe different households. Affordability describes the cost of entering the market. Credit data describe owners already inside. For anyone judging crash risk, that distinction changes the signal that matters: delinquencies and forced listings, not the affordability ratio, measure owner distress.
Owner distress turned decline into collapse
The last housing crash had a mechanism. Falling prices pushed borrowers into negative equity, defaults became foreclosures, and forced sales drove prices lower again. That feedback loop, the housing credit channel, turned a decline into a collapse.
Loan-level research explains the first step. Elul, Souleles, and their colleagues followed mortgages originated in 2005 and 2006 through the crisis. Their 2010 study in the American Economic Review found two default triggers with comparably sized effects. One was negative equity. The other was illiquidity: a borrower could no longer meet the payment from current income.
The test follows from those triggers. Are they present today, and are they creating forced supply?
Today's owners show little distress
Start with equity. CoreLogic counted 1.2 million mortgaged homes underwater in the fourth quarter of 2025, or 2.2% of the total. Roughly 11 million homes were underwater at the 2009 peak, and later CoreLogic analysis put the share at 26%. The current share has edged above its trough, but it remains roughly one-twelfth of its crisis peak.
Loans already failing tell the same story. The Mortgage Bankers Association reported that 4.44% of mortgages were at least one payment past due in the first quarter of 2026, up from a year earlier but far below the 10.06% crisis peak. That measure excludes loans already in foreclosure. Foreclosure inventory stood at 0.64%, against 4.63% at the 2010 peak.
A Federal Reserve series for single-family mortgages held by commercial banks reads lower still: 1.89% in the first quarter of 2026 against an 11.48% crisis peak. That is not a conflicting signal. The series covers a narrower, prime-skewed loan book and includes loans in nonaccrual status. Its lower reading reflects coverage as originations moved toward agency securitization and nonbank servicing.
Aggregate payment pressure also remains low. Required mortgage payments absorbed 5.88% of aggregate disposable income in the first quarter of 2026, at the 28th percentile of the series and far below the 8.95% crisis peak. The 2021 refinancing trough was lower still, at 4.76%. This measure can hide stress among particular borrowers, but it shows no crisis-scale burden in the aggregate.
This resilience has several sources. After 2008, the ability-to-repay and qualified-mortgage rules added by Dodd-Frank required lenders to document income and assets. Years of price gains built equity, while fixed-rate loans insulated existing borrowers from later rate increases. Tighter underwriting is part of the explanation, not the whole of it.
Owners have not entered distress at scale, so the first half of the credit channel is missing.
Forced supply has not arrived
Forced sellers would leave a visible mark in housing inventory. That mark is absent.
Months of supply measures how long current listings would take to sell at the current sales pace. The National Association of Realtors reported 4.6 months of existing-home supply in June 2026, against a peak above 11 months in 2008. The market has less than half the supply that accompanied the last crash.
Raw listings support the same conclusion. The market held 1.56 million existing homes for sale in June 2026, compared with about 4 million at the 2007 peak. Even that comparison understates the difference because the US housing stock has grown roughly 16% since then.
Vacancy offers a final check. The homeowner vacancy rate stood at 1.2% in the second quarter of 2026, below its 1.5% long-run median, though above the 0.7% series low reached in 2023.
Housing has an affordability problem. It does not have a forced-supply overhang.
The burden falls on entrants
If forced sellers are scarce, where does the affordability burden land? The mix of buyers reveals who remains able to enter.
The market favors buyers least exposed to mortgage rates. Redfin found that 32.6% of 2024 purchases across 40 large metros were all-cash. At the same time, the National Association of Realtors found that first-time buyers accounted for only 21% of purchases in the year to mid-2025, the lowest share in a series that began in 1981 and roughly half the pre-2008 norm near 40%.
Existing owners face a different calculation. In the first quarter of 2026, the Federal Housing Finance Agency found that 49.9% of outstanding fixed-rate mortgages carried rates below 4%. The Freddie Mac offer rate stood at 6.66% in late July. Selling means surrendering a cheap mortgage and borrowing again at a much higher rate. The gap suppresses listings: owners keep their cheap loans, and homes that could relieve the shortage remain unavailable.
That divide changes what the affordability measure means. It measures the cost of entering the market more directly than the solvency of owners already inside. In current data, the strain registers as exclusion, not distress.
Prices can fall without owner distress
In a demand-driven decline, transaction volume contracts, owners who do not need to sell withdraw, and the few transactions that clear set lower prices. That path needs no foreclosure wave. It can produce lower prices without producing a flood of forced supply.
The 2008 collapse followed a different path. Distress added inventory, lower prices created more negative equity, and each stage strengthened the next. Current data show that the credit channel is inactive, not impossible. A large price decline could erode equity, while a labor shock could raise delinquency.
Housing can correct, but owner distress is not yet large enough to activate the credit channel.
Income loss is the trigger to watch
One trigger can change faster than the others: income. Equity, underwriting, and mortgage lock-in usually change over years. Borrower income can change within quarters.
A borrower who loses income can miss payments even with positive equity, although that equity can allow a sale before foreclosure. A broad income shock can therefore move through the mortgage book much faster than equity or underwriting can change. That risk lives in the labor market.
No broad income shock is visible yet. The Bureau of Labor Statistics reported 4.2% unemployment in June 2026, at the 23rd percentile of its postwar history and far below the 10% peak in October 2009. Yet payrolls grew by only 57,000, and labor-force participation fell to 61.5%. The headline rate does not justify complacency.
During the last cycle, unemployment peaked in October 2009 and foreclosure inventory continued rising into the first quarter of 2010. One cycle does not establish a reliable lead. It does identify the sequence to watch: labor weakens, delinquencies rise, and foreclosure inventory follows.
Equity helps keep the credit channel dormant. Employment could activate it much faster.
Closing the gap still requires a large move
Even without a credit event, affordability can improve through prices, incomes, or both. The latest complete annual data, from 2024, show how far they must move. Returning the price-to-income ratio to its 2012 post-crash trough would require either a 28% fall in home prices or a 39% rise in incomes, holding the other fixed.
The arithmetic does not depend on the channel. A 28% price decline remains a 28% decline whether weak demand or forced sales delivers it. The path determines what happens around that decline: owners can withdraw from a quiet market, or distressed borrowers can add homes to a falling one.
The credit channel is dormant, not closed
Equity remains widespread, post-2008 underwriting contributes to borrower resilience, and mortgage lock-in restrains supply. None of those buffers is permanent. A large price decline would consume equity, while a labor shock could impair borrowers within quarters.
This analysis does not predict whether prices fall or whether the credit channel reactivates. It identifies the sequence that would change the diagnosis: equity erodes, income losses spread, delinquencies rise, and forced supply follows.
Today's affordability strain is closing the door on buyers, not breaking the mortgage book. Prices can still fall. What is missing, for now, is the forced-sale loop that turns a decline into a collapse.