Photo by Md Ishak Rahman on Unsplash
- As of September 27, 2026, coverage surfaced through Google News points to a Bloomberg report tying a renewed 7% handle on the 30-year fixed to a tightening mortgage lock-in effect — the weekly benchmark itself should be confirmed against Freddie Mac's Primary Mortgage Market Survey.
- FHFA research previously estimated the lock-in effect prevented roughly 1.3 million home sales between mid-2022 and the end of 2023. Spread across those roughly 18 months, that is on the order of 72,000 transactions a month that simply did not happen.
- The load-bearing stat is FHFA's finding that each percentage point a mortgage sits below market cuts the probability of sale by roughly 18%. Almost every headline quotes it; almost none stress-test it.
- Lock-in is not one national condition. It is a distribution — and the households it traps hardest are not the ones most commentary describes.
The Evidence
What if the lock-in effect is not really a rate story?
That sounds absurd on September 27, 2026, with the 30-year fixed once again pressing against the 7% threshold — the level Bloomberg, in reporting distributed via Google News, used as the frame for its latest look at frozen housing inventory. Rates first, headlines second: that is the right order. But the rate is the trigger, not the mechanism. The mechanism is a spread, and spreads behave differently from levels.
Here is the signal, stated plainly. The 30-year fixed has repeatedly crossed 7% across the 2023–2025 period, up sharply from the sub-3% lows of 2020 and 2021. A very large majority of outstanding U.S. mortgages carry rates well below today's market — a substantial share below 5%, and roughly 60%+ sat below 4% during the peak lock-in window. Those owners hold something close to a financial asset in their loan, and selling destroys it. Existing-home sales have hovered near multi-decade lows for most of the high-rate era, and constrained resale inventory has kept prices sticky even as affordability deteriorated. Readers should verify the current weekly 30-year figure at Freddie Mac's PMMS, which remains the authoritative benchmark rather than any single news citation.
None of that is new. What deserves more scrutiny is the number everyone repeats next.
The 18% Figure Is Doing All the Work
FHFA's estimate — each percentage point of below-market rate reduces sale probability by about 18% — is the quiet engine under nearly every lock-in article published in the last three years. It is worth asking what happens when you actually run it forward.
Take a household holding a 3% mortgage in a 7% market. That is four percentage points of lock. Applied linearly, 4 × 18% = 72% — a sale probability cut by roughly three quarters versus an equivalent household at market rate. Do the same for a 5% borrower and the gap is two points, or about a 36% reduction. Same market, same 7% rate, radically different behavior.
Chart: A straight-line extension of FHFA's roughly 18%-per-point finding. Illustrative only — the underlying relationship is unlikely to be perfectly linear.
And that is exactly where a careful skeptic should push back. FHFA's estimate is a marginal effect, not a formula meant to be stacked four times and sold as a forecast. Real behavior almost certainly flattens out: past some threshold, a locked-in owner is simply not selling, and the fifth point of lock cannot subtract much more probability than the fourth already did. The honest reading is directional. The gap between a 3% borrower and a 5% borrower is not a rounding difference — it is the difference between a household that might move and one that will not — but the precise percentages should be treated as a slope, not a schedule.
That distinction changes the story. Lock-in is usually described as something that happens to "homeowners." It does not. It happens to a specific cohort — the 2020–2021 refinance and purchase class sitting in the roughly 2.5% to 4% band — while a 2018 buyer at 4.75% and a 2023 buyer at 6.5% are barely constrained at all. When commentary treats the housing market as uniformly frozen, it is averaging together groups that face completely different math.
What It Means: Read the Cohort, Not the Headline
The second-order consequence is the one that actually moves money: lock-in transfers pricing power from resale sellers to builders.
Because existing inventory is suppressed, new construction has gained share simply by showing up — and builders can do something an individual seller structurally cannot. A homebuilder with a captive lending arm can fund a permanent rate buydown out of margin and advertise a payment, not a price. A family selling a 1994 colonial has no such lever. So the competitive question in many submarkets is not "which house is nicer" but "which seller can subsidize the borrower's rate." Under a 7% market rate, the builder wins that comparison nearly every time. Under a materially lower rate, the buydown loses its power and the resale seller's location advantage reasserts itself.
Geography matters here, and it is worth being precise about what the data does and does not support. The FHFA work quantifies lock-in nationally; it does not decompose neatly into metro-level scoreboards. What can be said is structural: in Sun Belt submarkets where new construction is a meaningful share of active listings — think the builder-heavy edges of metros like Phoenix, Atlanta and Austin — the buydown channel gives buyers a genuine alternative, so days on market for tired resale listings stretches out and the price-per-sqft delta between new and existing narrows. In land-constrained Northeast and coastal California submarkets, there is no builder pressure valve. Lock-in there does not redistribute demand; it just removes supply, and prices stay stickier for longer. Same national statistic, opposite submarket reality.
There is also a cost that does not show up in rate charts at all. Even an owner who never sells watches carrying costs drift upward through taxes and insurance — the mechanism Smart Insurance AI walked through on escrow-driven payment increases — which quietly erodes the very payment advantage lock-in is meant to protect. A 3% note is not a frozen monthly cost. It is a frozen principal-and-interest line inside a payment that keeps moving.
On the technology side, AI real estate tools are increasingly pointed at exactly this problem. Lenders and listing platforms now run automated valuation models (software that estimates a property's value from comparable sales and property characteristics, without a human appraiser visiting) alongside propensity models that try to predict which locked-in owners are closest to transacting anyway — typically because of a job change, a growing family, or retirement. That is the analysts' consensus on how lock-in unwinds: not in one thaw, but household by household as life events override rate math.
How to Act on This
Find your note rate, subtract it from the current PMMS 30-year figure, and that spread is your personal lock in percentage points. One point is friction. Four points is a wall. Home buying decisions made by someone at 2.75% and someone at 6.25% should not look remotely alike, and generic market commentary will not tell you which one you are.
A temporary 2-1 buydown (a subsidy that lowers your rate for the first two years, then steps up) is a very different product from a permanent rate buydown. In a 7% environment the incentive is real leverage, but only the permanent version changes your long-run payment. Compare the buydown-adjusted new-build payment against the resale listing's asking price — that side-by-side, not the sticker price, is the actual decision.
You cannot out-subsidize a national homebuilder's lending arm. You can out-locate it. Established schools, commute time, mature lot, finished basement — price to the submarket's realistic days-on-market, not to the 2021 comp, and check whether your loan is assumable (FHA and VA loans often are). An assumable low rate is the single asset in this market that a new build genuinely cannot replicate.
Bottom line: our read is that lock-in loosens on a slope, not a switch. Because the effect scales with the spread rather than the rate level, the first move that matters is not rates returning to 3% — it is the gap narrowing enough that the marginal 4.5% and 5% borrowers start transacting again, and they are a much larger group than the 2.5% holdouts who dominate the coverage. On balance, the more likely path is a gradual, cohort-by-cohort thaw that shows up in inventory before it shows up in prices — which means anyone watching for a dramatic housing market unfreeze is probably watching the wrong indicator. For property investment purposes, transaction volume is the tell.
Frequently Asked Questions
What is the mortgage lock-in effect, in plain English?
It describes homeowners who financed or refinanced at ultra-low pandemic-era rates — roughly 2.5% to 4% — and are financially discouraged from selling because any replacement mortgage would carry a far higher rate. Economists often call it a "golden handcuffs" situation: the house may no longer fit, but the loan is too good to give up. Note that the handcuffs are attached to the loan, not the property.
How many home sales has the lock-in effect actually prevented?
FHFA research estimated roughly 1.3 million home sales were prevented between mid-2022 and the end of 2023. That is a modeled counterfactual — an estimate of sales that would have occurred under normal conditions — not a direct count of cancelled transactions, so treat it as a magnitude rather than a precise tally.
Will the lock-in effect end when mortgage rates fall?
Housing analysts generally expect it to ease gradually rather than end. Two forces do the work: a narrowing spread between existing note rates and market rates, and life events — job relocations, new children, divorce, retirement — that eventually force moves regardless of rate math. Because the effect is tied to the size of the gap, partial rate declines unlock the shallowest-locked cohorts first.
What percentage of homeowners have a mortgage rate below the current market rate?
The vast majority of outstanding U.S. mortgages carry rates well below prevailing market levels, with a large share under 5% and roughly 60%+ below 4% during the peak lock-in period. Current composition shifts as new loans are originated and old ones pay off, so the figure should be checked against fresh agency data rather than assumed static.
How does a 7% mortgage rate affect home affordability for buyers today?
It works on two fronts at once, which is why it bites harder than the rate alone suggests. Directly, higher mortgage rates raise the monthly cost of a given loan amount. Indirectly, by freezing resale supply, lock-in keeps prices elevated even as fewer buyers can qualify — so buyers face both a costlier loan and less negotiating room. That combination is why existing-home sales have sat near multi-decade lows rather than simply repricing lower.
Disclaimer: This article is editorial commentary for informational purposes only and does not constitute financial or real estate advice. It reflects analysis of publicly reported data and does not represent independent testing or verification of any product or service. Rate figures change weekly; confirm current levels with Freddie Mac's PMMS before making any decision. Research based on publicly available sources current as of September 27, 2026.