US Housing & Mortgage Markets
TL;DR - The mortgage rate lock-in effect is measurably eroding: the share of outstanding loans over 6% has tripled since mid-2022, slowly restoring seller mobility - Erdmann's latest work reframes the affordability crisis as a $30 trillion accounting illusion — shortage-inflated land values look like wealth but are really a liability transfer from future buyers - California out-migrants are living proof of the affordability toll: those who left cut monthly housing costs by $700 and were 11 points more likely to become homeowners
The US housing market's defining tension in 2026 remains the collision between a slowly thawing supply picture and the deep structural distortions baked in by a decade of under-building. Two separate data releases this week illuminate opposite ends of that story — one granular and near-term, one sweeping and structural.
LOCK-IN EFFECT: THE ICE IS MELTING, SLOWLY
McBride's analysis of the FHFA's Q4 2025 National Mortgage Database puts hard numbers on a trend that's been building for 3 years. The share of outstanding fixed-rate mortgages carrying rates under 4% peaked at 65.1% in Q1 2022 and has now fallen to 50.6%. The sub-5% cohort similarly peaked at 85.6% and sits at 67.4% today — still a commanding majority of the outstanding stock, but trending the right direction.
The more telling metric is the other end of the distribution. Loans at 6% or above bottomed at just 7.3% in Q2 2022 and have now climbed to 21.9% — nearly tripling in 3 years. Each quarter that passes, more borrowers either originated at current rates or refinanced into them, diluting the pool of homeowners with powerful financial reasons to stay put.
The mechanism matters: this isn't rate cuts doing the work (we haven't seen dramatic relief on that front), it's time and turnover. New purchases, divorces, deaths, job relocations — the ordinary churn of life gradually dissolves the lock-in math. McBride has been consistent that this process unfolds over years, not quarters, and the data back him up. Inventory normalization is a slow grind, not a snap.
THE $30 TRILLION ILLUSION: SHORTAGE AS ACCOUNTING ERROR
Erdmann's new Mercatus Center paper offers the most clarifying reframe of the affordability crisis in recent memory, and a new California Policy Lab report gives it empirical teeth.
The core argument: roughly $30 trillion of US residential real estate value — half the total stock's ~$60 trillion market cap — reflects not genuine wealth but capitalized scarcity. US households pay approximately $3 trillion annually in rent (including imputed rent on owned homes). In a well-supplied market, those homes might be worth $30 trillion in structure and amenity value. The other $30 trillion is what Erdmann calls inflated land value — the premium extracted by artificial shortage. That premium shows up as homeowner wealth on one side of the ledger, but the matching liability (future tenants and buyers paying above-cost rents forever) never gets booked. Household net worth is overstated by approximately $30 trillion as a result.
This matters because it reframes the policy debate. Shortage-driven home price appreciation isn't a rising tide lifting all boats — it's a wealth transfer from the unhoused and future buyers to current landowners, with no net gain in real production. Real GDP is unchanged; the distribution of claims on it is just more skewed.
The California migration data the LA Times surfaced from the California Policy Lab illustrates this in human terms. People who left California for other states lowered monthly housing costs by an average of $700 and were 48% more likely (11 percentage points) to own a home than comparable neighbors who stayed. The catch Erdmann flags: those who left were already in worse financial shape — carrying $5,500 more in student debt on average than stayers. The mobility valve works, but it selects for the financially stressed, which means the people best positioned to absorb California's costs are the ones staying. That's the shortage sorting for you.
SYNTHESIS
These two data points tell a coherent story. The lock-in effect is the near-term friction keeping existing supply off the market; the shortage-as-wealth-illusion is the long-run structural distortion suppressing new supply and pushing marginal households out of high-cost markets entirely. McBride's data suggests the friction is easing at the margin — enough to matter for inventory over a 2-3 year horizon. Erdmann's framework suggests the structural problem is far larger than most accounting measures reflect, and that the "wealth" sitting in existing home values is partly a mirage that obscures the liability being handed to the next generation of buyers.
For buyers today: affordability math improves slowly as lock-in erodes and more inventory enters the market, but as long as supply constraints persist, they're bidding on assets partially priced as scarcity rents — not as productive capital.
(Note: Brian Potter's post this week covered the nature of technological progress — interesting reading, but outside the housing signal scope.)