What would a rate change mean for housing?
Interest-Rate Transmission Simulator
Choose a hypothetical change in the federal funds rate and follow it down the chain — mortgage rates, the payment on a real loan, and the housing activity that has historically moved with financing costs. Every number is tagged with its provenance, and every estimate carries a visible range.
Scenario controls
Starting point: 30-year fixed mortgage rate 6.55% (week of Jul 16, 2026) · median new-home price $403,200 (Q1 2026). Estimates describe changes over a 12-month horizon.
+1.00pp
Default: $322,560 — 80% of the median new-home price (20% down payment).
Hypothetical policy-rate change
+1.00pp over 12 months
A scenario you chose — not a forecast of what the Federal Reserve will do.
Implied 30-year mortgage-rate change
+0.35pprange +0.17 to +0.52 pp
Historically, a 1pp change in the federal funds rate over 12 months has been associated with about a 0.35pp same-direction move in the 30-year mortgage rate. Applied to your scenario, the rate moves from 6.55% to about 6.90% (range 6.72%–7.07%).
OLS on monthly data, 1990–2026, contemporaneous · slope 0.35 pp per pp · R² 0.34 · n = 438
Monthly payment on your loan
- At today’s rate (6.55%)
- $2,049/mo
- At the implied rate (6.90%)
- $2,123/mo
- range $2,086–$2,161
- Change
- +$74/mo (+3.6%)
Under the standard 28% front-end affordability rule, the gross annual income needed to carry this loan changes by +$3,172.
Standard fixed-rate amortization on a $322,560 loan over 30 years — exact arithmetic, no estimation. The only estimated input is the mortgage rate from the previous step.
Historically associated 12-month changes in housing and spending
- Housing starts
- -2.0%range -3.1% to -1.0%
- Applied to the mortgage-rate change above · OLS 1990–2026, driver lagged 6m · R² 0.09 · n = 438
- New single-family home sales
- -1.1%range -1.7% to -0.6%
- Applied to the mortgage-rate change above · OLS 1990–2026, driver lagged 6m · R² 0.03 · n = 437 — R² near zero: historically this driver explains almost none of the variation, so treat the point estimate as close to uninformative.
- Construction employment
- +0.2%range +0.1% to +0.3%
- Applied to the mortgage-rate change above · OLS 1990–2026, driver lagged 12m · R² 0.01 · n = 438 — R² near zero: historically this driver explains almost none of the variation, so treat the point estimate as close to uninformative.
- Durable-goods spending
- +0.1%range +0.1% to +0.2%
- Applied to the policy-rate change above · OLS 1990–2026, driver lagged 6m · R² 0.00 · n = 437 — R² near zero: historically this driver explains almost none of the variation, so treat the point estimate as close to uninformative.
How to read this chain
The mechanism runs through financing costs. A higher policy rate raises the cost of funds across the yield curve; mortgage rates reprice; monthly payments on new loans rise; fewer households clear the affordability bar at prevailing prices; builders respond to thinner demand by starting fewer homes; and, with a longer delay, hiring in construction and spending on big-ticket durables adjust. Each arrow in that story is plausible economics — but each estimated link above is an association drawn from history, not a measured causal effect, because the Federal Reserve moves rates in response to the very conditions these outcomes reflect.
Uncertainty widens as the chain gets longer. The pass-through step is the tightest link; the housing-activity steps depend on that estimate and add their own much weaker fits; the employment and durables links are close to uninformative on their own. The displayed ranges apply each step’s 0.5×–1.5× judgment band to the central input from the step before, so the true compound uncertainty is wider than any single range shown. Treat the cascade as a structured way to think, not a prediction machine.
Assumptions & limitations
- These are associations, not causal effects. Monetary policy is endogenous: the Federal Reserve raises rates when the economy is running hot and cuts them when it weakens, so historical rate changes coincide with the very conditions that were already moving housing and spending. A regression over that history cannot separate the effect of the rate change from the effect of the conditions that prompted it.
- Pass-through varies by regime. How much of a policy move reaches mortgage rates depends on whether the move was expected, how the Fed communicates, and conditions in the mortgage-backed securities market. A single slope estimated over 1990–2026 averages across very different regimes — including years at the zero lower bound and the 2022 tightening, when mortgage rates moved far more than the historical average pass-through implies.
- Ranges are judgment overlays. The low–high bands shown are 0.5× to 1.5× the point estimate — a deliberate honesty device, not a statistical confidence interval. They exist to keep the weakness of simple two-variable regressions visible, and the compound uncertainty across chained steps is wider than any single band.
- The horizon is fixed at 12 months. All associations are estimated on 12-month changes; the simulator says nothing about the first weeks after a policy move or about effects beyond a year.
- The estimation methodology, charts of the underlying relationships, and a fuller discussion of timing and endogeneity live in the rate-transmission research module.
Model provenance
Every estimated relationship used above, re-estimated from the committed data snapshots on each build. Slopes are per 1pp change in the driver.
| Relationship | Window | Lag | Slope | R² | n |
|---|---|---|---|---|---|
| 12m change in 30-year mortgage rate on 12m change in federal funds rate | 1990–2026 | 0m | 0.35 pp per pp | 0.34 | 438 |
| 12m % change in housing starts on 12m mortgage-rate change (lagged 6m) | 1990–2026 | 6m | -5.92 % per pp | 0.09 | 438 |
| 12m % change in new home sales on 12m mortgage-rate change (lagged 6m) | 1990–2026 | 6m | -3.23 % per pp | 0.03 | 437 |
| 12m % change in construction employment on 12m mortgage-rate change (lagged 12m) | 1990–2026 | 12m | 0.63 % per pp | 0.01 | 438 |
| 12m % change in durable-goods spending on 12m fed-funds change (lagged 6m) | 1990–2026 | 6m | 0.11 % per pp | 0.00 | 437 |
Source series
Board of Governors of the Federal Reserve System via FRED (FEDFUNDS) · Latest: Jun 2026
Freddie Mac (Primary Mortgage Market Survey) via FRED (MORTGAGE30US) · Latest: Jul 16, 2026
U.S. Census Bureau / HUD via FRED (HOUST) · Latest: Jun 2026
U.S. Census Bureau / HUD via FRED (HSN1F) · Latest: May 2026
U.S. Bureau of Labor Statistics (CES) via FRED (USCONS) · Latest: Jun 2026
U.S. Bureau of Economic Analysis via FRED (PCEDG) · Latest: May 2026
U.S. Census Bureau / HUD via FRED (MSPUS) · Latest: Q1 2026