Good data.
Clear explanations.
Our aim is to make mortgage history useful without suggesting that a national survey is a personal quote or a prediction. Every chart is based on the dated snapshot described below.
The official sources
Official source and documentation ↗ · Original data file ↗
Retrieved October 30, 2026. SHA-256 of the original downloaded file:
fb183b83bb5f65b8962c4bfd0b44642ee0cd6286b542e3f8a73ac6e14c57a4c2Official source and documentation ↗ · Original data file ↗
Retrieved October 30, 2026. SHA-256 of the original downloaded file:
f3c22c8592d9134fa5ab305fa2df6e37dfd2388f10168985fdbc84082deb4ff0Mortgage series coverage
- 30-year fixed: April 2, 1971 – October 29, 2026. 2,901 observations.
- 15-year fixed: August 30, 1991 – October 29, 2026.
- 5/1 ARM: January 6, 2005 – November 10, 2022. Discontinued; shown only as an optional historical rate overlay.
The archive also includes historical points/fees, ARM margins, and a rate spread. They are retained in the research snapshot. Selected-week points are shown where reported. Missing fields remain missing; we do not treat blanks as zero or invent observations.
A survey change worth knowing
On November 17, 2022, Freddie Mac changed PMMS from a lender survey to a method based on loan applications submitted through its system. That is a methodological break, marked in the chart. It also discontinued 5/1 ARM and points/fees reporting. See Freddie Mac’s explanation. The CSV readme describes the older survey; the newer methodology note takes precedence for post-change observations.
Inflation, measured separately
We use BLS series CUUR0000SA0: CPI for All Urban Consumers, all items, U.S. city average, not seasonally adjusted, with an index base of 1982–84 = 100. The imported index covers January 1913 – Sep 2026. The chart shows only the overlapping mortgage period. Inflation is calculated as 100 × (index this month ÷ index in the same month a year earlier − 1).
CPI is monthly and mortgage observations are weekly. We retain their separate dates rather than manufacturing weekly CPI observations. A missing month or missing prior-year value produces a gap in the inflation line. Source-file missing periods: 2025-10. The optional line provides context, not proof of causation and not an inflation-adjusted mortgage rate.
Where your rate stands
The comparison counts weeks in the selected fixed-rate series with a rate strictly higher than the entered rate, divided by the number of nonmissing weeks in that entire series. We display the result rounded to the nearest whole percentage. Equal-rate weeks are not counted as higher. The statistic always uses the full available series, regardless of chart zoom. This describes recorded survey weeks, not individual borrowers, deals, or your creditworthiness.
Payment comparisons
We use the standard level-payment amortization formula, M = P × r ÷ (1 − (1 + r)⁻ⁿ). P is the loan amount, r is the entered annual mortgage interest rate divided by 1,200, and n is 360 or 180 monthly payments depending on the fixed-rate series. At zero interest, M = P ÷ n. Here “rate” means the nominal mortgage interest rate, not APR, which can include additional financing costs.
For the historical comparison, we hold P and n constant and substitute the selected historical survey rate. The payment-mode chart does the same at every observation. Calculations use unrounded intermediate values; displayed dollar amounts are rounded. Taxes, insurance, PMI, fees, points, closing costs, changes in house prices and wages, and inflation adjustments are excluded. The ARM overlay is not used to model future adjustable payments.
Scenarios and extra payments
Scenario payments use the same fixed-rate formula with each scenario’s own loan amount, rate, and 15- or 30-year term. Total interest equals all scheduled principal-and-interest payments minus original principal. Different amounts and terms are not equivalent borrowing decisions.
The extra-payment tool starts with the entered principal before the first payment. Each month, interest is the outstanding balance times the annual rate divided by 1,200. The scheduled payment plus extra principal is capped at the remaining balance plus interest. The final payment may be smaller. Balances and totals use unrounded intermediate values, and the simulation stops at payoff. No prepayment fees or recasting are assumed. The month slider and schedule expose the underlying balances. Actual lender rounding and servicing practices may differ.
Refinance comparisons
Enter the actual remaining balance, fixed rate, and remaining term of the current loan. The current payment is inferred by amortizing that balance over the remaining months. The candidate loan starts with the same balance, plus closing costs if financed. Otherwise costs are paid in cash at the start. No cash-out is assumed.
Simple payment break-even is total entered fees divided by initial monthly payment savings, rounded up to whole months. It is unavailable when there are no initial payment savings or the calculated point exceeds either loan’s term. This is not an equity-adjusted break-even. The holding-period comparison instead subtracts the replacement loan’s payments made, cash closing costs, and debt remaining from the current loan’s payments made and debt remaining at the same month. Paid-off loans retain zero balance and no additional payments. Full-term savings compare total remaining interest on the old loan with total interest plus fees on the new loan. Financed fees are counted once, and their additional interest is included.
These comparisons exclude taxes, insurance, PMI, tax deductions, investment opportunity costs, discounting, future rate changes, and cash-out. A payment reduction can coexist with a higher total cost. See the CFPB guidance on comparing loan estimates and prepayment penalties.
Historical chapters and challenges
Chapters and quiz answers refer to specific published 30-year weekly observations from the same snapshot. They are optional educational interactions, not forecasts. Replay advances through selected observations rather than interpolating rates. The same dated attribution appears on generated comparison cards.
Averages and updates
The series average and annual averages are simple arithmetic means of the available weekly observations. They are not volume-weighted, daily-weighted, or borrower-weighted averages. Incomplete years are marked YTD. Historical high and low markers use the earliest date when an extreme occurs more than once.
This is a static snapshot, not a live feed. Mortgage observations are current through October 29, 2026; CPI through Sep 2026. A data refresh requires a new official snapshot and a site rebuild. Data may be revised by the source. We provide no forecast of rates.
Attribution and independence
Mortgage source: Freddie Mac Primary Mortgage Market Survey (PMMS). Inflation source: U.S. Bureau of Labor Statistics. Source data are provided as is, without warranties. Historic Mortgage Rates is independent and is not affiliated with, sponsored by, or endorsed by either source. Source names are used for identification and attribution.
Freddie Mac content-use policy · Freddie Mac terms · BLS linking and copyright information