PMMS 30Y signal
rising
+25 bps w/w (7.28%)
National benchmarks update from FRED (Freddie Mac PMMS weekly + Optimal Blue OBMMI daily indices). Your lender tables below stay separate for actionable comparisons.
Macro data as of 2026-10-06 (per series release schedule).
PMMS 30Y signal
rising
+25 bps w/w (7.28%)
Best 30Y (lenders)
6.290%
vs PMMS 30Y 7.28%
OBMMI − PMMS (30Y)
+18 bps
Conforming index vs survey average
FRED products
9
Top lenders
3
Scrolls automatically. Pause with reduced motion settings.
National survey averages, and the benchmark most often cited in headlines. New PMMS prints are weekly (commonly Thursday mornings US; confirm on FRED release calendar).
Source: Freddie Mac, Primary Mortgage Market Survey (MORTGAGE30US / MORTGAGE15US) via FRED, Federal Reserve Bank of St. Louis. See series pages on fred.stlouisfed.org for notes, copyright, and citation.
Aggregated from actual rate locks, which is useful for day-to-day movement between PMMS survey prints. Some calendar days may have no observation.
Source: Optimal Blue Mortgage Market Indices (OBMMI) via FRED, Federal Reserve Bank of St. Louis. See each series on fred.stlouisfed.org for copyright, citation, and permitted use.
The rates below are sample, illustrative figures for context and learning. They are not personalized quotes, advertisements of available credit, or commitments to lend. Always confirm costs and eligibility with a licensed lender for your situation.
The 30-year and 15-year figures here come from Freddie Mac's Primary Mortgage Market Survey, published every Thursday. It is the number nearly all news coverage is quoting when it says rates moved this week, which makes it the right benchmark for tracking direction over time.
It is the wrong number to expect on your own quote, and the reason is in the methodology. The survey reflects pricing offered to borrowers with strong credit, a conventional conforming loan amount, roughly 20 percent down, and points paid. Change any of those and your rate changes. A lower credit score, a smaller down payment, a condo, an investment property, a jumbo loan amount, or zero points will each move you away from the survey figure, usually upward.
Treat the benchmark as a thermometer for the market rather than a price list. If it has drifted down a quarter point over a month, that is genuinely useful information about the direction of your eventual quote. The quote itself only exists once a lender has your application.
Mortgage rates track the bond market, not the federal funds rate. The closest single reference point is the 10-year Treasury yield, which is why it sits on this page next to the mortgage benchmarks. Lenders price mortgage-backed securities against it, and the gap between the two is the spread.
That spread is not fixed, and watching it explains a lot of apparently strange rate behavior. When investors are nervous, when prepayment risk looks high, or when demand for mortgage bonds is thin, the spread widens and mortgage rates rise even if Treasury yields held steady. The historically normal spread has been in the neighborhood of 1.7 points, and it has spent meaningful stretches well above that.
This is also why a Federal Reserve announcement often produces no movement in mortgage rates at all. By the time the Fed acts, the bond market has usually priced the expected decision in weeks earlier. Inflation and employment reports frequently move mortgage rates more than the Fed meeting itself does, because they change what the market expects next. We cover the mechanics in why mortgage rates move when the Fed does nothing and how Treasury yields set mortgage pricing.
The most common mistake is waiting for a specific number. Rates are not on a schedule, and a borrower who decided in 2022 to wait for a 5 percent handle spent years renting while prices kept moving. The useful question is not where rates are headed, which nobody knows, but whether the payment at today's rate works for your situation.
Run the math before you form an opinion about the market. Put a realistic rate, a bit above the survey figure, into the affordability calculator, then check the result against your full debt picture in the DTI calculator. If you already own, the refinance workspace shows whether a move makes sense at current levels.
Two public series can still disagree with the chart on this page. A weekly survey average and a daily lock index are different clocks, and neither one is your credit score or your lock length. If you are months from a contract, the direction over a couple of weeks is enough. If you are inside a contract, ask the lender for a same-day par quote and compare it with the survey only as weather. The write-up of that gap is OBMMI versus PMMS.
When you are close to locking, read lock now or wait. Our month-by-month commentary lives in Learn, and the sourcing and refresh schedule behind every figure on this page is documented in our methodology and disclosures.
The headline 30-year and 15-year averages are Freddie Mac's Primary Mortgage Market Survey. Freddie publishes the release on its own site, and the St. Louis Fed republishes the 30-year history as MORTGAGE30US and the 15-year history as MORTGAGE15US. The survey is weekly. A Thursday print does not describe Friday afternoon's lock desk.
The 10-year Treasury yield next to those lines is DGS10 on FRED, a daily constant-maturity yield. It can move on a day the mortgage survey does not update. That is not a data error. It is two clocks. The daily mortgage indices in the depth chart are Optimal Blue's OBMMI series, also via FRED, and they describe locked loans rather than a survey of offered rates. The difference between a survey average and a lock tape is written up in OBMMI versus PMMS.
Inflation context, when we mention it beside these charts, comes from the Bureau of Labor Statistics Consumer Price Index, not from a mortgage desk. Employment context comes from the same agency's jobs release. Neither release is a rate. They are inputs the bond market uses before a lender prices your file. If a number on this page and the source disagree, the source wins, and the chart is the thing that gets corrected.