Why Your Rate Is Higher Than the Headline Rate

Freddie Mac says 6.71 percent, the daily index says 6.88, and your term sheet says something else. Here are the four spreads that explain the gap.

Re:InvestorHub Team · 2026-09-03 · Deal Analysis

On September 3, 2026, four different organizations published the average 30-year fixed mortgage rate. Freddie Mac said 6.71 percent. The Mortgage Bankers Association said 6.79 percent plus 0.65 points. Mortgage News Daily said 6.88 percent. The same day, Mortgage News Daily quoted the 30-year jumbo at 7.05 percent and noted that many borrowers were already seeing rates at 7 percent or higher. Every one of those numbers is accurate. None of them is the number on your term sheet, and the gap between them is not a lender marking you up. It is four measurable spreads, and once you know what they are you can underwrite the rate you will actually sign instead of the one in the headline.

Why do the published mortgage rates disagree with each other?

They disagree because they measure different things, and each publisher says so openly. The disagreement is definitional, not a discrepancy anybody is hiding.

• Freddie Mac’s Primary Mortgage Market Survey is built from purchase applications submitted through Loan Product Advisor, and it excludes upfront costs. It releases Thursdays at noon Eastern and averages the loan rates offered from the prior Thursday through Wednesday, so it is a backward-looking week, not today.
• The Mortgage News Daily index updates every weekday around 4pm Eastern and accounts for upfront costs. Mortgage News Daily states the distinction directly: their index accounts for upfront costs whereas the Freddie Mac weekly survey rate does not.
• The Mortgage Bankers Association weekly survey splits the two apart, reporting a contract rate and the points separately. For the week ending August 28, 2026, that was 6.79 percent with 0.65 points on conforming balances of $832,750 or less.

That is the whole first spread. Freddie Mac at 6.71 percent and Mortgage News Daily at 6.88 percent are 17 basis points apart largely because one counts the cash you bring to closing and the other does not. Neither is wrong. If you compare a quote that includes points against a survey that excludes them, you will conclude your lender is expensive when the two numbers were never measuring the same thing.

What are the four spreads between the survey and your term sheet?

Work through them in order. Each one moves your rate in the same direction, which is why the total gap surprises people who only expected one adjustment.

1. Upfront cost. Points are prepaid interest expressed as a percentage of the loan. At 0.65 points, a $300,000 loan costs $1,950 in cash at the closing table to hold the quoted rate. A survey that excludes that cost will always print lower than a quote that includes it.
2. Timing. The Freddie Mac survey averages the prior Thursday through Wednesday. Your lock is priced off where mortgage-backed securities trade this afternoon. In a week when rates are moving, you are comparing last week’s average to today’s market. On September 2 the daily index rose into the 6.9s for the first time in more than a year, which a weekly average will not show you until the following Thursday.
3. Occupancy. Every rate you see quoted in the news is for an owner-occupied primary residence. A non-owner-occupied investment property carries a pricing add-on from the agencies, and it is applied on top of whatever the market rate is that day. The headline is never quoting your rental.
4. Product and credit overlays. Agency conforming, jumbo, and debt service coverage ratio loans price on separate ladders. On the same day the daily 30-year index sat at 6.88 percent, the 30-year jumbo sat at 7.05 percent. Your credit score, loan-to-value ratio, and whether the loan is cash-out add further adjustments.

The fourth spread is where most investor confusion lives, because a DSCR loan (Debt Service Coverage Ratio, a loan underwritten to the property’s income rather than yours) prices differently from a conventional loan for reasons that have nothing to do with the day’s market. That comparison deserves its own treatment, and we have written it: see DSCR vs conventional loans for when the premium pays for itself, and hard money and DSCR as sequential financing for how the two legs fit together.

What does the spread actually cost on a real loan?

Take a $300,000 loan on a 30-year fixed and run principal and interest at each of this week’s published numbers. The monthly differences look small in isolation and compound into real money across a hold.

• At 6.71 percent, the Freddie Mac survey rate: $1,938 per month.
• At 6.79 percent, the MBA contract rate: $1,954 per month, or $191 more per year.
• At 6.88 percent, the daily index: $1,972 per month, or $408 more per year.
• At 7.05 percent, the jumbo rate: $2,006 per month, or $818 more per year.
• At 7.30 percent, a plausible investor quote once the occupancy add-on lands: $2,057 per month, or $1,427 more per year.

The distance from the headline to the investor quote is $119 per month on this loan. That is $1,427 a year of cash flow that exists in your spreadsheet and not in your bank account, and it is the difference between a deal that clears your threshold and one that does not. Add the $1,950 in points and the first year of the gap is closer to $3,400.

How should you underwrite a rate you have not locked yet?

Underwrite the quote, not the survey, and get the quote in a form you can actually compare. Three habits do most of the work.

1. Ask every lender for rate and points together, on the same day, for the same lock period. A 6.75 percent quote with 1.5 points and a 7.0 percent quote with zero points are not the same loan, and you cannot rank them until both are stated the same way.
2. Ask for the par rate, meaning the rate at zero points. It gives you a clean baseline to compare lenders against and tells you what you are really paying to buy the rate down.
3. Underwrite the investment property add-on explicitly rather than assuming the headline applies. If you do not know your lender’s current add-on, underwrite a range and see whether the deal survives the top of it.

One more piece of context for anyone planning to refinance out of a current position. The MBA Refinance Index fell 1 percent for the week ending August 28 and sat 19 percent below the same week a year earlier, with the refinance share of applications at 41.8 percent. Mike Fratantoni, the MBA’s chief economist, attributed the drop to rates reaching their highest levels in four weeks as concerns about inflation and deficits pushed yields higher. A refinance exit is a plan that depends on a market you do not control, and that market is currently moving away from you.

What should you do this week?

Pull the term sheet on any deal you have in underwriting and check which number you built the model on. If it came from a headline, rebuild it at the quoted rate with the points included and the occupancy add-on applied. The four spreads above are not going to narrow because you did not budget for them, and a deal that only works at the survey rate was never a deal. It was a rounding error you were counting on.

If you are underwriting a deal whose exit depends on refinancing at a lower rate than you bought at, that assumption deserves its own stress test right now. We work through the floor of that argument in why dating the rate is dead, and the market has spent the last week pricing the opposite direction.