This matters more than usual right now. Rates have been sitting above 7% on some measures while housing supply has quietly climbed to its highest level in more than a decade — two facts that sound contradictory until you understand that they’re not measuring the same thing, on the same schedule, in the same way.
Two "current" mortgage rates, both correct, both incomplete
Consider two numbers that could both appear in the same week’s news cycle. One survey puts the average 30-year fixed rate at 7.12% APR for the week ending September 24, calculated from daily annual percentage rates recorded by Zillow over the prior five business days. Another official report, released earlier in the same reporting cycle, cites Freddie Mac’s average 30-year fixed rate at 6.67% for August, up from 6.54% in July.
Neither number is wrong. They simply answer different questions. One is a short, rolling window measured in APR — which folds in certain lender fees and costs on top of the note rate. The other is a broader monthly average that Freddie Mac typically reports closer to the plain note rate, not APR. Add in the fact that they cover different calendar periods, and a half-point gap between "current" mortgage rates stops looking mysterious and starts looking like ordinary measurement noise.
The practical lesson isn’t that one source is more trustworthy than the other. It’s that "the mortgage rate" is not a single, stable object — it’s a family of related but distinct measurements, and a headline rarely tells you which one you’re looking at.
Benchmark, meet borrower
The distance between a national average and your actual offer is where a real transaction gets decided. A benchmark tells you the weather; a Loan Estimate tells you what to wear.
| Aspect | National benchmark average | Your actual lender quote |
|---|---|---|
| What it measures | A broad sample across many lenders, aggregated over a short window | One lender’s price for your specific loan file |
| Rate basis | May be reported as APR or as a plain note rate, depending on the source | APR reflects your rate plus lender fees and points on this loan |
| Time horizon | A weekly window or a monthly average — already slightly in the past by publication | Typically locked for a defined period, often 30–60 days, once you apply |
| What moves it | Treasury yields, inflation expectations, Fed policy expectations, bond-market sentiment | Credit score, down payment or equity, loan type and term, points paid, the property itself |
| Best use for you | Orientation — is the environment tighter or looser than a few months ago | The number that actually determines your monthly payment |
The takeaway isn’t that benchmarks are useless. They’re excellent for spotting direction and magnitude of change. They are a poor proxy for what you, personally, will be offered, because that number depends on variables no national survey can capture — your credit profile, your down payment, the term you choose, and how many points you’re willing to pay upfront to buy the rate down. A longer term lowers the monthly payment by spreading it over more months, but it typically raises the total interest paid over the life of the loan — a trade-off no weekly average can make for you.
Budgeting for a moving target
If the rate you’ll get next year is genuinely uncertain even to forecasters — and it is; many expected rates near 6% by this point in the cycle and didn’t get there — the more useful skill isn’t predicting the number. It’s budgeting around a range.
One rule of thumb from a Realtor.com analysis of rate swings since 2000 offers a rough way to size that range: for a purchase three months out, build in roughly a half-percentage-point cushion; six months out, three-quarters of a point; a year out, a full point. The analysis found this kept a buyer’s budget realistic about 80% of the time. Run your affordability math at the higher end of that range, not the lower one — if today’s rate is around 7%, and you’re a year from buying, test whether the payment still works at 8%.
That kind of stress-testing is more actionable than watching weekly headlines, because it converts an unpredictable input into a planning boundary you control.
Which housing number answers which question
The rate isn’t the only place where headlines can seem to contradict each other. In the same month, it’s entirely possible to read that home sales fell, that supply hit a decade-plus high, and that prices climbed to a new record — all at once, all true. That’s not a data error; it’s four different rulers measuring four different things.
| Statistic | What it actually measures | What it does NOT tell you |
|---|---|---|
| Existing-home sales (seasonally adjusted annual rate) | Closed transactions, annualized — sales fell 2.0% month over month to 3.98 million units in August, and were down 1.2% year over year | Nothing about current asking prices or how many homes are newly listed |
| Months’ supply of inventory | How long current stock would last at the present sales pace — 4.9 months in August, the highest in more than ten years | Whether that inventory is priced attractively or sitting stale |
| Median sales price | The midpoint price, less skewed by ultra-high-end sales than an average — $429,100 in August, up 1.6% year over year | Nothing about volume; a rising median can occur even while total sales fall, partly because the mix of what’s selling can shift toward pricier homes |
| Days on market | How long listed properties typically sit before selling — 31 days in August versus 29 in July | Whether that’s due to weak demand, seasonal slowdown, or simply more choices for buyers |
The National Association of Realtors is explicit that its sales figures are based on closings pulled from multiple listing services, which is why they can diverge from other series, such as the Census Bureau’s new-home sales data, based on signed contracts rather than closings. It also notes that medians, not averages, are used for price because medians resist distortion from a handful of very expensive sales — a good general habit to borrow whenever you see a housing statistic described as an "average" versus a "median".
None of this means the housing slowdown is illusory, or that ample supply automatically favors buyers everywhere; NAR’s chief economist has pointed out that rates and sales tend to move in opposite directions, which is consistent with elevated rates weighing on the closings recorded in August. But a single month’s print — even an unusually high inventory figure — is a data point in a longer series, not proof of a durable new trend on its own.
Where the Fed actually fits in
It’s tempting to treat Fed meetings as mortgage-rate announcements. They aren’t. The Federal Reserve sets short-term policy rates; it does not directly set the 30-year mortgage rate. What connects the two is the bond market: mortgage pricing tracks yields on longer-term Treasurys and mortgage-backed securities, which move on expectations about where Fed policy — and inflation — are headed. When the 10-year Treasury yield climbs, as it reportedly did to its highest level since 2007 amid a bond-market selloff, mortgage rates tend to follow, even before the Fed itself acts. Markets were pricing in a high probability of further rate hikes into the Fed’s late-October meeting and beyond, according to futures-based tracking cited in the same reporting — but futures-implied odds are a market expectation, not a guarantee of what the Fed will do or where mortgage rates will land.
The question worth asking instead
"Are rates high or low right now?" is the wrong question to build a decision on — it invites you to react to a single printed number. The better question is: what is this specific statistic measuring, over what window, and does it tell me anything about my situation, or only about the market’s average temperature?
Weekly and monthly rate benchmarks, sales counts, inventory levels, and median prices are all legitimate, carefully constructed data. They’re just built to describe the market in aggregate, not to describe you. The number that matters for an actual purchase — your Loan Estimate, with your credit profile, your down payment, your negotiated points or seller concessions — only exists once you’re in a real transaction with a real lender. Everything before that is context: useful for calibrating expectations and stress-testing a budget, not for predicting what you, personally, will be offered next month or next year.
