You Can Refinance a Rate. You Can’t Refinance a Purchase Price

There’s a sentence I’ve heard in this industry for four years now, usually delivered with real conviction: you can always refinance the rate, but you can only buy the house once. It’s true. That’s what makes it dangerous.

The statement is factually correct and the conclusion most people draw from it is backwards. Used properly, the asymmetry between price and rate is the strongest argument for negotiating harder and buying less house than you qualify for. Used the way it’s usually deployed — as a reason to stop worrying about the payment and sign — it’s the single most expensive piece of advice given to American homebuyers in the last decade.

I want to work through why, with actual numbers, because the arithmetic is more persuasive than the slogan.

What’s actually permanent and what isn’t

Start by separating the two sides cleanly.

Your purchase price is fixed at closing and never changes. It sets your loan amount. It sets your initial equity position. In most jurisdictions it resets your property tax assessment. It establishes the replacement-cost basis your insurer uses. If you put less than 20% down, it determines your mortgage insurance premium. And it’s the number you have to climb back above before you can sell without writing a check.

Your interest rate is a contract you may be able to replace. Emphasis on may. More on that shortly.

So far this supports the slogan. Now run the numbers on what the price side is actually worth.

Take the national median existing-home price, which NAR put at $429,100 in August. Suppose you negotiate $25,000 off — entirely plausible in a market where Realtor.com found price cuts on 20.8% of September listings, the highest September share since 2018, and where builders in high-supply metros like Austin have been discounting by double digits.

What does that $25,000 actually buy you?

At the closing table. With 10% down, you need $2,500 less cash. That’s immediate and it’s often the difference between closing and not closing.

On the loan. You borrow $22,500 less. Over a 30-year term you repay that principal plus three decades of interest on it. Depending on where rates sit, the total cost of that borrowed $22,500 lands somewhere between one and a half and two times the principal itself. Call the lifetime saving somewhere in the range of $45,000 to $60,000.

On property taxes. A $25,000 lower assessment saves roughly $250 to $300 annually in many jurisdictions. Over thirty years, with reassessments compounding off a lower base, that’s commonly $9,000 to $12,000.

On insurance and PMI. Lower coverage basis, lower premium. Lower loan amount, lower mortgage insurance, and you reach the equity threshold to cancel it sooner.

On your equity position. You start $25,000 further from underwater. In a market where ICE counted roughly 813,000 underwater borrowers — up 44% year over year — that cushion is not theoretical.

Add it up and a single $25,000 negotiation is worth somewhere around $55,000 to $75,000 across the life of the loan, and it’s unconditional. It doesn’t require the Fed to do anything. It doesn’t require you to qualify for anything later. It happens whether or not the market cooperates.

That’s the part of the asymmetry people skip.

The conditions nobody mentions

Here’s where the slogan quietly falls apart. “You can always refinance” contains three assumptions, and all three can fail.

You have to qualify again

This is the one that surprises people most. A refinance is a new loan application, underwritten from scratch. Your income, credit, debt-to-income ratio, and appraised value all get re-examined — and not as they were when you bought, but as they are on the day you apply.

Lenders generally want a minimum credit score, sufficient equity, a debt-to-income ratio inside their threshold, documented income, and a seasoning period on the existing loan. Plenty of people who qualified at purchase cannot qualify three years later. They changed jobs. They went self-employed. They had a child and income dropped. They took on a car payment. Their appraisal came in light because half the country’s major metros have been flat or declining.

The industry-wide numbers are sobering. Of roughly 12 million mortgage applications in a recent year, about a quarter were withdrawn and nearly a fifth were declined. Something in the neighborhood of 43% never became loans. If you want a clear picture of what underwriting will actually ask of you before you build a plan around it, read up on what it takes to refinance a fixed-rate mortgage rather than assuming approval.

Rates have to fall enough to beat closing costs

Refinancing isn’t free. Closing costs typically run 2% to 5% of the loan amount. A small improvement in your rate doesn’t clear that hurdle — you need a reduction substantial enough that the monthly savings recover several thousand dollars in costs within a reasonable window.

Which means “rates came down a little” doesn’t help you. They have to come down meaningfully, and they have to do it while you still qualify and still own the house.

The forecast has to be right

This is the one the 2022 cohort learned the hard way, and it’s the reason I’m writing this piece rather than repeating the slogan.

When “marry the house, date the rate” was coined, the assumption underneath it was not reckless — most of the industry genuinely believed rates were headed back down within a year or two. Major institutions forecast a substantial decline within a short window. Buyers were told the high payment was temporary.

It wasn’t. Four years on, borrowing costs have been elevated the entire time, and meanwhile property taxes and insurance premiums rose, pushing those payments up rather than down.

Nicole Rueth of CrossCountry Mortgage, who never endorsed the phrase, has said it “put a lot of people in tough situations.” Mortgage professional Jeff Winchell put it more memorably: outside of Shakespeare, he noted, a rhyming couplet is somebody’s marketing slogan. Another lender’s advice to clients was blunter still — look to lower your rate when there’s a benefit, but “don’t bet the house on it.”

I’d add something from the lending side. A loan officer and a real estate agent are both paid on the transaction closing. Neither is paid on whether your refinance materializes in 2029. That’s not an accusation of bad faith — most people in this business are trying to help — but it is a structural reality you should factor into whose forecast you’re relying on.

And the current forecast backdrop does not favor patience. The Federal Reserve raised rates in September for the first time since 2023, and its own projections now imply no net easing through 2027. The question of will mortgage rates go down in 2026 has a very different answer today than the consensus had a year ago.

So what does the asymmetry actually tell you to do?

Here’s my argument, and it’s the opposite of how this framing usually gets used.

Because the price is permanent and the rate is conditional, you should spend your energy on the thing that’s permanent.

The slogan says: don’t worry about the rate, just buy. What it should say: the rate might fix itself, so put all your negotiating leverage into the price — because that’s the part nobody will fix for you later.

In practice, that means five things.

1. Buy a payment you can carry unchanged, indefinitely. This is the non-negotiable. If the purchase only works on the assumption of a future refinance, it doesn’t work. Treat any future rate improvement as a bonus, not as load-bearing structure. A Michigan lender put the standard well: the approach works when buyers can comfortably afford the home under today’s terms and view refinancing as upside rather than a requirement.

2. Use the leverage that exists right now. NAR counted 1.62 million existing homes for sale in August, the first reading above 1.6 million since November 2019, with 4.9 months of supply — the most in over a decade. Realtor.com’s median list price has declined year over year for ten consecutive months. Sellers are negotiating. This leverage is cyclical and it will disappear the moment borrowing costs fall and sidelined buyers return.

3. Negotiate terms, not just price. Seller-paid closing costs, a temporary buydown funded by the seller or builder, repair credits, a home warranty. These reduce your cash at closing and your early payments without requiring anything from the bond market. Builders in oversupplied markets are especially willing here.

4. Buy less house than you’re approved for. Your approval is a ceiling, not a recommendation. The gap between what you qualify for and what you comfortably carry is the margin that protects you if the refinance never comes.

5. Stack every program you’re eligible for. Down payment assistance, forgivable second mortgages, state housing finance agency programs, and first-generation buyer grants are substantially underused. Non-savings sources now fund 29% of all purchase down payments, a seven-year high. Before you decide you can’t afford the price you want, check what down payment assistance programs exist in your state, and get familiar with the first time home buyer loan programs that may reduce your required cash.

The counterargument, taken seriously

Let me give the other side its due, because there’s a real version of it.

Waiting also has a price, and it’s not zero. Every year you rent, you pay housing costs and build nothing. If home prices rise while you wait — and most forecasters expect modest national appreciation rather than declines — the price you’re protecting yourself from gets higher, not lower. And if rates do fall significantly, the buyers who waited will be competing against each other for the same inventory, with the leverage that exists today gone.

That’s a genuine tension and I won’t pretend otherwise. There’s a real scenario where lower rates in 2028 produce a more competitive, higher-priced market than the one in front of you right now, and the buyer who waited ends up worse off on both sides.

The resolution isn’t “always buy” or “always wait.” It’s this: the asymmetry argument is valid as a reason to act when the payment works at today’s terms. It is not valid as a reason to stretch beyond what works.

If you can comfortably carry the payment on a house you’ve negotiated hard for, the fact that the rate might improve later is a legitimate point in favor of moving now. If you can only carry the payment on the assumption that it improves later, the same fact is being used to sell you something.

Same sentence. Completely different decisions.

When waiting is the right answer

I’d rather lose a reader than have someone buy into trouble because an article sounded encouraging. Waiting is correct if:

• Your reserves after closing would be thin. You need months of expenses available, not zero.
• Your income is unstable, newly self-employed, or heavily commission-based — both for carrying the payment and for qualifying later.
• You may relocate within about three years. Transaction costs on both ends will likely exceed any equity you build.
• The payment would consume an uncomfortable share of your take-home pay.
• You’re buying in a metro with meaningfully declining values and putting minimal money down. Thin equity in a soft market is how people get stuck.

None of those are fixed by a future refinance. Several of them are made worse by it.

The bottom line

The asymmetry is real. Price is permanent; rate is replaceable. I’d just ask people to follow the logic all the way through to its actual conclusion.

If the price is the permanent part, the price is where your scrutiny belongs. Negotiate it like it’s the only number that matters, because in a meaningful sense it’s the only number you fully control. Then buy a payment that works without requiring anything from the Federal Reserve, and treat any future refinance as found money.

The buyers who got hurt over the last four years weren’t hurt by high rates. They were hurt by paying a price that only penciled if the rate came down — and then it didn’t.

Get the price right, and the rate becomes a detail you can fix later or live with. Get the price wrong, and no refinance will save you.

References

Bankrate. (2026, June 29). ‘Buy now, refinance later,’ they said. Mortgage rates said otherwise. 

ConsumerAffairs. (2025). What does “marry the house, date the rate” mean? 

Realtor.com. (2026, September). Monthly housing market trends report.