15-Year Refinance Calculator

What a 15-year refinance changes

A refinance is a new loan that pays off the old one. Set the new term at 15 years and three things move at once: the rate, the monthly payment, and the month the balance reaches zero.

The rate usually falls. A lender carries less risk over 180 payments than over 360, and Freddie Mac's survey shows the 15-year fixed rate sitting under the 30-year for years. In the last week of October 2026 the gap was 0.64 points.

The payment usually rises. The same balance now spreads across 180 payments instead of the 324 that were left, so every check carries more principal. That is the part people miss when they hear the words lower rate.

Even with an unchanged rate the payment would climb, because 180 payments must clear a balance that was spread over 324. The lower rate softens that jump. It rarely erases it.

So the decision is a trade, not a free win. You take on a bigger bill each month to cut the total interest and the years. The calculator above shows both sides, plus the closing costs that decide whether the trade pays.

Term choice is part of that trade. The same balance at the same rate costs more each month at 10 years than at 15, and less at 30. The quote also moves with the term, since a 30-year loan carries a higher rate than a 15-year one. Set the selector to 20 or 30 years and the payment drops, but the interest total climbs back toward where it started.

Where 15-year rates stand in late 2026

Freddie Mac publishes the Primary Mortgage Market Survey every Thursday. For the week ending October 29, 2026, the 15-year fixed rate averaged 6.60 percent, down from 6.68 percent a week earlier. The 30-year averaged 7.24 percent that same week, so 7.24 - 6.60 = 0.64 points separated the two. A year before, the 15-year averaged 5.41 percent.

The survey is a benchmark, not a quote. It averages rates on applications that lenders sent to Freddie Mac, and your own offer depends on credit score, equity, loan size, property type and state. A borrower with a 780 score and 40 percent equity can beat the average. A cash-out loan on a condo can land above it.

Weekly moves are small. Through October 2026 the 15-year averaged between 6.60 and 6.73 percent, a band of 0.13 points. Calling three lenders on the same afternoon will move your payment further than waiting a week for the next survey.

Points change the picture as well. A quote with 1.5 points can carry a rate well under the survey average, while a no-point quote sits above it. Compare the annual percentage rate and the total closing costs, not the headline rate alone.

Break-even: closing costs divided by the monthly saving

Closing costs are the price of admission. You pay them once, at closing, and the lower payment has to win them back. The count is one division:

Break-even months = closing costs / monthly payment saving.

Say the new payment is $54.11 lower and the costs are $4,800. Then 4,800 / 54.11 = 88.7 months. Divide by 12 and you get 88.7 / 12 = 7.4 years. Move, sell or refinance again before that month and the closing costs are money the payment never returned.

When the payment rises instead of falls, there is no break-even in months. The saving sits in the interest column and arrives at the end of the term. That is not a reason to skip the loan. It is a reason to be sure you can carry the higher payment for 180 straight months.

Costs vary by lender and by state. The Federal Reserve's refinancing guide puts typical refinance fees at 3 to 6 percent of the outstanding principal. On a $250,000 balance that is $7,500 to $15,000, and discount points come on top of it.

Worked example: $250,000 at 6.50 percent with 27 years left

Take a balance of $250,000 at 6.50 percent with 324 payments left. The principal and interest part of the payment is $1,638.89.

Now refinance to 15 years at 5.75 percent with $4,500 in closing costs. The new principal and interest payment is $2,076.03, so 2,076.03 - 1,638.89 = 437.14 more each month. That is a jump of 26.7 percent, and there is no monthly break-even to compute because nothing is being saved month to month.

The interest side runs the other way. Left alone, the loan would charge $280,999.53 in interest across its remaining 27 years. The 15-year loan charges $123,684.54 across 15 years. The gap is 280,999.53 - 123,684.54 = 157,314.99.

Add the $4,500 and the total outlay still falls, from $530,999.53 to about $378,185. That leaves $152,815 unspent over the life of the loan, which is the real prize here rather than a smaller check.

The payoff date tells the same story. Close in January 2027 and the 15-year loan ends in January 2042. The old loan would have run to January 2054, so the refi removes 12 years of payments.

Points, the fee list, and the tax rule

A discount point is one percent of the loan amount, paid at closing to buy a lower rate. On $250,000, one point costs 250,000 x 0.01 = 2,500.

A point only pays off if you keep the loan long enough. Take a $250,000 15-year loan at the survey rate of 6.60 percent, where the payment is $2,191.54. Buying the rate down a quarter point to 6.35 percent cuts the payment to $2,157.21, a saving of $34.33 a month. The point costs $2,500, so 2,500 / 34.33 = 72.8 months pass before it repays itself. That is about six years.

Points carry a tax rule too. IRS Publication 936 treats them as prepaid interest. On a refinance you generally deduct them a little at a time over the life of the loan, so $2,500 of points works out to 2,500 / 180 = 13.89 a month of deduction instead of one large write-off. The full deduction in the year paid is for points on a loan used to buy, build or improve the home, and only when several tests are met. Ask a tax preparer about your own return.

The other closing costs are ordinary: appraisal, credit report, title search, title insurance, recording fees, and the lender's origination charge. The CFPB lists them on its mortgage terms page and requires a Loan Estimate within three business days of your application, so you can compare two offers before you commit.

The three business days you can still walk away

Refinancing a home you already own is a non-purchase loan. Federal law gives you three business days to cancel it, and the rule sits in Regulation Z at 12 CFR 1026.15. Saturdays count as business days for this clock. Sundays and federal holidays do not.

The clock does not start until all three events have happened: you signed the note, you received the Truth in Lending disclosure, usually the Closing Disclosure, and you received two copies of the notice of your right to rescind. Then you have until midnight of the third business day to cancel in writing. A phone call does not do it. After a valid cancellation the lender has 20 calendar days to return what you paid.

This matters to the math above. Cancel inside the window and the old loan stands while the closing costs come back. Let the window close and those costs are sunk, which is why the break-even count starts at closing day and not at application day.

The notice is not a formality. It is the last moment when the numbers on the calculator still match the numbers on the paper, and the only moment when walking away costs nothing but time.

You should also hold the Closing Disclosure at least three business days before closing. Use that gap to line up the Loan Estimate against the final numbers, with attention on the points line and the lender fees, since those two decide most of the break-even.

When a 15-year refinance is the wrong move

Four cases argue against it, even when the rate looks good.

• Thin reserves. If the higher payment would eat more than a third of your cash cushion, one repair or layoff turns a good loan into a problem.
• A short stay. The example above needs 88.7 months, or 7.4 years, to return $4,800 in costs. Sell in year three and you paid for a benefit you never collected.
• Costlier debt. A credit card at 22 percent costs more per dollar than a mortgage at 6.60 percent saves. Clear that balance first.
• Retirement money. Borrowing from a 401(k) or draining an emergency fund to cover closing costs trades a known gain for an unknown risk.

One last check is worth the effort. Run the numbers twice, once with the quote in hand and once with a rate half a point higher plus $2,000 more in costs. If the plan still works at the worse set of inputs, the risk is small. If it only works at the perfect quote, it is a bet rather than a plan.