Are Mortgage Points Worth It? The Break-Even Math, Explained

Short answer: paying points is worth it if you'll keep the loan past the break-even point — usually somewhere in the 4-to-6-year range on today's rates — and it's a bad deal if you might sell or refinance before then. Everything else in this article is just the math behind that sentence, because "it depends" isn't useful without the numbers to plug in yourself.

WTF is a discount point, actually

A discount point is a fee you pay your lender at closing, upfront, in exchange for a lower interest rate for the life of the loan. One point costs 1% of your loan amount — on a $400,000 loan, one point is $4,000, paid at closing on top of your down payment and other closing costs.

In exchange, the lender typically knocks about 0.25 percentage points off your rate per point, though the exact ratio moves with the lender and the broader rate environment — some quotes will show more or less than a quarter point per point, so treat 0.25 as a rule of thumb to sanity-check a Loan Estimate against, not a guarantee (Bankrate, as of August 2026). You can usually buy fractional points too (0.5, 1.5, 2) — it's not all-or-nothing.

The break-even formula

This is the whole decision, boiled down to one calculation:

Cost of points ÷ monthly payment savings = months to break even

Below that number of months, you're still in the hole. Past it, you're banking savings for as long as you keep the loan.

Here's what that looks like with real numbers. As of August 13, 2026, Freddie Mac's Primary Mortgage Market Survey put the average 30-year fixed rate at 6.67% (Freddie Mac PMMS). On a $400,000 loan at that rate, your principal-and-interest payment is about $2,573/month. Buy one point for $4,000, and if it buys the standard 0.25-point discount down to 6.42%, your payment drops to about $2,507/month — a savings of roughly $66/month.

$4,000 ÷ $66 ≈ 61 months, or about 5 years, to break even.

Stay in the house and keep that loan for 10 years and you come out ahead by roughly $3,900. Sell or refinance in year 3, and you've just handed your lender $4,000 for nothing. The math is genuinely that binary — points are a bet on how long you'll hold the loan, not a universal upgrade.

That's also why points make more sense for a "forever home" purchase or a rate-and-term refinance you plan to sit on, and less sense if there's any real chance you relocate for a job, outgrow the place, or refinance again if rates drop further. If you're not sure how long you'll stay, run the break-even math at your actual best-guess timeline, not the optimistic one.

What the data actually shows

Theory is easy to write and easy to get wrong when lenders round differently or a "special" is really an LLPA (loan-level price adjustment) dressed up as a discount. So we pulled real, originated loans from the HMDA (Home Mortgage Disclosure Act) database — the federal, loan-level dataset covering nearly every mortgage originated in the U.S. — to see whether paying points actually correlates with a lower rate in practice, not just in a lender's marketing PDF.

Using the 2025 HMDA loan/application register (13.5 million records), filtered to originated (action_taken = '1'), conventional, 30-year fixed, first-lien, owner-occupied, closed-end, non-commercial purchase loans, and bucketing by discount points paid as a percentage of loan amount:

The pattern is clear at the high end: borrowers who paid 2% or more of their loan amount in discount points averaged a rate about 0.57 percentage points lower than borrowers who paid none — real, checkable confirmation that buying points does buy a lower rate, at scale. Running the same filters against the 2024 HMDA file (12.2 million records) shows the same shape: 6.71% for no points vs. 6.17% for 2%+ in points, a nearly identical spread.

One honest wrinkle worth flagging: the smaller point-payment bucket (0–2% of loan) came in roughly flat with — or even a hair above — the no-points group, rather than showing a smooth, straight-line discount. Our read is that HMDA's discount_points field doesn't cleanly separate "I voluntarily bought down my rate" from "the lender's risk-based pricing charged me points because of my credit profile or loan type," and both get reported in the same field. That's a real limitation of the data, not a reason to distrust the headline finding — the 2%+ tier, where voluntary buydowns are more likely to dominate the mix, tells the more reliable story, and it's consistent across two separate years of data.

Lender credits: the mirror image

If points are "pay more now, pay less over time," lender credits are the reverse: you accept a higher interest rate, and the lender gives you money toward closing costs instead of you paying them out of pocket. Same mechanism as points, opposite direction (Chase).

Our same HMDA sample shows this playing out too, if more subtly: loans with zero lender credit averaged 6.48% interest, while loans that used a lender credit averaged 6.53% — a smaller gap than the points comparison, but in the direction you'd expect (2025 data, same filters as above).

Lender credits make sense when cash at closing is the actual constraint, not the monthly payment — a common, entirely reasonable position, whether you're stretching to hit a down payment or you'd simply rather keep more cash liquid after closing than shave a few hundred dollars off your rate. They tend to make less sense if you're financially able to bring more cash to the table and plan to hold the loan a long time, since the extra interest compounds for as long as you keep the mortgage.

Putting it together

Ask yourself three questions before you pay for points or take a credit:

1. What does the Loan Estimate actually show? Ask your loan officer for the payment and rate at zero points, and at each point increment they offer, side by side. Do the break-even math yourself — don't take their word for "it pays for itself."
2. How long will you realistically keep this loan? Be honest, not aspirational. If there's a decent chance you move or refinance inside the break-even window, points are a discount you'll never redeem.
3. What's the actual constraint — cash today, or payment forever? Tight on cash at closing points toward a lender credit. Flush on cash but want the lowest possible monthly payment for a home you're not going anywhere from points toward buying points, potentially several of them if the math keeps improving.

None of this requires you to be broke or flush to make a smart call — it's a straightforward tradeoff between money now and money over time, and the break-even formula is the same whether you're stretching to close or writing a check for extra points without blinking.

Sources

• Bankrate — What Are Mortgage Points and How Do They Work? (as of August 2026)
• Freddie Mac Primary Mortgage Market Survey — 30-year fixed rate 6.67% as of August 13, 2026
• Chase — Lender Credit, Explained
• CFPB/FFIEC HMDA Data Browser — methodology reference for the public HMDA dataset
• HomebuyingWTF local HMDA database, hmda_2025.sqlite and hmda_2024.sqlite (full public LAR, table lar), queried directly: originated loans (action_taken='1'), purchase loans (loan_purpose='1'), conventional (loan_type='1'), first lien (lien_status='1'), 30-year term (loan_term='360'), owner-occupied primary residence (occupancy_type='1'), non-reverse/closed-end/non-commercial, excluding NA/Exempt codes on discount_points, interest_rate, and lender_credits, grouped by discount points (and separately lender credits) as a share of loan amount.