“Points” sound like a casino term. They’re just prepaid interest. That reframing does most of the work in understanding them: when you buy points on a mortgage, you’re handing the lender cash on day one in exchange for a lower rate for the life of the loan. Sometimes that trade is excellent. Sometimes it’s the most expensive coffee you’ll ever buy. The difference is one number – the break-even – and almost nobody computes it before signing. Let’s compute it.
What a point actually is
One point costs 1% of the loan amount. On a $300,000 loan, a point is $3,000. In exchange, the lender lowers your rate – historically by around 0.125% to 0.25% per point, though the exact discount moves with the market and varies by lender, so treat any quoted ratio as negotiable and verify it. Points paid at closing on a purchase are often tax-deductible (mortgage interest deduction rules apply, and I’m not a tax professional – confirm with one).
The mechanic is simple. You’re converting liquid cash today into a slightly smaller payment for the next 30 years. The question is only: when does the smaller payment catch up to the cash you gave up?
The break-even math
Here’s the worked example, fully checkable. Take a $300,000 loan at 6.75% over 30 years. The payment is about $1,946 a month (300 x $6.486, the standard per-$1,000 factor). Now buy one point for $3,000 and drop the rate to 6.5%: the payment falls to about $1,896 (300 x $6.321). The savings are $50 a month, and the point cost $3,000. Break-even: 3,000 / 50 = 60 months – five years.
But the discount you’re offered matters enormously. Here’s the same loan where the point only buys 0.125% (a 6.625% rate):
| Scenario | Monthly payment | Savings/mo | Break-even |
|---|---|---|---|
| 6.75%, no points | $1,946 | – | – |
| 1 point ($3,000) for 0.25% off | $1,896 | $50 | ~5 years |
| 1 point ($3,000) for 0.125% off | $1,921 | $25 | ~10 years |
Look at that second row again. The same $3,000, buying a smaller discount, takes ten years to break even – and the average homeowner moves or refinances well before ten years. This is why “just always buy points” is bad advice and “never buy points” is bad advice. The answer is arithmetic, and the two inputs are the point cost and the payment delta your lender actually offers. Make them write both down.
Three things kill the points trade
- Moving early. Every year you stay past break-even, the points win. Every year before it, they lose. If there’s a real chance you relocate within five years – job mobility, family plans, a starter home you’ll outgrow – points are usually the wrong purchase.
- Refinancing later. If rates fall and you refinance in three years, you paid $3,000 for a benefit you only used for 36 months, and you’ll pay closing costs all over again on the new loan. The same break-even logic applies to refi decisions – I break that down separately in the refinance break-even post.
- Opportunity cost of the cash. $3,000 is also a bigger emergency fund, or a down-payment top-up, or invested money. If buying the point drains the account you’d want liquid in year one of a mortgage, that’s a hidden cost. Points make sense with money you won’t need, in a home you won’t leave.
Seller credits and the 2-1 buydown question
Now the version that can be free money. In a slower market, sellers sometimes agree to pay closing costs or points for you – a “seller credit.” A 2-1 buydown is a structure where the rate is reduced by 2% in year one and 1% in year two, then returns to the note rate. Whether that’s good depends entirely on who pays for it:
- If the seller funds it as part of a negotiated deal, it’s a genuine discount – the money was coming out of their pocket instead of a price cut, and your cost is roughly nil. Take it.
- If you fund it – or if the “free” buydown is baked into a higher price or a higher rate elsewhere – you’re prepaying interest for a temporary discount. That’s often worse than plain points, because the benefit expires after two years while a bought point lowers the rate forever.
This is what I mean when I call some 2-1 buydowns “theater.” They show up in marketing as a gift from the lender, but the money comes from somewhere. The test is always the same: what does the total package cost compared to the same loan priced without the gimmick? Ask your lender to price the loan both ways, in writing. A lender who won’t show you both versions is telling you something.
The decision rule
Here’s the whole post compressed into a rule: buy points only if the break-even is meaningfully shorter than your honest expected stay in the home. If you’d plausibly move or refinance inside five years, keep the cash. If you’re a ten-year-plus hold-everything household and the lender’s discount ratio is decent (0.25% per point, not 0.125%), points can be one of the better guaranteed returns available in the deal – a $3,000 payment that saves $50 a month for 30 years totals $18,000 of savings on a $3,000 spend, if you actually stay the whole term. That “if” is the entire game.
A few practical notes. First, points and rate are always negotiable together – you can also take a slightly higher rate for a lender credit that offsets closing costs, which is the mirror-image trade and often the better choice for cash-tight buyers. Second, this math changes with rates; at higher rates the per-point discounts tend to be richer, at lower rates thinner, so the break-even from a blog post (including this one) is never a substitute for your loan estimate. Third, none of this is financial advice – it’s the framework, and your loan officer’s loan estimate is the data.
Next step: before you sign anything, ask for two versions of your loan estimate – one with points, one without – and compute the break-even exactly like the table above. If the lender’s discount is thin, the table will show it in one line. And if you’re weighing the bigger purchase decision first, start with whether to buy while prices are high or the full cost stack in the house hacking math post.
