If you've been browsing lender rate sheets this week, you've probably seen a 5.75% staring back at you in bold — with a footnote that reads "1.5 discount points required." On a $400,000 loan, that footnote costs you $6,000 before you move a single box into the house. The 30-year fixed currently sits at 6.22% according to Freddie Mac's Primary Mortgage Market Survey (week of March 19, 2026) — and lenders are actively pitching rate buydowns as a way to close the gap. Sometimes they're right. Often they're not. Here's the actual math, and a 2026-specific reason to be more skeptical than usual.
What Discount Points Are (and Aren't)
One discount point equals 1% of your loan amount, paid upfront at closing in exchange for a lower interest rate. On a $400,000 loan, one point costs $4,000. The rate reduction you get per point varies by lender — typically somewhere between 0.20% and 0.30% — and it is not standardized. Two lenders offering the same buydown can price it very differently.
It's also worth distinguishing discount points from origination points. Origination points are a lender fee — they compensate the lender for processing your loan and do not reduce your rate. If a lender quotes you "2 points," ask them point-blank: how much of that is discount versus origination? The CFPB explains lender credits and discount points here — it's a two-minute read worth doing before any points conversation.
According to a CFPB Data Spotlight on discount points, roughly 9 in 10 borrowers who took cash-out refinances during the 2022–2023 peak-rate environment paid discount points. The CFPB flagged this as a consumer protection concern — the complexity of the points tradeoff consistently leads to confusion, and borrowers with lower credit scores were disproportionately represented in the group paying points. This is worth knowing: points aren't always pitched because they're in your interest.
The Breakeven Math at 6.22%
Let's run the numbers on a $400,000 loan at today's Freddie Mac PMMS rate of 6.22%. The baseline monthly principal and interest payment is approximately $2,451. Here's what happens when you buy down the rate:
| Scenario | Rate | Monthly P&I | Upfront Cost | Monthly Savings | Breakeven |
|---|---|---|---|---|---|
| No points (baseline) | 6.22% | ~$2,451 | $0 | — | — |
| 1 discount point | ~5.97% | ~$2,384 | $4,000 | ~$67/mo | ~60 months (5 yrs) |
| 2 discount points | ~5.72% | ~$2,318 | $8,000 | ~$133/mo | ~60 months (5 yrs) |
Notice something counterintuitive: the breakeven period is essentially the same whether you buy one point or two. More points get you more monthly savings, but proportionally more upfront cost. The math is linear, which means adding points doesn't accelerate your breakeven — it just amplifies both sides of the equation.
The most important number here: the breakeven clock resets to zero if you refinance or sell. If you pay $8,000 in points and refinance 36 months later (well before month 60), you've recovered $4,788 in monthly savings but you're still $3,212 in the hole — and that money is simply gone. According to the Freddie Mac PMMS historical data, borrowers have historically refinanced within 5–7 years during rate-declining cycles. In today's environment, that's exactly the cycle we're entering.
The 2026 Wrinkle: Spread Compression Risk
Here's where the standard breakeven calculator misses something important — and it's specific to where we are in March 2026.
The 30-year fixed mortgage rate doesn't just track the Federal Reserve's benchmark rate. As we've covered in our breakdown of what actually sets your mortgage rate, the 30-year fixed is anchored to the 10-year Treasury yield plus a spread that reflects MBS market conditions and investor risk appetite. Right now, the 10-year Treasury yield sits at 4.25% (FRED series DGS10). With the 30-year fixed at 6.22%, that spread is 197 basis points — well above the historical norm of 150–170 bps.
That excess spread doesn't require Fed action to compress. Oil prices stabilizing, macro uncertainty clearing, or a mild improvement in MBS demand can do it independently. Run the scenario:
"If the 30yr/10yr spread normalizes to 170 basis points while the 10-year Treasury holds at 4.25%, the 30-year mortgage rate falls to approximately 5.95% — naturally, with no rate cut from the Fed."
— Derived from FRED DGS10 (4.25%, March 2026) + historical PMMS spread analysis
At 5.95%, you'd save roughly $70/month on a $400,000 loan compared to 6.22% — and you'd have spent zero dollars on points to get there. Compare that to paying two points today ($8,000) to buy down to 5.72%. You save $133/month, but if the spread normalizes to 5.95% within 18–24 months, you'd refinance and reset your amortization clock. Your $8,000 investment would have generated maybe $3,200 in savings before the refinance erased the breakeven. That's a losing trade.
When Discount Points Actually Make Sense
We're not saying points are always bad. They work — in specific circumstances:
- Definite long-term hold (10+ years): If this is a forever home and refinancing isn't on the table, the lifetime interest savings on a 2-point buydown can exceed $47,000 on a $400K loan. Over 30 years, the math is compelling. The problem is that "forever" rarely means forever.
- Seller-paid points: In a buyer's market, sellers sometimes offer to cover points as a concession. That's free rate reduction — always worth accepting. Ask your agent about seller concessions before dismissing points entirely.
- Rates are clearly near a structural ceiling: If macro conditions point to rates staying elevated or rising — not the case in March 2026 — the risk of the natural rate decline disappears and the breakeven math improves.
In today's environment specifically, the combination of an elevated spread and a Fed that has already signaled two potential cuts in its latest dot plot makes the "definite long-term hold" condition harder to bet on. The rate education case for points in 2026 is weaker than it would be in a stable, structurally-high-rate environment.
What to Do Before You Agree to Any Points
Here are four concrete steps before signing anything that includes discount points:
- Always request the zero-point rate first. Every lender must provide a Loan Estimate within 3 business days of application. Ask explicitly for a zero-point quote. Then compare the APR — not just the interest rate — between the zero-point and points scenarios. APR captures the true cost including upfront fees.
- Run your personal breakeven, not the rule of thumb. The "5–7 years" estimate above is based on $400,000 — your loan size will shift the number. Larger loans mean larger point costs but also larger monthly savings, so the breakeven timeline stays roughly similar. But your confidence in your hold period is the real variable.
- Ask about seller concessions. In this market, sellers have more motivation to help. Seller-paid points of 1–2% as part of a purchase offer negotiation is a legitimate and often underutilized strategy.
- Watch the 10-year Treasury and the PMMS spread. You can track the 10-year at FRED series DGS10 and the 30-year mortgage rate at FRED MORTGAGE30US. If the spread starts compressing from 197 toward 150 bps, natural rate improvement is coming — and points look worse with every step of that compression.
The bottom line is that discount points are not inherently good or bad — they're a prepayment of interest, and like any prepayment, the return depends entirely on how long you keep the loan. In a market with an elevated spread and a rate cycle pointing toward eventual improvement, the burden of proof for paying points is higher than usual. Don't let an advertised rate seduce you into writing a $4,000 or $8,000 check before you've done this math yourself.
