Why a 20-Point Credit Score Jump Sometimes Changes Nothing

By Kevin Havard | | Market Analysis | 6 min read
Data as of: May 18, 2026

Mortgage pricing moves in tiers, not straight lines. A 40-point credit score jump can save a borrower roughly $26,000 over a $400,000 loan, or it can save nothing at all. The difference is whether the jump crosses a tier boundary.

A borrower whose credit score climbs from 702 to 718 often sees no change in the rate a lender quotes. A borrower who moves from 719 to 720 can shift into a cheaper pricing tier and save thousands over the life of the loan. Same 16-point improvement, very different outcomes. Mortgage pricing does not reward credit score improvement smoothly. It rewards crossing lines drawn by the agencies that buy most loans in the United States.

Fannie Mae and Freddie Mac publish identical pricing grids for purchase mortgages, effective January 28, 2026. Those grids decide what the average conventional borrower pays as a credit-driven add-on. Lenders pass the cost through as a higher rate or as cash at closing. This framing has surfaced in earlier coverage of mortgage pricing, but the practical implication for borrowers deciding whether to delay a purchase is worth stating plainly: improvement inside a tier is mostly noise. Crossing a tier is the lever.

Agency pricing sets the framework; lenders still vary. What follows is the framework, not the final quote.

Credit Pricing Snapshot
680 to 720 FICO move ($400K loan) ~$26,000 lifetime savings Cheaper
780+ FICO ceiling Mostly maxed out Flat
Down payment near 15% (80-85% LTV) Pricing cliff Highest penalty
Cash-out refi penalty below 680 FICO More than 2x penalty Steeper
FHA below 680 FICO Often beats conventional Crossover
MBA Credit Availability (April 2026) 107.9 (-0.4% MoM) Tightening

The Three Rules of Credit-Based Mortgage Pricing

Rule 1: Crossing a tier matters more than improving within one.
The agency grid sorts borrowers into nine credit buckets: 780+, 760-779, 740-759, 720-739, 700-719, 680-699, 660-679, 640-659, and 639 or below. Moves inside a bucket carry no pricing change. Moves across a boundary can shift the price by 25 to 75 basis points in fees, which lenders typically translate into 0.06% to 0.50% on the quoted rate. Time spent improving from 745 to 758 buys nothing. Time spent reaching 760 buys real money. A borrower can spend months improving a score and still receive the same quote if the improvement stays inside a single tier. If your score is already deep inside a tier, delaying a purchase to gain another 10 to 15 points may accomplish nothing.
Rule 2: Down payment changes pricing almost as much as credit score.
The pricing grid is two-dimensional. A 740-759 borrower at 60% loan-to-value pays no credit add-on. The same borrower at 80.01-85% LTV pays 1.000%, or $4,000 on a $400,000 loan. Counter-intuitively, the LTV penalty peaks at 80.01-85% and then declines as LTV climbs higher, because private mortgage insurance absorbs part of the agency's risk above 80% LTV. The 80% threshold is a cliff, not a floor.
Rule 3: Conventional pricing can become punitive below 680.
A 660-679 borrower at 80.01-85% LTV pays 2.125% in credit add-ons, roughly $8,500 on a $400,000 loan. The 639-or-below tier pays 2.875% at the same LTV, roughly $11,500. For borrowers in this range with a low down payment, FHA financing, which prices credit through mortgage insurance rather than a credit-score grid, frequently produces a lower all-in cost. The crossover is real and worth checking before assuming conventional is cheaper.

The Canonical Reference: Fannie and Freddie Purchase Grid

This is the table that decides credit-driven pricing on most conventional purchase mortgages. Each cell shows the credit add-on a borrower pays at a given credit tier and loan-to-value combination. Credit scores in mortgage underwriting are reduced to nine rows; loan-to-value compresses into nine columns. The intersection is the price.

Credit Score≤60% LTV60.01-70%70.01-75%75.01-80%80.01-85%85.01-90%90.01-95%>95%
780+0.000%0.000%0.000%0.375%0.375%0.250%0.250%0.125%
760-7790.000%0.000%0.250%0.625%0.625%0.500%0.500%0.250%
740-7590.000%0.125%0.375%0.875%1.000%0.750%0.625%0.500%
720-7390.000%0.250%0.750%1.250%1.250%1.000%0.875%0.750%
700-7190.000%0.375%0.875%1.375%1.500%1.250%1.125%0.875%
680-6990.000%0.625%1.125%1.750%1.875%1.500%1.375%1.125%
660-6790.000%0.750%1.375%1.875%2.125%1.750%1.625%1.250%
640-6590.000%1.125%1.500%2.250%2.500%2.000%1.875%1.500%
≤6390.125%1.500%2.125%2.750%2.875%2.625%2.250%1.750%
Fannie Mae and Freddie Mac purchase-money credit add-ons, effective 01/28/2026. Values are percent of loan amount. Adjustments are commonly converted to rate at roughly 0.125% rate per 0.500% in fees.

Two things stand out. First, the 60% LTV column is almost entirely zero. Borrowers putting 40% or more down escape credit pricing across nearly every tier. Second, the gaps between rows are not linear. Moving from 700-719 to 720-739 at 80.01-85% LTV saves 0.250%. Moving from 660-679 to 680-699 at the same LTV saves 0.250%. Moving from 640-659 to 660-679 saves 0.375%. The penalty accelerates downward.

What a 40-Point Move Actually Buys

The dollar impact of crossing tiers is easier to see than read. Here is the same $400,000 loan at 80.01-85% LTV, priced across four FICO tiers. The baseline rate is 6.36%, the May 14, 2026 Freddie Mac PMMS reading for the 30-year fixed. The rate impact column reflects the standard conversion from agency credit add-ons to rate, not surveyed rates by tier; actual quotes will vary by lender, day, and loan structure.

FICO TierCredit Add-On$ Cost at ClosingApprox RateMonthly P&Ivs 780+
780+0.375%$1,5006.36%$2,490baseline
740-7591.000%$4,0006.51%$2,531+$41 / mo
720-7391.250%$5,0006.57%$2,547+$57 / mo
680-6991.875%$7,5006.83%$2,619+$129 / mo
640-6592.500%$10,0006.99%$2,664+$174 / mo
Illustrative cost by credit tier, $400,000 conventional loan at 80.01-85% LTV. Rate values reflect the conversion of agency credit add-ons to rate (approximately 0.125% rate per 0.500% in fees) on top of a 6.36% baseline (Freddie Mac PMMS, 2026-05-14). Individual lender quotes will differ.

A borrower with a 685 FICO who reaches 720 saves about $72 a month and roughly $26,000 over 30 years on this loan. A borrower who moves from 725 to 745 saves about $16 a month, or $5,760 over the life of the loan. Same 20-point improvement, very different outcomes, because the first move crosses two tier boundaries and the second crosses one. Targeted credit work aimed at reaching the next tier is worth real money. Generic credit work that improves a score without crossing a boundary is not.

Where Conventional Loses to FHA

Below 680 FICO, the conventional grid gets steep enough that FHA financing often produces a lower all-in cost. FHA does not use the agency credit-add-on grid; it prices credit through an upfront mortgage insurance premium of 1.75% of the loan amount plus annual mortgage insurance at roughly 0.50% to 0.55%. The structure is different. The comparison runs program against program, not score against score.

Scenario ($400K loan, 660 FICO)Conventional 80.01-85% LTVFHA 96.5% LTV
Upfront credit-driven cost2.125% credit add-on ($8,500)1.75% UFMIP ($7,000)
Ongoing insurancePrivate mortgage insurance until 80% LTVAnnual MIP roughly 0.50-0.55% for the life of most FHA loans
Down payment floor~15% in this LTV bucket3.5% with FICO 580+
Illustrative comparison of conventional vs FHA cost structure at 660 FICO. FHA's flat insurance pricing often outperforms conventional credit add-ons below 680, especially at higher LTVs. Lender overlays apply to both programs.
The Cash-Out Penalty: Sub-680 FICO Pays Double
A 660-679 borrower doing a cash-out refinance at 70.01-75% LTV pays 4.000% in credit add-ons. The same borrower doing a purchase at the same FICO/LTV pays 1.875%. More than double. Cash-out refinancing at low credit tiers carries the steepest credit add-ons in the entire agency matrix. Borrowers below 680 considering a cash-out should explicitly model the credit add-on cost before committing.

Outlook and What This Means for Borrowers

The MBA Mortgage Credit Availability Index fell to 107.9 in April 2026, down 0.4% from March after three months of loosening. The conventional component dropped 0.6%, driven by lenders tightening high-LTV and lower-credit programs. If that trend continues, marginal borrowers will see fewer aggressive lender overlays and slower exception approval. The grid itself will not change in the near term; what may change is which lenders are willing to work near the bottom of it.

For borrowers deciding whether to delay a purchase, the practical implication is specific. Look at where the current credit score sits inside the nine-tier grid, then ask how many points stand between it and the next boundary. A 30-point cushion above the next boundary is wasted. A 12-point gap to the next boundary may be worth a few months of targeted credit work. A score that is already at 780 is mostly done; further improvement does not unlock additional savings. For everyone else, the question is not "is my credit good enough?" but "am I one tier away from a meaningfully better quote?"

See whether improving your score would actually change your quote.

Build My Mortgage Plan