Trended data: when your balance history starts to count

Two people walk into the same lender with 30% utilization on the same credit limits. One got there by paying down from 70% over the past year. The other drifted up from almost nothing. A FICO 8 score sees two identical ratios. A trended-data score sees two people moving in opposite directions.

For most of the history of credit scoring, the utilization part of your score was a photograph: whatever your cards reported last month. Trended data turns the photograph into a short film, and the newer scoring models are built to watch it.

Snapshot vs. trend: what each kind of score reads

Your card issuers send updates to the bureaus roughly monthly. The bureaus keep that history. The question is how much of it the score uses.

myFICO’s description of FICO’s versions says that scores without trended data often rely on “the most recently reported month of data” for certain components. The old advice to pay your cards down before your statement closes works well on FICO 8 for exactly this reason. The score reacts to the latest report and has no memory of last month’s balance for that calculation.

Trended data adds the months before. According to myFICO, FICO 10T looks at balances and credit limits over “a longer historical time frame (the previous 24 months or longer)” to see whether balances are trending up, down, or staying the same. Experian’s VantageScore 4.0 page lists what that history can show:

• changes in balances
• changes in credit utilization
• payment amounts relative to amounts due
• payment behavior across multiple months

People overlook the third item. If your card issuer reports the amount you actually paid, a trended model can tell the difference between someone who pays $35 on a $35 minimum every month and someone who pays $600.

How far back each model looks

One caution on VantageScore: its own fact sheet says trended data “reflects changes in credit behaviors over time” without naming a month count. The 24-month figure comes from Experian, one of the three bureaus that sell the score. If you see a different number quoted elsewhere, that’s why.

Who is actually using these scores

FICO 8 is still the score myFICO describes as the one most widely used by lenders, so most card and auto decisions are still snapshot-based. Trended data is spreading in mortgages.

According to FHFA’s credit score page, updated October 2, 2026:

• VantageScore 4.0 became available to all Fannie Mae and Freddie Mac approved lenders on September 9, 2026, with no prior approval required. Lenders choose Classic FICO or VantageScore 4.0 for each loan and must use the same model for every borrower on that loan.
• FICO 10T was added as an approved model in April 2026 but is not yet eligible for delivery, and no date has been set.

Separately, Fannie Mae’s Desktop Underwriter has used trended data in its own risk assessment since DU Version 10.0 in June 2016, regardless of which score is on the file. The release notes describe a borrower with low revolving utilization and/or regular payoff of revolving balances as lower risk, and one with high utilization who only makes the minimum payment each month as higher risk. Mortgage credit scores covers how these pieces fit together.

A worked example: same score input today, different stories

Example: two borrowers, each with $10,000 in total card limits.

A snapshot score reading only last month sees 30% for both. A trended model sees Borrower A cutting $4,000 off card debt over two years, and Borrower B adding $2,500 and still climbing. Neither FICO nor VantageScore publishes how many points that difference is worth, so nobody can tell you the gap. The direction is what the model was built to tell apart.

Now give Borrower B a windfall: they pay the balance to $0 the month before applying. A snapshot score treats them as a 0% user. A trended model still sees nearly two years of rising balances followed by one month of zero. That one month helps, but it doesn’t rewrite the pattern.

What changes for you in practice

Steady beats sudden. Under trended models, a balance that falls a little every month builds a better record than a balance that sits high and drops once. If you’re working on your cards anyway, fixed payments above the minimum are worth more than they used to be.

Pay above the minimum when you can. Payment amount relative to the amount due is one of the things trended data captures. Even a modest amount over the minimum separates you from minimum-only payers in that history.

Last-minute paydowns still help. Snapshot scores aren’t going away soon, and paying down before a statement date still lowers what gets reported. Paying down balances before an application covers timing. Just don’t expect a one-month drop to carry the same weight under 10T or VantageScore 4.0.

Consolidating and re-borrowing shows. Moving card debt to a loan and then letting the cards fill up again is exactly the kind of pattern a two-year view picks up. Using a personal loan to pay off cards goes into that risk.

Where to start this month

1. Pull your reports free at AnnualCreditReport.com and look at the balance history on each card, if your report shows it.
2. Pick a monthly payment above the minimum on each card that you can keep making, and set it on autopay.
3. Stop adding new card balances you can’t pay in full that month.
4. If a mortgage is 12 to 24 months away, start now. That window is roughly what trended models read. What changed in credit scoring tracks the rollout.

Trended data can’t see what was never reported. Models can only work with the balances, limits and payment amounts your card issuers actually send to the bureaus.