Why Your Score Moved When You Didn't Do Anything

You checked in March and saw 712. You check in April and see 689. Nothing happened — no missed payment, no new account, no unusual spending. The instinct is that something must be wrong. Usually nothing is. A score is a model run against a file that updates on schedules you don't control, using balances captured on dates you didn't choose, from creditors reporting whenever they report. Most month-to-month movement reflects when things were measured rather than anything about your behavior. This guide separates the movement that's noise from the movement that's signal — and explains the specific cases where doing something sensible makes the number go down.

A score is a snapshot, not a state

The mental model most people carry is that a score measures their creditworthiness, which changes when their behavior changes. The accurate model is narrower and explains everything else in this guide:

A score is one model, applied to one bureau's file, as it existed at one moment.

Change any of the three and the number changes without anything about you changing:

• The model. Many exist, in multiple versions, with different scales and different weightings.
• The bureau. Three files, not identical, because not every creditor reports to all three — a substantial share of tradelines report to fewer than three.
• The moment. Files update continuously as creditors send data on their own cycles.

Which means the useful question is almost never "why did my score change" but "what changed on the file the score was calculated from?" That's answerable, and it's what the rest of this covers.

The statement date problem

The single largest cause of unexplained movement, and it's almost entirely mechanical.

Credit card issuers report the balance as of your statement closing date — not the balance after you pay, and not your current balance. If your statement closes on the 8th and you pay in full on the 22nd, the file shows the 8th's balance for the entire month.

The consequence: a person who pays in full every month and has never carried a balance still shows utilization on their credit file, and that reported utilization varies with ordinary spending.

Worked example. A $10,000 total limit across cards.

Month	Spending before statement close	Reported utilization	Effect
January	$800	8%	Favorable
February	$1,100	11%	Slightly less favorable
March	$3,400 — annual insurance and a repair	34%	Meaningful drag
April	$900	9%	Recovered

Every balance was paid in full. Nothing was ever carried. The March score drop was real, and it reflected a timing artifact rather than any change in financial condition. It also reversed the following month without any action.

Two practical implications:

• If you need a strong score on a specific date — a mortgage application — pay balances down before the statement closes rather than before the due date. That single change moves reported utilization directly, and it's the highest-leverage short-term lever available, per our utilization guide.
• Otherwise, ignore it. Month-to-month utilization noise on a file with no delinquencies is not worth managing.

Three moves that lower your score

Sensible financial actions with counterintuitive scoring effects. Worth understanding not to avoid them, but so the drop doesn't alarm you.

Paying off and closing a card. The account's limit leaves your available credit, so utilization on remaining balances rises. Eliminating a $6,000 limit while carrying $2,000 elsewhere moves you from $2,000 of $16,000 (12.5%) to $2,000 of $10,000 (20%) — you owe less and score worse. If the card was old, you also eventually lose its age contribution. Our closure guide covers when to close anyway.

Paying off an installment loan. Ending an auto loan or personal loan closes an active account and can reduce the mix of account types. The effect is usually modest and temporary, and it is not a reason to keep debt you can retire — but if you're applying for something in the next few months, it's worth sequencing.

Being removed as an authorized user. If you were on someone else's account, its full history — age, limit, payment record — was in your file. Removal takes all of it out at once, which can be a large single-month move for someone whose file depended on it. The mechanics are in our authorized user guide.

The general lesson: scoring models measure the file, not your finances. Actions that improve your finances while shrinking your file can move the number down, and that's a limitation of the measure rather than a reason to make worse decisions.

Changes you didn't cause

Things that move your file without any action by you:

• Negative items aging off. A late payment or collection reaching the end of its reporting period disappears, and the score improves — with no action at all. This is the most common cause of an unexplained increase.
• Accounts aging. Average age rises passively each month.
• Inquiries aging out of scoring consideration.
• A creditor changing your limit. An issuer raising your limit lowers utilization; one reducing it raises utilization instantly — and reductions happen without notice in many cases.
• A creditor changing its reporting. Some report balances differently, some stop reporting limits, some change reporting frequency. Any of these moves the calculation.
• A closed account falling off entirely after its retention period, removing its age contribution — which can cause a drop years after you closed it.
• Joint account activity. A co-borrower's behavior on a shared obligation appears on your file, per our joint finances guide.
• Bureau data corrections, in either direction.
• Scoring model updates, where a provider changes the version they show you.

Reading that list, the point is clear enough: a substantial share of score movement originates with decisions made by creditors and bureaus rather than by you. That's not a flaw to be fixed — it's what it means for a score to summarize a file that other parties write.

Why you have several different scores

The question underneath most confusion about score movement.

There isn't one score. There are many models, several versions of each, three bureaus, and industry-specific variants used for particular products. A number from a card issuer's free tool, a number from a credit monitoring app, and the number a mortgage lender pulls can all differ substantially on the same day, for the same person, with no error anywhere.

What follows practically:

• Compare a score only against itself over time, from the same source. Comparing across sources tells you nothing.
• A change in the number you're shown may be a change in the model, not the file — providers update versions.
• Mortgage lenders frequently use older model versions than consumer-facing tools, which is why the number at application can surprise people.
• The underlying file is the real object. Two people with the same score can have very different files, and lenders look at both.
• Auto and card lenders may use industry-specific versions weighted toward their own product's history.

The reframe worth adopting: the score is a summary statistic, and the file is the thing. A file with no delinquencies, low reported utilization, and reasonable age produces good numbers across every model — which makes managing the file a more reliable strategy than chasing any particular score.

Telling noise from signal

Movement	Likely cause	Action
A few to ~20 points, either way	Reported balance timing	None
~20–40 points down	A large reported balance, a new account, a hard inquiry, or a closed card	Identify it; usually resolves
Sudden large drop	Newly reported delinquency, collection, or public record	Investigate now
Unexplained large drop	Possibly an account you didn't open	Pull all three reports immediately
Gradual rise	Aging, negative items expiring, utilization declining	Nothing — it's working
Sudden large rise	A major negative aged off, or a correction posted	Verify it's accurate

The heuristic worth carrying: the size of a movement tells you roughly which category it's in. Small movements are almost always the reporting mechanics described above. Large sudden movements are almost always a specific discrete event, and discrete events have identifiable causes on the report.

When to investigate

Investigate a drop when it's large, sudden, and you can't name the cause. The sequence:

1. Pull all three reports, not just the one your score came from. The change may be on a file you haven't looked at.
2. Compare against your last pull, if you have one. Line-by-line comparison finds the change fast.
3. Look for anything new — accounts, inquiries, collections, public records.
4. Look for anything changed — a limit reduced, a status updated, a balance that looks wrong.
5. Check accounts you don't recognize. This is the reason large unexplained drops warrant urgency rather than patience, and the response is in our identity theft guide.
6. Dispute anything inaccurate, following our error correction guide.
7. Freeze your files if you find unauthorized activity.

And the case where speed genuinely matters: a new account you didn't open is the one score change you should treat as urgent. Everything else on this list can wait a week.

What you actually control

Ranked by effect over the horizon that matters:

• Paying on time, every time. Dominates everything else, and it's the input with the least noise — per our delinquency guide.
• Keeping reported utilization low, which means managing against statement dates rather than due dates.
• Keeping old accounts open, since age accrues only with time and is destroyed instantly.
• Applying for credit deliberately rather than opportunistically.
• Adding positive tradelines where your file is thin, per our file building guide.
• Checking your reports periodically so errors are caught while they're correctable.
• Not reacting to monthly noise, which is a real discipline — chasing month-to-month movement produces worse decisions than ignoring it.

The framing that holds up: the score is an output. The file is the input. Manage the input and the output takes care of itself over any horizon that matters — and month-to-month is not a horizon that matters.

Frequently asked questions

---

This guide is educational and does not constitute financial advice. Scoring models, their weightings, and the reporting practices of individual creditors vary and change; worked examples are stylized illustrations of the mechanism rather than predictions of any specific score movement.