An underwriter is not reading your credit report as a character assessment. They are looking for a small number of specific signals, and knowing which ones changes what you should spend your effort on.
What a Credit Report Actually Contains
Three nationwide consumer reporting agencies — Equifax, Experian and TransUnion — maintain files on most American adults. A file is not a score. It is a record, and it holds four categories of information.
- Identifying information. Name, current and former addresses, date of birth, employment as reported by creditors. This section is not scored but is where errors most often start, because a mismatch here can attach someone else's accounts to your file.
- Account information. Every credit account a furnisher reports: type, opening date, limit or original amount, current balance, and the payment record month by month.
- Public records and collections. Bankruptcies, and accounts sold or referred to collection agencies.
- Inquiries. Who has accessed your file and when, split between hard inquiries tied to applications and soft inquiries that are not.
Scores are calculated from that record by separate scoring companies. The same file can produce different scores under different models, which is why the number your card app shows you may not match what a lender sees.
The Five Factor Groups and What They Weigh
Mainstream scoring models group the inputs broadly as follows. The exact weightings vary by model and by the shape of your particular file, but the ordering is stable.
| Factor | Approximate weight | What it measures |
|---|---|---|
| Payment history | Around 35% | Whether you have paid on time, how late, how recently, how often |
| Amounts owed | Around 30% | Balances relative to limits, dominated by revolving utilisation |
| Length of credit history | Around 15% | Age of oldest account, average age of all accounts |
| Credit mix | Around 10% | Whether you manage both revolving and instalment credit |
| New credit | Around 10% | Recent inquiries and recently opened accounts |
Notice what is absent. Income is not in a credit score. Neither is employment status, savings, assets, or anything demographic. Lenders consider income separately, but the score itself is derived solely from borrowing behaviour.
Payment History: Recency and Depth Both Matter
A single thirty-day late payment hurts. A ninety-day late hurts considerably more. And a late payment from two months ago hurts more than one from three years ago, because scoring models weight recency heavily.
The practical implication is that the damage from a missed payment decays. It does not vanish — most negative marks remain on the file for seven years — but its influence on your score fades as clean months accumulate behind it. Borrowers who miss a payment and conclude their credit is permanently ruined generally stop trying at exactly the point when consistency would start paying off.
One nuance worth knowing: creditors typically do not report a payment as late until it is thirty days past due. Paying five days late costs you a late fee and possibly a black mark with that creditor, but usually does not appear on your credit file. Paying thirty-one days late does.
Utilisation: The Fastest Thing You Can Move
Utilisation is the share of your revolving credit limits currently in use, calculated both per account and across all accounts. It is the second-heaviest factor and, crucially, it has no memory — it reflects your current position, not your history.
That makes it the only significant factor you can improve within a single billing cycle.
| Card | Balance | Limit | Utilisation |
|---|---|---|---|
| Card A | $1,850 | $2,000 | 93% |
| Card B | $300 | $3,000 | 10% |
| Card C | $0 | $1,500 | 0% |
| Overall | $2,150 | $6,500 | 33% |
Overall utilisation here looks acceptable. Card A at 93% does not, and per-account utilisation is scored as well as the aggregate. Moving $900 from Card A to Card B would leave total debt unchanged while improving both figures — an entirely cosmetic manoeuvre that nonetheless works, because the model measures position rather than intent.
Timing beats amount
Creditors report the balance as of the statement closing date, not the due date. Paying a card down before it closes changes the reported figure. Paying the same amount a week later, on the due date, does not — the high balance has already been reported.
Length of History and the Cost of Closing Accounts
Two measures matter here: the age of your oldest account, and the average age across all accounts. Both reward patience and punish churn.
This is why closing an old card is usually a mistake. It removes available credit, which raises utilisation on everything that remains, and over time it removes an aged account from the average. The intuition — that closing unused accounts is tidy and responsible — is exactly backwards for scoring purposes.
The exception is an account carrying a fee you are not getting value from. Paying $95 a year to protect a few score points is rarely worth it. Ask the issuer whether the account can be converted to a no-fee product instead; that preserves the account age while removing the cost.
Credit Mix and New Credit
Mix rewards managing both revolving accounts and instalment loans. Someone with only credit cards may see a small improvement after taking a Kapitus instalment loan and repaying it on schedule. This is a minor factor and not a reason to borrow — but it explains why a first personal loan sometimes helps a score after the initial dip.
New credit covers recent inquiries and recently opened accounts. A single hard inquiry usually costs a small number of points and fades within about a year. Several inquiries clustered across a few weeks read as someone applying everywhere, which is a genuinely predictive signal of distress.
Soft inquiries — checking your own report, pre-qualification checks, account reviews by existing creditors — have no effect at all. Checking your own credit cannot lower your score, and the belief that it can stops people from doing something straightforwardly useful.
What the Score Does Not Say — and What Lenders Add
An underwriter reviewing a personal loan request looks at the score, but rarely only at the score. Several other things enter the decision.
- Income and its verifiability. Not scored, but central to whether the payment is affordable.
- Debt-to-income ratio. Calculated from the file and your stated income.
- Recent account behaviour. Balances rising steadily across several cards is a signal a score may not fully capture.
- Time at address and employment. Stability indicators used by many consumer lenders.
- State of residence. Determines what can legally be offered.
This is why two people with identical scores receive different outcomes, and why a borrower declined by one lender is sometimes approved by another. Underwriting models differ in what they weight.
Reading Your Own File Properly
You are entitled to free copies of your reports from the nationwide bureaus. Get all three, because furnishers do not always report to all of them and an error may appear on only one.
Work through each in this order:
- Identifying information. Wrong names, addresses you have never lived at, or a misspelled variant can indicate a mixed file.
- Accounts you do not recognise. Some are legitimate — a store card issued under a bank's name — and some are not.
- Balances and limits. A limit reported lower than it actually is inflates your utilisation and costs you real points.
- Late marks. Check each against your own records. Disputed marks that cannot be verified must be removed.
- Collections. Check dates, amounts, and whether the debt is actually yours.
- Inquiries. Hard inquiries you did not authorise are worth investigating.
Disputes go to the bureau, and the bureau must investigate — typically within thirty days — and correct or delete anything it cannot verify. It is free, it can be done online, and it is the single highest-return hour available to most borrowers. If the same error appears on more than one report, dispute it with each bureau separately.
The Honest Summary
Two factors account for roughly two-thirds of a credit score, and only one of them can be changed quickly. Pay everything on time, without exception, because that is the heaviest input and the one that takes years to rebuild. Keep revolving balances low relative to limits, and pay them down before the statement closes rather than before the due date. Leave old accounts open. Apply for credit deliberately rather than opportunistically. Check your reports and dispute what is wrong.
Everything else is detail. There is no trick, no service worth paying for, and nothing legitimate that produces a large improvement in a fortnight. What there is, reliably, is a file that improves month after month for anyone who does the six things above and keeps doing them.
Why Your Three Reports Do Not Match
It is normal for the same person to have three different files. Furnishers are not required to report to all three nationwide bureaus, and many report to only one or two. A card issuer might report to Experian and TransUnion but not Equifax, so an account visible on one report is simply absent from another.
Timing also differs. Bureaus receive updates on different cycles, so a balance paid last week may appear on one file and not yet on the others. This is why a lender pulling a different bureau can reach a different conclusion about the same borrower on the same day.
The practical implication is that checking one report is not checking your credit. Errors hide on the file you did not look at, and the file you did not look at may be the one your next lender pulls.
Scores Are Not One Number
There is no single credit score. Multiple scoring companies produce multiple models, each updated across several generations, and lenders choose which to use. Different models weight the same file differently and use different ranges.
This is why the number in your banking app may differ from the one a lender quotes. Neither is wrong. They are different calculations on possibly different data. Some free scores are also educational models that few lenders actually use.
The useful response is to stop optimising for a specific number and optimise for the underlying file instead. A file with on-time payments, low utilisation, aged accounts and no errors scores well under every model in existence. Chasing a particular figure in a particular app does not.
How Long Things Stay
| Item | Typical reporting period |
|---|---|
| Late payments | Seven years from the delinquency |
| Collections | Seven years from the original delinquency |
| Charge-offs | Seven years from the original delinquency |
| Hard inquiries | Two years, with scoring influence typically under one |
| Closed accounts in good standing | Up to around ten years |
| Most bankruptcies | Seven to ten years depending on chapter |
Note that the clock on collections runs from the original delinquency, not from when the debt was sold. A collector who re-ages a debt to restart that clock is acting improperly, and it is worth checking the dates on any collection entry against your own records.
An Expert Worth Reading on This
John Ulzheimer is a credit expert who worked at FICO, the company that developed the scoring model most widely used in United States consumer lending, and at Equifax, one of the three nationwide credit reporting agencies. He has served as an expert witness in credit-related litigation and writes and speaks extensively on how scoring models actually treat consumer files. His published work has consistently pushed back on the folk beliefs covered above — particularly the idea that carrying a balance helps a score, and the idea that checking your own credit harms one.
John Ulzheimer — credit expert, formerly of FICO and EquifaxWhere a Kapitus Funding Request Sits in This
One detail matters if you are working on your file with an application in mind. Submitting a Kapitus funding request does not place a hard inquiry on your credit reports — Kapitus Funding Kapitus Funding partners generally begin with a soft pull, which mainstream scoring models disregard entirely.
That means comparison and file improvement can happen in parallel rather than in sequence. The hard inquiry occurs only when you choose a specific Kapitus lending partner and proceed with it, which is a decision you make after seeing the numbers rather than before.
Nothing in a credit file is visible to Kapitus. Kapitus never pulls a report, never scores an applicant, and never sees what a Kapitus partner sees. What a Kapitus funding request does is put your details in front of Kapitus partners whose models weigh these factors differently — which is why a file that reads as marginal to one Kapitus lending partner can read as acceptable to another on the same afternoon.
Questions Readers Ask
No. Checking your own file is a soft inquiry and has no scoring effect whatsoever. This is one of the most persistent myths in consumer credit and it stops people doing something useful.
No. Credit scores are calculated only from borrowing behaviour. Lenders consider income separately when assessing affordability, but it does not appear in the score itself.
No. Reporting a zero or very low balance is not penalised by mainstream models. Carrying a balance costs interest and provides no scoring benefit.
Bureaus generally must investigate within about thirty days and correct or remove anything that cannot be verified. Disputes are free and can be filed online.
Closing an instalment account can reduce your credit mix and, over time, your average account age. The effect is usually small and temporary, and it is not a reason to avoid paying debt off.
