Money

Credit scores: what actually moves the number

By Leandro Bruzaferro · · 5 min read
A printed credit report on a desk beside a pen
SharePostLinkedInFacebook

This is informational and not financial advice.

Credit scoring produces more folklore than almost any other area of personal finance, largely because the number changes without explanation and people construct theories to fill the gap. The underlying model is published, the factors are weighted, and the weights are not equal. Knowing which ones carry real weight tells you where effort is worth spending and where it is wasted.

What the score is built from

The most widely used scoring model assigns the following weights, published by the company that produces it.

Factor Weight What it measures
Payment history 35% Whether you paid past accounts on time
Amounts owed 30% How much of your available credit you are using
Length of credit history 15% How long your accounts have been open
Credit mix 10% The variety of account types you hold
New credit 10% How many accounts you have opened recently

Two of those five account for nearly two thirds of the outcome. Payment history and amounts owed are where the score is decided; the remaining three are refinements.

The weights describe general importance across the whole population. Your own profile can shift them: for someone with a very short history, the length factor carries more practical weight than the table suggests, and for someone with a recent missed payment, payment history dominates everything else.

The weight of each factor

Payment history, 35%. Whether accounts were paid as agreed. A single payment reported thirty days late does measurable damage, and the damage is worse the more recent and the more severe the delinquency. This factor rewards nothing except consistency, which makes it simultaneously the most important and the least interesting: there is no technique, only the habit.

Amounts owed, 30%. Chiefly the ratio of balances to credit limits, both overall and per card. This is the factor that responds fastest, because it recalculates whenever a lender reports a new balance. It is also the one people most often influence by accident, since a large purchase paid off in full can still be reported at its peak and depress a score for a month.

Length of credit history, 15%. The age of your oldest account and the average age of all of them. It can only be improved by waiting, which is why closing an old account is usually a poor idea.

Credit mix, 10%. Holding different types, revolving cards and instalment loans, is viewed slightly more favourably than holding one type. The effect is small and not worth taking on debt to engineer.

New credit, 10%. Applications generate hard inquiries. Several in a short window suggest either expansion or difficulty, and the model treats that as elevated risk.

Myths that cost people points

These beliefs are common, wrong, and each one has a real cost.

Belief Reality What it costs you
Checking my own score lowers it Checking your own report is a soft inquiry and does not affect the score People avoid monitoring, so errors go undetected for years
I should carry a balance to build credit Paying in full builds history identically Unnecessary interest at an average card rate above 22%
Closing an old card I don’t use helps It can hurt, by reducing available credit and shortening average account age Two factors worsen at once, worth 45% of the model combined
Income is part of the score Income is not a scoring factor Misplaced confidence that a raise fixes a score
Paying off a collection erases it The account is updated as paid; it does not necessarily vanish Disappointment, and sometimes a worse outcome than negotiating first
All scores are the same number Multiple models and versions exist; lenders use different ones Surprise when a lender quotes a different figure than an app did

The second row is the expensive one. Carrying a balance to “build credit” achieves nothing the model rewards and, at rates currently averaging above twenty-two per cent on accounts assessed interest, costs real money every month to accomplish it.

How fast changes register

Different factors move on different timescales, which explains why improvement feels unpredictable.

Balances update whenever a lender reports, usually monthly, so paying down a card can show up within a billing cycle. That makes utilisation the fastest available lever, and it is the one to use before a mortgage application.

Hard inquiries fade in relevance over months and drop off entirely after a couple of years.

A missed payment takes far longer. It stays on the report for years, though its effect diminishes as it ages and as clean history accumulates behind it.

Account age only improves with time, which cannot be accelerated by any action.

The practical sequence, for someone with a deadline: reduce card balances first, avoid new applications, and leave old accounts open. Those three actions address the two heaviest factors and the one that is easiest to damage inadvertently.

Reading your own report

You are entitled to obtain your credit reports from the major bureaus at no cost through the official channel, and the reports are worth reading rather than glancing at.

Look specifically for accounts you do not recognise, balances that are wrong or stale, payments marked late that were not, and negative items older than the reporting period allows. Errors in these files are not rare, and every one of them is depressing a number that lenders use to price your borrowing.

Disputes go to the bureau, which must investigate. It is unglamorous administrative work, and it is the highest-return activity available in this entire subject, because correcting an error is free and instant compared with the years of behaviour it would otherwise take to offset.

Sources