Credit scores govern the price of borrowing for most American households, and the industry that produces them has been deliberately opaque for decades. It is less opaque than it used to be. The broad structure of the dominant scoring models is published, the underlying data is available free to consumers, and most of what circulates as advice is either outdated or invented.
The five factors, weighted
Payment history is the largest single component, accounting for roughly a third of a typical score. A single payment reported thirty days late can reduce a strong score substantially, and the damage decays slowly. Amounts owed, dominated by credit utilisation — balances as a percentage of available limits — is the second largest and the most volatile, because it updates monthly.
Read moreIndex Funds Now Own Wall Street. What Changes
Length of credit history, credit mix and recent applications for new credit make up the remainder. None of these is trivial, but none moves a score as fast as a missed payment or a spike in utilisation. Consumers are entitled to free reports from each nationwide bureau, and the official channel is described by the Consumer Financial Protection Bureau.
Utilisation is measured at a snapshot
The most useful practical detail about scoring is that utilisation is calculated from the balance reported on the statement date, not from average balance or from whether the bill is paid in full. Someone who charges heavily and pays in full every month can still show high utilisation, because the reported figure is the statement balance. Paying down before the statement closes, or requesting a higher limit, lowers reported utilisation without changing spending at all.
Closing an old unused card is often counterproductive for the same reason: it removes available limit, raising utilisation, and eventually shortens average account age. The advice to close cards you do not use is one of the most persistent pieces of harmful folklore in the category.
Read moreWhy State Tax Competition Reshapes Business
What is not in the score
Income is not an input. Neither is employment, savings, net worth, marital status or the interest rate being paid. Checking your own report is a soft inquiry with no effect. Applying for credit produces a hard inquiry with a small and temporary effect, and multiple mortgage or auto applications within a short window are treated as one inquiry precisely so consumers can shop rates.
Rent and utility payments have historically been absent unless delinquent, though several programmes now allow positive reporting on request. Medical collection reporting has been narrowed considerably following regulatory and industry changes, which materially improved scores for a large number of consumers.
Errors, disputes and rights
Report errors are common enough that reviewing all three bureau files annually is worth the time. The dispute process, and the obligations of both the bureau and the furnisher of the data, are established under federal law and explained by the Federal Trade Commission. A dispute must be investigated within a defined period, and unverifiable items must be removed.
Read moreHow Public Companies Actually Set Their Prices
Security freezes are free and are the strongest available protection against new-account identity fraud. Placing and lifting a freeze at each bureau takes minutes online. Broader analysis of household credit conditions, including delinquency trends by loan type, is published quarterly by the Federal Reserve Bank of New York, and it is a better guide to where consumer stress is building than any individual score.
What the score cannot see
A credit score is a prediction of delinquency built from a lender-reported file, and everything outside that file is invisible to it. Income does not appear. Assets do not appear. Rent paid on time for a decade generally does not appear unless the landlord or a rent-reporting service furnishes it, which is why thin-file consumers with impeccable payment habits can be scored as risky simply because there is too little data to model. Employment history, education, and savings balances are likewise absent, and this is a deliberate design choice rather than an oversight, because the file is restricted to what furnishers report under the Fair Credit Reporting Act framework that governs consumer reporting.
The practical consequence is that two borrowers with identical scores can carry very different real risk, and lenders know it. That is why mortgage underwriting layers debt-to-income ratios, reserve requirements, and documentation on top of the score rather than treating the number as sufficient. The score is a filter and a price input, not a verdict.
The levers that actually move the number
For most consumers, three actions dominate everything else. Pay every account on time, because payment history is the single heaviest factor and a thirty-day late mark can cost a good file a substantial number of points for years. Keep revolving utilisation low, ideally in the single digits at the moment the file is pulled, since utilisation is measured on the statement balance rather than on annual spending. And leave old accounts open, because average age of accounts rewards patience and closing a long-held card can shorten the file and raise utilisation at the same time. Everything else, including the rate-shopping windows and the mix-of-credit factor, is comparatively marginal.
