No File, No Score, No Loan, No Way to Start One
Tens of millions of American adults have no credit file or too thin a file to generate a score. The system that decides whether they can borrow requires a history of borrowing they were never able to start.
Three Different Situations
Public discussion collapses several distinct conditions into the phrase bad credit, and separating them is the whole subject.
Credit invisible means no file exists at the major credit bureaus at all. Nothing has ever been reported.
Unscorable means a file exists but contains too little information, or nothing recent enough, to generate a score under conventional models. A single account opened last month, or several accounts last active a decade ago, produce this outcome.
Subprime means a score exists and it is low, reflecting a record of missed payments or high utilisation.
The third is a judgement about behaviour. The first two are an absence of information, and a lender declining them is not saying the applicant is risky. It is saying it does not know.
| Status | What the Lender Sees | Typical Outcome |
|---|---|---|
| Credit invisible | No file | Automatic decline in most systems |
| Unscorable | Insufficient or stale data | Manual review or decline |
| Subprime | A low score | Approved at a high rate |
Who It Affects
Research by consumer financial regulators has estimated that roughly one in ten American adults is credit invisible, with a further several percent unscorable, together amounting to tens of millions of people.
The distribution is not random. Rates are substantially higher among young adults, recent immigrants, residents of low income neighbourhoods, and Black and Hispanic consumers. That concentration is what turns a technical data problem into a policy question, because the absence of a file compounds every other disadvantage.
A thin file is not a fact about a person, it is a fact about which of their financial behaviours somebody chose to report. Paying rent for fifteen years produces no record. Missing two credit card payments produces a permanent one.
The Reporting Asymmetry
The credit system records a specific and narrow set of behaviours: loans, credit cards, and other extensions of credit reported by furnishers who choose to participate.
It has historically excluded the largest and most regular payments most households make. Rent is the biggest monthly obligation for tens of millions of people and is generally not reported. Utilities, telecommunications, and insurance premiums are typically reported only when they go badly wrong and are sent to collections.
The asymmetry is stark. Paying rent on time for a decade generates no positive record; failing to pay a utility bill generates a collection item that damages a score for years. The system captures failure in categories where it does not capture success.
What Has Been Tried
Rent reporting services report on time rent payments to bureaus, either through a landlord or through a third party that verifies payment. Studies have found meaningful score improvements and, importantly, that many previously unscorable consumers become scorable. Coverage remains limited because most landlords have no incentive to participate.
Alternative data in underwriting incorporates cash flow information from bank accounts, showing income regularity and spending patterns, rather than relying solely on bureau data. Several lenders have adopted this and regulators have issued guidance describing how it can be used consistently with fair lending obligations.
Consumer permissioned data lets an applicant grant a lender access to their bank account data directly, which is particularly useful for someone with steady income and no borrowing history.
Utility and telecom reporting programmes allow consumers to opt in to having those payments reported.
The Fair Lending Complication
Alternative data is not automatically an improvement, and the caution is genuine.
Any variable used in underwriting must not produce unlawful discrimination, and variables correlated with protected characteristics can produce disparate impact even without any intent. Data such as educational institution attended, shopping patterns, or device information has been challenged on exactly this basis.
Cash flow data is generally viewed more favourably because it measures the thing underwriting is actually trying to assess, which is capacity to repay, rather than serving as a proxy for it.
There is also an explainability requirement. Lenders must provide specific reasons for adverse decisions, which constrains the use of models whose outputs cannot be attributed to identifiable factors.
The Products That Bridge the Gap
Several products exist specifically to create a file where none exists.
A secured credit card requires a cash deposit that becomes the credit limit, so the lender takes no credit risk while the account reports normally.
A credit builder loan inverts a loan entirely: the borrowed amount is held in a locked account, the borrower makes payments, and receives the money at the end. It is a savings plan that generates a payment history.
Both work. Both require some cash upfront, which is the constraint for exactly the population that needs them most.
The Bottom Line
Credit invisibility is a data problem that behaves like a creditworthiness problem, because automated underwriting treats absence of information the same way it treats bad information. It falls hardest on populations already disadvantaged, and it persists because the system records borrowing rather than payment reliability generally. The fixes that work are unglamorous: report the rent, use the bank account data, and offer products that manufacture a record safely. None of them requires a new theory of credit risk, only a wider definition of what counts as evidence.