A payday loan advertises a fee, not an interest rate, and the fee sounds small. Fifteen dollars per hundred borrowed reads like a service charge. Converted to the annual rate every other lender must quote, it is 391 percent. The conversion is not a trick played by critics. It is arithmetic, and the gap between how the price is presented and how it accrues is the entire business model.
Here is how the math actually runs.
Turning a flat fee into a rate
The Consumer Financial Protection Bureau reported that the median storefront payday loan carries a fee of $15 per $100 borrowed, a median term of 14 days, and a median loan size of $350. Those three numbers produce an annual percentage rate of 391 percent.
The calculation is simple. Fifteen dollars on a hundred is 15 percent for the period. A 14-day period repeats about 26 times in a year. Multiply 15 percent by 26.07 and you get roughly 391 percent.
Lenders object that nobody holds the loan for a year, which is true and also the wrong objection. An annual percentage rate is a unit of comparison, not a prediction. Quoting price per year is how a two-week product gets compared against a credit card at 24 percent or a personal loan at 12 percent. Without it, the borrower has no way to know that the cheapest credit card available to them costs about a sixteenth as much.
On the median $350 loan, the fee is $52.50 for fourteen days.
The part the two-week frame hides
The defense of payday pricing assumes the loan gets repaid in two weeks. The lender’s own regulator measured whether it does.
The Consumer Financial Protection Bureau’s 2014 analysis covered more than 12 million storefront payday loans across 30 states. Its findings:
- Over 80 percent of loans were rolled over or followed by another loan within 14 days.
- Across all states studied, 82 percent of loans were renewed within fourteen days, and that figure varied by only three percentage points between states with cooling-off requirements and states without.
- Half of all loans sat in a sequence at least ten loans long.
- Sixty-two percent of loans were in sequences of seven or more.
- Among new borrowers, 64 percent became renewers. Only 15 percent repaid without borrowing again within fourteen days.
- For more than 80 percent of sequences lasting beyond one loan, the final loan was the same size as or larger than the first.
That last finding rules out the most sympathetic reading. If borrowers were climbing out, loan sizes would shrink across a sequence. They do not.
What a sequence costs
Run the median numbers forward. A borrower takes $350 and owes $402.50 in fourteen days. Short of the full amount, they pay the $52.50 fee and roll the principal.
Two weeks later the same choice arrives. The Bureau found that 22 percent of new loans were renewed six or more times. Six renewals on $350 costs $315 in fees against $350 of principal that has not moved. The Bureau’s own press release made the point directly: a consumer taking an initial loan and six renewals will have paid more in fees than the original loan amount.
Ten renewals, which describes half of all loans by the Bureau’s sequence measure, costs $525 in fees on $350 borrowed.
Nothing about that requires an unusual borrower. It requires a borrower whose budget was short by $350 once and did not become $402.50 longer two weeks later.
Why the structure produces this result
A conventional installment loan amortizes. Each payment covers interest and reduces principal, so the balance falls and the loan ends.
A single-payment payday loan does not amortize. The borrower owes the entire principal plus the fee on one date. Someone who could produce $402.50 in a lump on payday could usually have produced $350 two weeks earlier. The repayment structure asks for more money, sooner, from a household that just demonstrated it was short.
Rolling over resolves the immediate shortfall and resets the same problem with a new fee attached. The product works exactly as designed. The design assumes repayment capacity that the borrowing itself contradicts.
What the rules currently require
The Consumer Financial Protection Bureau finalized a payday lending rule in 2017 that included a requirement to assess a borrower’s ability to repay. The Bureau revoked that portion in July 2020, removing the mandatory underwriting provisions.
The payment provisions survived and became operative on March 30, 2025. They limit repeated withdrawal attempts after two consecutive failures and require advance notice before payment attempts. Two days before they took effect, the Bureau announced it would not prioritize enforcement or supervision actions regarding penalties associated with those provisions, and said it was considering a rulemaking to narrow the rule’s scope.
So the payment protections are binding on paper while the federal regulator has publicly said it will not police them. State law and private litigation are unaffected.
Where the product is not available
States diverge sharply. The Consumer Federation of America counts 18 states and the District of Columbia that effectively prohibit high-cost payday lending through usury caps, with most setting the limit at 36 percent APR, plus three more that block it through other mechanisms. The Center for Responsible Lending counts 20 states and the District of Columbia with caps around 36 percent, and finds triple-digit rates still legal in 28 states, ranging from 140 to 662 percent on a $400 single-payment loan.
The counts differ because the definitions differ, not because either organization is wrong. Both point at the same threshold: 36 percent is the line legislatures keep choosing.
The broader arithmetic
Payday lending prices a specific condition, which is a household with income but no buffer. The Federal Reserve’s household survey has repeatedly found large shares of adults unable to cover a modest unexpected expense from savings, and a household in that position facing a car repair has few options that are not expensive.
Fight For A Living Wage, a nonpartisan grassroots 501(c)(3), argues that the squeeze is an affordability problem spanning housing, healthcare, childcare, food, transport and education rather than a wage-floor problem alone. Short-term lending sits downstream of that. A household whose fixed costs consume its income has no buffer, and the absence of a buffer is what 391 percent is priced against.
The Consumer Financial Protection Bureau publishes the loan-level research cited here, and the arithmetic above uses only its published medians. Anyone can check it with a calculator.