U.S. credit-card balances reached $1.26 trillion in the second quarter of 2026, up $21 billion from the previous quarter and close to the $1.28 trillion peak recorded at the end of 2025. Those figures, reported by ABC News from the Federal Reserve Bank of New York's household-credit data, describe a large debt stock. They do not, by themselves, reveal whether the average borrower is running out of room.
The better diagnosis combines three views: who holds the balance, how accounts move into delinquency, and how much unused credit remains. That approach produces a less dramatic but more useful conclusion. Stress is real and expensive, yet the aggregate data still point to concentration and cohort effects rather than a uniform consumer break.
A near-record stock says nothing about who is carrying it
The New York Fed's 2026 second-quarter report puts total household debt at $18.77 trillion, down $13 billion from the first quarter. Mortgage balances fell by $74 billion while non-housing balances increased by $48 billion. Credit cards accounted for $21 billion of that increase. The mix matters: card growth occurred during a quarter in which the largest household-debt category contracted.
A nominal record also needs scale. Prices, incomes, the adult population and available credit limits change over time. A balance can rise even if the typical borrower's debt relative to income or credit capacity is stable. Conversely, an unchanged national balance can hide severe pressure if debt migrates toward households with less ability to repay. The headline cannot distinguish those cases because it aggregates millions of different cash-flow positions.
That is why the $1.26 trillion figure should begin a distributional question, not end the analysis. Useful cuts include balances by credit score and age, the share held by borrowers already behind, and the movement of individual accounts from current to late.
Old balances can deteriorate while new spending looks resilient
ABC reports that the percentage of card balances more than 90 days delinquent rose from 7.6% in mid-2022 to 12.8% in early 2026. The same report quotes New York Fed researchers saying that older outstanding debts help explain the increase, rather than a sudden wave of missed payments on new purchases. That distinction is easy to lose: a stock of troubled loans can age into worse delinquency categories even when the flow of newly delinquent accounts is steady.
The official quarterly report supports caution around the flow. It says transitions into early delinquency were largely steady for credit cards in the second quarter and transitions into serious delinquency were largely unchanged across products. It also says aggregate delinquency edged down, with 4.7% of all outstanding household debt in some stage of delinquency. These facts do not cancel the high card-delinquency stock. They locate it: legacy stress may be persisting without accelerating at the same rate.
For consumer demand, both can matter. New spending can remain resilient while older revolving balances absorb more monthly cash. That pattern is not a contradiction; it can continue until households cut discretionary spending, refinance, default or receive income relief.
Utilization is the denominator hiding behind the headline
Credit-card debt is revolving credit: borrowers can draw, repay and draw again up to a limit. The Federal Reserve's G.19 methodology explains that card loans make up most, but not all, revolving consumer credit. The New York Fed separately defines utilization as outstanding balance divided by credit limit and warns that its data may overestimate the ratio when issuers do not report limits.
That denominator changes the interpretation of a rising balance. The Q2 report says aggregate card limits continued to increase, with an $85 billion rise. If limits grow alongside balances, the system is not necessarily approaching its collective ceiling. But the aggregate can still conceal constrained borrowers whose individual utilization is near 100%. Averages are least reassuring precisely when unused capacity belongs to low-risk households while high-risk households have none.
The price of carrying a balance also remains material. The Federal Reserve's consumer-credit release reported a 21.00% average rate across commercial-bank card accounts in the first quarter and 21.52% for accounts assessed interest. Those are observed terms, not a forecast. At such rates, a balance that is merely stable can still demand substantial cash from a borrower who revolves it.
Bank losses will test the household-stress thesis
For lenders, delinquency becomes economically material through provisions, charge-offs, collection expense and tighter underwriting. If older bad balances are the principal problem, losses can remain elevated even as new account performance stabilizes. If new vintages also begin deteriorating, the risk shifts from cleanup to continuing credit-cycle weakness.
The strongest counterargument to a measured reading is that household damage appears before an aggregate break. High interest costs can squeeze consumption, and lenders can reduce limits or reject marginal applicants while total credit remains available to safer borrowers. A broad national balance therefore may look calm longer than the vulnerable tail does.
Evidence that would change the assessment is straightforward: a sustained rise in the flow of current card accounts entering delinquency; worsening performance among recently originated accounts; utilization rising because balances outrun limits; and bank charge-offs spreading beyond prior vintages. Falling serious-delinquency balances with stable utilization would support the opposite conclusion.
The record-adjacent total deserves attention, but it is the wrong alarm by itself. The credit cycle lives in transitions and distributions. Investors and policymakers should watch who is losing payment capacity — and whether that group is growing — before treating $1.26 trillion as a verdict on the U.S. consumer.