For a while now, credit-card delinquency numbers have looked alarming, feeding a steady narrative of a consumer-credit crisis. But a closer look, which The Wall Street Journal reports, reveals that a meaningful part of the frightening figure is an artifact of measurement rather than a measure of fresh distress. A good deal of it is old debt: loans that banks already gave up on and charged off long ago, but that still linger on consumers' credit reports, quietly inflating the delinquency rate that everyone cites.

Understanding why is a small but useful lesson in how a number can be perfectly accurate and still mislead. The delinquency rate is not lying. It is simply counting something slightly different from what most people assume it counts.

Two measures, two pictures

Start with a puzzle. There are two main ways to measure credit-card delinquency, and they have been telling noticeably different stories. The Federal Reserve's measure, built from what banks report on their own balance sheets, sits at around 2.9% as of early 2026, down from its recent peak and close to its pre-pandemic normal. The New York Fed's measure, built from consumer credit reports, has looked more elevated and more worrying.

Same economy, same borrowers, two different readings. When two measures of the same thing diverge, the divergence is usually a clue rather than a contradiction, and here the clue is in how each one handles a loan that has been charged off.

What a charge-off does, and doesn't do

When a borrower falls far enough behind, typically around 180 days past due, the bank charges the loan off, meaning it writes the balance off as a loss. On the bank's books, that loan then leaves the delinquency count entirely; it has been removed and recognized as a loss, so it no longer shows up as delinquent in the bank-based measure.

But the debt does not vanish from the borrower's life, or from their credit report. Charged-off debt can linger on a consumer credit report for years, still showing up as a derogatory, delinquent mark. So on the consumer-report measure, that same loan keeps being counted as delinquent long after the bank has written it off and moved on. One recent analysis estimated that stale, charged-off card debt lingers in the consumer-report delinquency data for roughly two years before aging off. That lag, between when a bank is done with a loan and when it finally disappears from the credit report, is the whole story.

The number is measuring a stock, not a flow

Here is the deeper point beneath the accounting. What most people think the delinquency rate measures is current distress, how many people are falling behind right now. That is a flow, a rate of new events. But the consumer-report measure actually counts a stock, the total pile of delinquent debt sitting on credit reports at a given moment, and that pile includes both genuinely new distress and the lingering residue of old distress that has already been charged off.

Those are two different things, and they behave differently at exactly the wrong moment. When a wave of charge-offs occurs, as happened across 2023 and 2024, the stock of lingering charged-off debt swells and then stays elevated for a couple of years while it slowly ages off reports, holding the delinquency rate up even after the flow of new distress has calmed down. So a high reading can reflect a wave that already crested rather than water still rising. The statistic, in part, is looking backward at a storm that has largely passed, not forward at one still gathering.

Why this trips people up

This is a specific instance of a general trap worth recognizing, because it shows up all over economic data. A great many headline statistics measure a stock, an accumulated state, when what people actually care about is the flow, the rate of change right now. And a stock and a flow diverge most sharply precisely at turning points. After conditions begin to improve, the stock stays elevated for a while, because it takes time for the old accumulation to drain away, even as the flow of new problems has already slowed.

Read a stock as though it were a flow, and a receding problem looks like an ongoing crisis. That is exactly what has happened with credit-card delinquency measured off consumer reports: a stock, swollen by a past wave of charge-offs that are still aging off, read as if it were a live flow of new distress. The number looks like a crisis in part because the crisis it reflects is already in the rearview mirror.

But don't swing to "everything's fine"

It would be a mistake to overcorrect here, and the caveats matter as much as the main point. The stale charged-off debt is not fake. It is the echo of genuine distress that really happened; people truly did fall behind and lose those accounts. The argument is only that this distress is past rather than ongoing, so counting it as current overstates today's conditions, not that it never occurred.

And, crucially, "the card-delinquency alarm is overstated" is a long way from "consumers are fine." There are real signs of strain that this artifact does nothing to explain away. Credit-card balances stand at record highs, around $1.25 trillion and up nearly 6% in a year. Auto-loan serious delinquency has climbed to roughly 5.6%, its highest in the published series and above even its financial-crisis-era peak, a genuinely worrying figure that has nothing to do with stale card debt. And delinquency at smaller, subprime-heavy banks runs far above the national average, near 6.4%, revealing real stress concentrated among the most vulnerable borrowers. The honest picture is mixed: the specific credit-card delinquency headline is padded by old debt, while the cleaner measures show real pockets of strain elsewhere that deserve attention.

How to read a scary credit number

So when a frightening consumer-credit statistic crosses your screen, a few questions turn alarm into understanding. Which measure is this, bank-based or consumer-report-based? Does it include stale charged-off debt still slowly aging off reports? And what do the cleaner, flow-based measures, and the other categories such as auto loans and subprime borrowers, actually say?

The same underlying reality can be made to look like a crisis or a recovery depending on which number gets quoted, and the difference lives in the plumbing, in how each measure treats a debt once it has been written off. Knowing that plumbing is the entire difference between being alarmed by a statistic and actually understanding it, and it is not knowledge reserved for economists; it is just a matter of asking what a number counts before reacting to how big it is.

In the end, the credit-card delinquency scare is a clean illustration of how a figure can be entirely accurate and still point the wrong way. The rate is not wrong. It is counting exactly what it counts, delinquent debt sitting on credit reports. But that count includes the fading echo of a distress wave that has already passed, so it reads as an ongoing crisis when much of it is aftermath. The remedy is not to distrust the data but to learn its plumbing, to know whether you are looking at a stock or a flow, whether old charged-off debt is padding the total, and what the cleaner measures report. Do that, and the consumer-credit picture resolves into something more honest than either the crisis headline or a complacent all-clear: a card-delinquency figure inflated by old debt aging off reports, set against real and separate strain in auto loans and among subprime borrowers. The numbers most worth worrying about, as is so often the case, turned out not to be the loud one everyone was citing, but the quieter ones they were not.

Primary sources

  1. The Wall Street Journal for the framing that stale, charged-off loans lie behind elevated credit-card delinquency figures.
  2. PNC Economics research for the finding that the New York Fed's consumer-report delinquency data includes credit-card debt already charged off but lingering on consumer credit files, with an estimated roughly two years of stale charged-off debt persisting in the data, and for the argument that delinquencies are not at financial-crisis levels.
  3. The Federal Reserve Board of Governors, via FRED, for the bank-based credit-card delinquency rate of about 2.9% as of the first quarter of 2026, down from its recent peak.
  4. The New York Fed Household Debt and Credit Report for total credit-card balances of roughly $1.25 trillion, up about 6% year over year, and consumer-report-based delinquency measures.
  5. Industry data compilations for the charge-off rate near 3.8%, auto-loan serious delinquency around 5.6%, a series high above the financial-crisis peak, and delinquency near 6.4% at banks outside the largest 100.
  6. TransUnion's 2026 consumer-credit outlook for the view that stable delinquency rates and slowing balance growth reflect relative consumer resilience.