Card rate hits 22.36% as revolving balances fall 4.2% in August
The Fed's G.19 for August 2026 shows credit card balances shrinking at a 4.2% annual rate while the card rate ticked up to 22.36%, before the September hike.
- What happened
- The Federal Reserve's G.19 release of October 7, 2026 put total consumer credit at $5,196.8 billion in August 2026, growing at a seasonally adjusted annual rate of 1.9%. Revolving credit, which is mostly credit cards, fell at a 4.2% annual rate, while nonrevolving credit rose at 4.1%.
- Why
- A monthly growth rate on the G.19 is one month's seasonally adjusted flow divided by the previous month's level, expressed as an annual rate. A $4.8 billion fall in card balances over one month becomes −4.2% when multiplied up to a year. The interest-rate table in the same release comes from a separate quarterly bank survey taken in the first week of August, which is why it shows 22.36% on accounts assessed interest even though the Federal Open Market Committee did not raise rates until September 16, 2026.
- The numbers that matter
- $5,196.8bnTotal consumer credit, August 2026, seasonally adjusted−4.2%Revolving credit, annual rate of change, August 202622.36%Card rate, accounts assessed interest, August 2026 survey7.54%60-month new car loan rate, August 2026 survey
- Why it matters
- The card rate the Fed measured in August is 0.21 points above the May reading, which on a $5,000 balance is $0.88 a month more interest, before any effect of the September 16 hike shows up in the November data.
- What happens nextas we saw it on 13 Oct 2026
- The next G.19, covering September 2026, is scheduled for release at 3:00 p.m. Eastern on November 6, 2026. The next quarterly bank interest-rate survey, by the pattern in the release, covers the first week of November 2026.
The August 2026 reading is above May 2026 but below the 2024 annual average and below August 2025.
| 2023 average | 22.1% |
|---|---|
| 2024 average | 22.9% |
| Aug 2025 (Q3) | 22.8% |
| Feb 2026 (Q1) | 21.5% |
| May 2026 (Q2) | 22.1% |
| Aug 2026 (p) | 22.4% |
Americans paid down their credit cards in August 2026. The Federal Reserve's G.19 release of October 7, 2026 shows revolving consumer credit, which is mostly card balances, falling at a seasonally adjusted annual rate of 4.2% in the month. It is the only negative figure in the growth-rate columns the release prints, after rises at annual rates of 6.0% in June and 2.5% in July. At the same time the average rate charged on card accounts that carry a balance rose to 22.36%, from 22.15% in the May survey.
The two facts point in different directions, so this piece takes them one at a time, and explains why the rate table is older than the release date suggests.
The August balance figures
The seasonally adjusted level of consumer credit outstanding was $5,196.8 billion in August 2026 (preliminary), up from a revised $5,188.5 billion in July. Revolving credit was $1,352.4 billion, down from $1,357.2 billion. Nonrevolving credit was $3,844.4 billion, up from $3,831.3 billion. Source: Federal Reserve Board, G.19 Consumer Credit, released October 7, 2026.
The release's opening line reads: “In August, consumer credit increased at a seasonally adjusted annual rate of 1.9 percent.” Revolving credit fell at an annual rate of 4.2% and nonrevolving credit rose at an annual rate of 4.1%. The seasonally adjusted flows, at an annual rate, were $99.4 billion in total, −$57.4 billion revolving and $156.8 billion nonrevolving.
The G.19's own definition is that a monthly growth rate is the seasonally adjusted flow for the month divided by the previous month's seasonally adjusted level. Revolving credit: −$57.4 billion ÷ $1,357.2 billion = −4.23%, which the release rounds to −4.2%. The same arithmetic for the total gives $99.4 billion ÷ $5,188.5 billion = 1.92%, rounded to 1.9%, and for nonrevolving $156.8 billion ÷ $3,831.3 billion = 4.09%, rounded to 4.1%. The flows are annualised, so the change in the level across one month is roughly one twelfth: $1,352.4 billion − $1,357.2 billion = −$4.8 billion for revolving.
A reader who sees “−4.2%” and pictures a twentieth of card debt disappearing has misread it. In level terms, card balances fell by $4.8 billion on a stock of $1,357.2 billion, which is 0.35% in one month. The convention exists so that months, quarters and years can share one table.
| Period | Total | Revolving | Nonrevolving |
|---|---|---|---|
| 2024 | 2.0 | 4.1 | 1.2 |
| 2025 | 2.2 | 3.1 | 1.8 |
| Q1 2026 | 2.6 | 4.2 | 2.0 |
| Q2 2026 (r) | 3.1 | 4.9 | 2.4 |
| June 2026 (r) | 3.6 | 6.0 | 2.7 |
| July 2026 (r) | 4.1 | 2.5 | 4.7 |
| August 2026 (p) | 1.9 | −4.2 | 4.1 |
The same release prints the figures before seasonal adjustment. On that basis revolving credit was $1,315.2 billion in August, down from $1,316.1 billion in July, a flow of −$10.8 billion at an annual rate against −$57.4 billion after adjustment. The not seasonally adjusted total was $5,162.9 billion, against $5,196.8 billion adjusted.
When the adjusted decline (−$57.4 billion at an annual rate) is much larger than the raw decline (−$10.8 billion), the seasonal factor is telling you that August usually brings an increase in card borrowing, and that 2026 did not deliver one. That is what seasonal adjustment is for. It is an inference from the two published numbers, not a statement the Fed makes.
July was revised, slightly
The September 8, 2026 release had reported July's total growth as 4.2% at an annual rate, from a preliminary level of $5,186.2 billion. The October 7 release shows July at 4.1% and $5,188.5 billion. Revolving credit for July is unchanged at $1,357.2 billion, which is the figure the site's credit card interest piece used on October 4. Nonrevolving credit for July was revised up from $3,829.0 billion to $3,831.3 billion. The revision is $2.3 billion on a $5 trillion stock, and it moved the headline rate by a tenth of a point, which is the kind of change that should be expected every month rather than read as news.
What the rate table shows, and when it was measured
The second half of the G.19 is a table of average interest rates at commercial banks. Unlike the balance figures, which are monthly, these come from a quarterly survey. Footnote 5 of the release says the new-car and personal-loan rates are simple unweighted averages of each bank's most common rate, collected during the first calendar week of the middle month of each quarter. That is why the June and July columns show “n.a.” and the August column shows a number: August is the middle month of the third quarter, and the survey was taken in its first week.
| Product | 2024 | 2025 | Q3 2025 | Q1 2026 | Q2 2026 (r) | Aug 2026 (p) |
|---|---|---|---|---|---|---|
| Credit cards, all accounts | 21.58 | 21.22 | 21.39 | 21.00 | 20.94 | 21.19 |
| Credit cards, accounts assessed interest | 22.89 | 22.32 | 22.83 | 21.52 | 22.15 | 22.36 |
| New car loans, 60-month | 8.16 | 7.65 | 7.64 | 7.53 | 7.14 | 7.54 |
| New car loans, 72-month | 8.29 | 7.80 | 7.80 | 7.53 | 6.97 | 7.17 |
| Personal loans, 24-month | 12.27 | 11.50 | 11.14 | 11.36 | 11.86 | 11.90 |
Accounts assessed interest: 22.36% (August 2026) − 22.15% (May 2026, the Q2 column) = +0.21 points. Against the Q3 2025 column, which is the August 2025 survey: 22.36% − 22.83% = −0.47 points. The gap between the two card measures is 22.36% − 21.19% = 1.17 points; the all-accounts figure includes cards that were paid in full and charged no interest.
The Federal Open Market Committee raised its target range to 3.75%–4.00% on September 16, 2026. The card and loan rates in the October 7 G.19 were collected in the first calendar week of August 2026, five to six weeks before that decision. By the pattern footnote 5 describes, the first bank survey taken after the hike is the one for the first calendar week of November 2026, and the September and October columns of the rate table should be expected to read “n.a.”. Anyone reading 22.36% as the post-hike card rate is reading a number that predates the hike.
The 60-month new car rate was 7.54% in the August 2026 survey, up from 7.14% in May 2026 and down from 7.64% in the August 2025 survey. The 72-month rate was 7.17%, up from 6.97% and down from 7.80% a year earlier. The 24-month personal loan rate was 11.90%, against 11.86% in May.
Who holds the $5.2 trillion
On a not seasonally adjusted basis, the release's holders table shows depository institutions with $2,099.9 billion, the federal government with $1,621.5 billion, credit unions with $734.1 billion and finance companies with $697.3 billion in August 2026. The total across all holders was $5,162.9 billion.
$1,621.5 billion ÷ $5,162.9 billion = 31.4%. Depository institutions, which are mainly banks: $2,099.9 billion ÷ $5,162.9 billion = 40.7%. Together they hold just over seven dollars in ten.
What the G.19 does not cover matters as much as what it does. The Fed's description of the release is credit extended to individuals for household, family and other personal expenditures, excluding loans secured by real estate. Mortgages and home equity lines are therefore outside the $5.2 trillion. The site's household debt piece covers the broader New York Fed measure that includes them.
What it means for your money
Two numbers in this release touch a household directly.
$5,000 × 22.36% ÷ 12 = $93.17. At the May survey rate of 22.15% the same month cost $92.29. The difference is $0.88 a month, or $10.50 over a full year ($5,000 × 0.21 points). Monthly rate taken as the annual rate divided by 12, no fees, no new purchases, interest rounded to the cent. The credit card arithmetic piece works through the full minimum-payment path at 22.15%; each figure there is now 0.21 points stale, upward.
Monthly payment at 7.54% over 60 months on $40,000 is $802.28; at 7.14% it is $794.69. The 0.40-point rise in the survey adds $7.59 a month, or $455 over the life of the loan. Standard amortising-loan formula, payment = P × r ÷ (1 − (1 + r)^−n), with r the annual rate ÷ 12 and n the number of months, no fees.
The mechanism for what comes next is the one described in the site's interest rates and how the Fed works pages: card rates at US banks are mostly set as a margin over the prime rate, and prime moves with the Fed's target range. The September 16 increase of the target range to 3.75%–4.00% is not in any figure in this release. Whether and how much it shows up in the November 2026 survey is something a later G.19 will report; this piece does not estimate it.
On balances, the August fall in revolving credit is one month, preliminary, and will be revised on November 6, 2026. The site's one month is not a trend piece applies in full. The quarterly figures in the table above, which smooth the monthly noise, still show revolving credit growing at 4.9% in the second quarter of 2026.
Sources
- Consumer Credit - G.19, October 7, 2026 (August 2026 data)Federal Reserve Board · Oct 12, 2026 · Primary source
- G.19 Consumer Credit, August 2026, released October 7, 2026 (PDF: not seasonally adjusted levels, flows, holders, footnote 5)Federal Reserve Board · Oct 12, 2026 · Primary source
- Consumer Credit - G.19, September 8, 2026 (July 2026 data, preliminary)Federal Reserve Board · Oct 12, 2026 · Primary source
- Consumer Credit - G.19: AboutFederal Reserve Board · Oct 12, 2026 · Primary source
- Federal Reserve Board calendar, November 2026 (G.19 release November 6, 2026, 3:00 p.m.)Federal Reserve Board · Oct 12, 2026 · Primary source
- FOMC statement, September 16, 2026Federal Reserve Board · Sep 16, 2026 · Primary source
Related analysis
At a 22.15% card APR, $5,000 on minimums takes 21 years to clear
The Fed's second-quarter 2026 card-rate reading is 22.15% on accounts charged interest. What that does to a $5,000 balance, month by month, and how prime moves it.
Household debt fell $13bn in Q2 2026; card stress held at 6.97%
The New York Fed puts household debt at $18.771 trillion for June 2026. The total dipped; new card delinquencies ran at 4.6 times the mortgage pace.
The Fed stopped arguing with itself, and raised rates
In July three officials wanted higher rates and lost. On September 16 the vote was 12–0. Nothing about the committee changed — the data did.
August added 162,000 jobs. The average of the prior twelve months was 31,000.
The strongest hiring month in a year sits on top of a twelve-month average five times smaller. Both numbers are true, and only one of them is a trend.