Blog

One maxed-out card among four: does per-card utilization matter?

Three hypothetical households owe the same total on the same limit, spread differently across four cards, and show how overall and per-card ratios compare.

Two people can owe exactly the same amount on the same total credit limit and look very different on paper. One has four cards each a fifth full. The other has one card at its limit and three nearly empty. Overall credit utilization is identical, and the pictures are not. Scoring models are generally described as looking at both the overall ratio and the ratio on individual cards, and the exact weights are not published, so the sensible reading is that a maxed-out card is worth attention even when the household total looks fine.

This post builds three labeled hypothetical households with the same totals, shows the per-card arithmetic, and then asks what each would do about it. The households and numbers are inventions chosen so the math is easy to verify. They are not typical limits or balances.

The setup: same totals, three shapes

Every household below has four cards, a combined credit limit of $20,000 and a combined statement balance of $4,000. Overall utilization is $4,000 ÷ $20,000 = 0.20, so 20 percent, in all three. What differs is how the limits and the balances are distributed.

For reference, amounts owed make up about 30 percent of a FICO Score in the general model, according to FICO's published description, and Experian says utilization above roughly 30 percent starts to weigh more heavily on scores. FICO does not publish a magic number, and lower is generally better. Whether a model weighs a single card's ratio heavily is not published, which is the reason to look at all three shapes and not at one.

The three households

Each table lists limit, statement balance and the per-card ratio.

Household A: the even spread

Lena has four equal cards, each with a $5,000 limit, and she keeps a balance on each.

Card Limit Statement balance Ratio
Card 1 $5,000 $1,000 20%
Card 2 $5,000 $1,000 20%
Card 3 $5,000 $1,000 20%
Card 4 $5,000 $1,000 20%
Total $20,000 $4,000 20%

Every per-card figure matches the overall figure, so there is nothing for a model to disagree about. The highest single-card ratio is 20 percent. If the household needed to lower any number, any payment would help evenly.

This household also has the most flexibility. Each card has $4,000 of available credit, which is the limit minus what is owed, so a surprise expense does not run into a ceiling.

Household B: a small card maxed out

Omar has the same $20,000 of total limits, but the limits are uneven, and one small card is at its limit.

Card Limit Statement balance Ratio
Card 1 (small) $1,500 $1,500 100%
Card 2 $4,500 $1,000 22.2%
Card 3 $6,000 $1,000 16.7%
Card 4 $8,000 $500 6.25%
Total $20,000 $4,000 20%

Check the sums: limits are 1,500 + 4,500 + 6,000 + 8,000 = 20,000, and balances are 1,500 + 1,000 + 1,000 + 500 = 4,000. Per-card arithmetic: 1,500 ÷ 1,500 = 1.00; 1,000 ÷ 4,500 = 0.222; 1,000 ÷ 6,000 = 0.167; 500 ÷ 8,000 = 0.0625.

The overall figure of 20 percent is the same as Lena's, but the highest single-card ratio is 100 percent. A model that reads only the overall number sees two identical files. One that reads the per-card ratios sees Omar's small card as fully used. Notice what the simple average of the four per-card ratios gives: (100 + 22.2 + 16.7 + 6.25) ÷ 4 = 36.3 percent, far above the 20 percent overall figure. The two methods diverge sharply here because one small card carries a lot of weight in the average and very little in the total.

Omar also has a practical problem: Card 1 has $0 of available credit. Any new charge there either is declined or goes over the limit. Under the CARD Act rules in Regulation Z, an issuer can charge an over-limit fee only if the cardholder opted in to over-limit transactions after a separate notice, and then at most one per billing cycle. Without opt-in the issuer may still approve the transaction but cannot charge the fee. Whether a given card allows over-limit transactions at all is a matter for the card agreement.

Household C: everything on one card

Priya uses one card for nearly everything and keeps the rest for emergencies.

Card Limit Statement balance Ratio
Card 1 $5,000 $4,000 80%
Card 2 $5,000 $0 0%
Card 3 $5,000 $0 0%
Card 4 $5,000 $0 0%
Total $20,000 $4,000 20%

The overall ratio is again 20 percent. The highest single-card ratio is 80 percent. The simple average of per-card ratios is (80 + 0 + 0 + 0) ÷ 4 = 20 percent, which matches the overall figure only because the limits happen to be equal. That is a coincidence of the setup, and a good reminder that the average of percentages and the ratio of totals are different calculations that agree only in special cases.

Priya's situation is also fixable without paying anything, which is a point worth seeing in arithmetic below.

Side by side

Household Overall ratio Highest single card Cards at or above 50%
A: even spread 20% 20% 0
B: small card maxed 20% 100% 1
C: everything on one card 20% 80% 1

A model that reads only the first column cannot tell these households apart. A model that also reads the second or third column can. How much it cares is unpublished. A human underwriter reviewing a file by hand may read all three columns, and a card issuer's own internal model may use its own mix.

What each household could do

The goal here is not to reach a particular score but to show the arithmetic of the options.

Option 1: pay the maxed card down before the statement closes

Reported balances commonly reflect the statement closing date, so a payment before that date is what shows. Choose a per-card marker, say 30 percent.

  • Omar: 30 percent of $1,500 is $450. Card 1 is at $1,500, so a $1,050 payment before closing brings it to $450, which is 30 percent. Total balance becomes $4,000 − $1,050 = $2,950, so overall is $2,950 ÷ $20,000 = 14.75 percent.
  • Priya: 30 percent of $5,000 is $1,500. Card 1 is at $4,000, so she needs to pay $2,500. Total becomes $1,500, and overall is $1,500 ÷ $20,000 = 7.5 percent.

In both cases the payment fixes the single-card problem and the overall ratio together.

Option 2: shift spending, not money

Priya has a second option that Omar does not. Her other cards have the same $5,000 limit and no balances, so she can stop adding to Card 1 and put new spending on the others. If over the next cycles she brings Card 1 down to $1,000 by paying, and her purchases land on Cards 2, 3 and 4 at about $1,000 each, each statement shows 20 percent and the household resembles Lena's. This takes no balance-transfer fee because no balance is moved; it simply changes where new charges go.

If Priya instead moved the existing $3,000 from Card 1 to the others through balance transfers, the balance transfer fee, which is typically about 3 to 5 percent of the amount moved according to commonly cited ranges, would add roughly $90 to $150 on $3,000. Those fees are real cost for a ratio benefit, and issuers often do not allow transfers between cards they issue, so the cost-free route is usually better.

Option 3: raise the small limit

Omar could request a credit line increase on Card 1. If the issuer raised it from $1,500 to $3,000 and the balance stayed at $1,500, the ratio would fall to 50 percent. Overall would become $4,000 ÷ $21,500 = 18.6 percent. The increase helps only if the balance does not grow into the new room, and it depends on the issuer's decision. See how a credit limit change moves your utilization, worked out by hand for the break-even arithmetic.

Option 4: do nothing and accept the snapshot

If no application is coming, one month of a maxed card is a snapshot that the next closing replaces. What matters more than the ratio is paying at least the minimum on time. pay the highest-rate card first, or the one closest to its limit? discusses how to choose between paying the card with the highest interest and paying the card nearest its limit.

When a small card makes the ratio look worse than it is

One caveat protects against over-reading. A card with a small limit makes the per-card ratio swing on tiny amounts. A $300 limit with a $290 balance is 97 percent, and a $100 payment brings it to $190 ÷ $300, about 63 percent. That is a $100 move in cash and a 34-point move in the ratio. Models and lenders know that small-limit cards behave this way, and a person with a $300 card and a clean payment history is not in the same position as a person with a $15,000 card at its limit. The ratio is a signal, and size is part of the context.

How a lender or a model might read each household

Because the weights are unpublished, the honest approach is to describe what each reader would see, not to predict an outcome.

A scoring model reads the file as data. It receives each card's limit and balance, computes whatever ratios it was built to use, and produces a number. For Lena's file every computation points the same way. For Omar's file the overall measure looks modest and the per-card measure on the small card looks extreme. For Priya's file the overall measure looks modest and the per-card measure on the heavy card looks high, with three clean cards beside it. If a model leans on the per-card figures, Omar and Priya are likelier to see a larger effect than Lena, and Omar's 100 percent card is the more extreme case.

A human underwriter reads the same file differently. Reviewing a mortgage or a large loan by hand, the underwriter may see a card at its limit and ask why. The explanation matters, and a one-time medical bill or a large repair that is being paid down looks different from a card that has been maxed for a year. This is why it helps to know your own story before an application: you can explain the number only if you have looked at it.

An issuer reads your file for its own purposes, too. Issuers can use internal data, such as how you pay on cards you hold with them, along with bureau data. A card at its limit with on-time payments is one signal, and a card at its limit with late payments is another. None of that is published, and it is a reason to treat all of this as a direction and not a prediction.

Why one card ends up maxed

A single card at its limit is rarely a mystery once you trace it. The usual causes are ordinary.

  • A small starting limit. A first card with a $500 or $1,500 limit fills quickly. Omar's Card 1 is this case, and the same card can look maxed after one repair bill.
  • One big purchase. A medical expense, a car repair, a move or a flight lands on whichever card is in the wallet. If that card is the one with the smallest limit, it fills first.
  • A balance transfer. Moving a balance onto a new card can fill most of its limit by design. A transferred balance that is being paid down on a plan is a different story from spending, but the ratio looks the same on the report.
  • Habit. One card for everything, as in Priya's case, concentrates the balance whatever the limits are.
  • A limit cut. A lowered limit makes a stable balance look heavier, as the section above showed.

Knowing the cause tells you the fix. A small limit points to a request for more room or a shift of spending. One big purchase points to a payoff plan. Habit points to spreading the load across cards.

A worked recovery: Omar in three cycles

To see how fast a maxed card can recover, follow Omar's Card 1 across three statements. This is an illustration that ignores interest and assumes no new charges, so it slightly understates how long a real payoff takes. At an illustrative 24 percent APR, $1,500 would accrue about $1,500 × 0.24 ÷ 12 = $30 of interest in a month, so a real balance would fall a little more slowly than the table shows.

Say Omar can set aside $400 a month and sends it to Card 1 before each statement closes.

Cycle Payment before closing Card 1 balance reported Card 1 ratio Household total Overall ratio
Start none $1,500 100% $4,000 20%
1 $400 $1,100 73.3% $3,600 18%
2 $400 $700 46.7% $3,200 16%
3 $400 $300 20% $2,800 14%

Check a few cells: 1,100 ÷ 1,500 = 0.733, 700 ÷ 1,500 = 0.467, 300 ÷ 1,500 = 0.20. Household totals are $4,000 less $400, $800 and $1,200. Overall ratios are 3,600 ÷ 20,000 = 0.18, 3,200 ÷ 20,000 = 0.16 and 2,800 ÷ 20,000 = 0.14.

By the third statement, Card 1 is below the 30 percent marker and the household looks better on both measures. The overall ratio moved by six points while the per-card ratio moved by eighty, which shows why the maxed card is the one to attack first. If any of those months needed the $400 for rent or food, Omar would skip the extra payment and keep paying the minimums. A utilization plan never outranks a due date.

What happens when the issuer cuts a limit

One way a household changes shape without any new spending is a limit cut. If Omar's issuer cut Card 4 from $8,000 to $2,000, Card 4's ratio would jump from 500 ÷ 8,000 = 6.25 percent to 500 ÷ 2,000 = 25 percent, and the overall limit would drop to $14,000, giving $4,000 ÷ $14,000 = 28.6 percent overall. Nothing about Omar's spending changed. My issuer lowered my credit limit. What does that do to my score? covers why issuers do this and what to check.

A decision list for your own cards

Run the table for your own statements, then ask these questions in order.

  1. Is any single card above 80 or 90 percent of its limit? That is a maxed card. Decide whether a payment before closing is possible this cycle.
  2. Is your overall ratio fine but one card high? Compare your shape with Households B and C. Shifting new spending to a card with room may fix it without any payment.
  3. Is the high card also your highest-interest card? Then paying it down both lowers the ratio and cuts interest, and the choice is easy.
  4. Is the high card a small-limit card? Then a modest payment moves its ratio a lot. Consider whether a limit increase request is reasonable.
  5. Is an application coming in the next two or three months? Plan payments for the cycle before it.
  6. Is anything unaffordable? If you cannot make the minimum payments, utilization is not the first problem. Call the number on the back of the card to ask about hardship options, or see a nonprofit credit counselor.

The credit utilization calculator and work out your credit utilization three ways with your own numbers help with the arithmetic.

Key takeaways

  • Overall and per-card utilization can disagree: three households each owed $4,000 on $20,000 of limits and showed 20 percent overall, with single-card highs of 20, 100 and 80 percent.
  • Models are generally described as reading both, and the weights are unpublished, so watch both.
  • The average of per-card percentages is not the overall ratio. In Household B the average was 36.3 percent against 20 percent overall.
  • A payment before the statement closes fixes a maxed card fastest: $1,050 for Omar, $2,500 for Priya to reach 30 percent on the heavy card.
  • A small limit exaggerates swings, and shifting new spending to cards with room can help without paying anything extra.

Terms used in this article

Related tools

Latest on this topic