Two Medicare late enrollment penalties are described the same way in almost every summary: sign up late, pay more for life. They are built on different units. The Part D drug penalty counts single months and multiplies them by a premium figure that the Centers for Medicare & Medicaid Services resets every year. The Part B medical penalty counts only completed twelve-month periods and multiplies them by a much larger premium. For 2026 the two base figures are $38.99 and $202.90. What follows is what each formula does with the same gap, and why one of them costs a different amount every January while the gap behind it does not change. What Starts the Part D Count The condition is not lateness in general. The Medicare drug coverage cost page states: “You may owe a late enrollment penalty if at any time after your Initial Enrollment Period is over, there’s a period of 63 or more days in a row when you don’t have Medicare drug coverage or other creditable prescriptio...
Two payoff orderings dominate the advice: pay the smallest balance first (the snowball), or the highest interest rate first (the avalanche). The arithmetic case for the avalanche is airtight — money aimed at the most expensive debt costs less than money aimed anywhere else. What the argument almost never includes is the size of the difference.
Run the two methods against the same four balances, the same monthly payment, and the same rates, in the configuration that favors the avalanche as much as a realistic set of debts can, and the avalanche wins by $442 in interest over roughly three and a half years on $13,670 of debt. That is about 3.2 percent of the principal. It is real money, and it is roughly one-fortieth of what raising the monthly payment to $500 saves on these same four balances against paying minimums alone.
That ratio is the subject here. The choice between orderings is a small optimization sitting next to a very large one, and the two are routinely presented as though they were the same size.
What each ordering specifies
Both methods assume the same two things. Every debt receives at least its minimum payment, because missing one triggers late fees and delinquency reporting regardless of strategy. Whatever the budget holds beyond the sum of the minimums goes to a single target debt; when that target clears, its payment joins the pool and the next target absorbs everything. The methods differ on one point only: which debt is the target.
Neither method changes the total monthly outlay. Neither changes any interest rate. The only variable is the sequence.
The test case
The scenario below uses four revolving balances totaling $13,670, with a fixed $500 a month available. The rates are set so that balance order and rate order are exact opposites: the smallest balance carries the lowest rate, the largest the highest. This is the arrangement in which the two methods disagree about every debt, and where the avalanche's advantage is largest.
| Debt | Balance | APR | Snowball rank | Avalanche rank |
|---|---|---|---|---|
| Card A | $520 | 16.99% | 1st | 4th |
| Card B | $1,150 | 22.15% | 2nd | 3rd |
| Card C | $3,600 | 24.24% | 3rd | 2nd |
| Card D | $8,400 | 27.99% | 4th | 1st |
The rates are anchored to published averages. For the second quarter of 2026 the Federal Reserve reported an average rate of 22.15 percent on credit card accounts assessed interest and 20.94 percent across all accounts, in the G.19 Consumer Credit release of August 7, 2026. The Consumer Financial Protection Bureau, in its 2025 report to Congress on the consumer credit card market, put the 2024 average annual percentage rate at 25.2 percent for general purpose cards and 31.3 percent for private label cards, the highest levels since at least 2015. The 16.99 percent floor and the 27.99 percent ceiling used here sit inside that observed range.
One modeling assumption has to be stated plainly, because it drives the result and no agency sets it: the minimum payment is treated as 1 percent of the balance after interest posts, plus that month's interest, with a $30 floor. Issuers set their own formulas; the one that applies to a given account is in the cardholder agreement. This structure is used here because it always amortizes. A flat percentage-of-balance minimum does not — a minimum of 2 percent of the balance fails to cover interest at any APR above 24 percent, and the balance grows even when the payment is made on time.
The result
| Snowball | Avalanche | Difference | |
|---|---|---|---|
| Months to clear all four | 43 | 42 | 1 month |
| Total interest paid | $7,680 | $7,238 | $442 |
| Total paid | $21,350 | $20,908 | $442 |
| As a share of principal | 56.2% | 53.0% | 3.2 points |
The avalanche is cheaper, as the arithmetic guarantees — by $442 spread across 43 months, or about $10.30 a month.
The snowball clears its first debt in month 7 and its second in month 15. The avalanche clears nothing until month 21, then retires three debts in its final eight months.
Why the gap compresses
Three structural features hold the difference down, none of which appear in the usual framing.
Minimum payments already do part of the snowball's work
Card A clears in month 21 under the avalanche despite being last in priority and never receiving a dollar of surplus: the $30 minimum floor retires a $520 balance on its own in under two years. Small balances disappear under either method, because a percentage-based minimum with a dollar floor pays them down fast relative to their size. The avalanche is not carrying them to the end; it is simply not accelerating them.
The rate spread applies only to the dollars that move
In the first month of this scenario the four minimums total $454, leaving $46 of the $500 budget free to direct. That $46 is the entire scope of the ordering decision in month one. Sending it to 27.99 percent instead of 16.99 percent captures an 11-point spread on $46 — worth about 42 cents. The surplus grows as balances and minimums fall, and the cumulative reallocation is what eventually produces $442. But the spread never applies to $13,670. Applying the rate difference to the whole debt rather than to the marginal dollars whose destination changed is the arithmetic error behind inflated expectations.
Both methods end on the same day, roughly
Total interest is driven mostly by how long the money is borrowed, and the payoff date is set by the budget, not the ordering. Forty-three months against forty-two is the entire timing difference.
What widens the gap, and by how much
The $442 is not universal. Two variables move it. The first is the spread between the highest and lowest rate, holding the same four balances and the same $500 payment, with the two middle rates spaced evenly between the endpoints. That even spacing is why the 11-point row below is not identical to the $442 base case, whose middle rates are not evenly spaced:
| Rate spread | Snowball interest | Avalanche interest | Avalanche advantage |
|---|---|---|---|
| 3 points (20.99%–23.99%) | $6,172 | $6,041 | $131 |
| 7 points (18.99%–25.99%) | $6,885 | $6,579 | $306 |
| 11 points (16.99%–27.99%) | $7,658 | $7,177 | $480 |
| 16 points (13.99%–29.99%) | $8,431 | $7,749 | $682 |
| 30 points (0%–29.99%) | $7,637 | $6,501 | $1,136 |
Where every card sits within three points of every other — common for cards opened in the same few years — the ordering is worth $131 across three and a half years. The spread has to reach into promotional-zero-percent territory before the advantage clears a thousand dollars.
The second variable is the size of the payment, and it does not behave monotonically:
| Monthly payment | Months (snowball / avalanche) | Avalanche advantage |
|---|---|---|
| $460 | 50 / 49 | $414 |
| $475 | 47 / 46 | $432 |
| $500 | 43 / 42 | $442 |
| $600 | 33 / 32 | $400 |
| $750 | 24 / 24 | $321 |
| $1,000 | 17 / 17 | $236 |
The table starts at $460 because the four minimums come to $454 in month one, and below that figure the schedule is not a payoff plan at all. The advantage peaks in the middle and falls off at both ends: at a large payment everything clears quickly, leaving little time for ordering to matter, and at a payment barely above the minimums there is almost no surplus to direct. At $460 a month, six dollars above the month-one minimums, the ordering decision is worth $414 across fifty months.
The number that dwarfs both
Take the same four debts and pay only the minimum on each, with no surplus directed anywhere. The last balance clears in month 270 — twenty-two and a half years — after $24,501 in interest, for $38,171 paid on $13,670 borrowed.
Set against the avalanche's $7,238, that is a difference of $17,263. The gap between the two accelerated orderings is $442. The payment-size decision is thirty-nine times the size of the ordering decision, on the same debts, under the same assumptions, in the configuration chosen to flatter the ordering decision.
A second comparison makes the same point at closer range. In the base scenario, adding $20 a month to the snowball makes it cheaper than the avalanche at the original payment: $7,177 against $7,238. A 4 percent increase in the payment more than offsets the entire ordering decision in the case built to make the ordering matter most.
The rate environment behind all of this
The Federal Reserve's quarterly series shows why the arithmetic changed character after 2022. From 2019 through early 2022, the average rate on accounts assessed interest sat between roughly 15.8 and 17.1 percent. It reached 23.37 percent in the third quarter of 2024 and was 22.15 percent in the second quarter of 2026 — six to seven points above where it spent the pre-2022 period.
Higher rates raise the cost of carrying any balance under either method, but they do not on their own widen the gap between the methods, which depends on the spread between rates rather than their level. What the level changes is the return on the payment-size decision: at 22 percent, every additional dollar of payment retires debt compounding at 22 percent.
The aggregate scale is visible in the same release. Revolving consumer credit outstanding stood at $1,351.1 billion in June 2026, seasonally adjusted. The CFPB reported that consumers were assessed $160 billion in interest charges in 2024, up from $105 billion in 2022.
What the statement is required to tell you
Some of the relevant math arrives every month without being requested. Under Regulation Z, at 12 CFR 1026.7(b)(12), a credit card periodic statement must carry a repayment disclosure. It includes the warning that "if you make only the minimum payment each period, you will pay more in interest and it will take you longer to pay off your balance," an estimate of how long minimum-only repayment would take, its total cost, and — when minimum-only repayment would run more than three years — the monthly payment that would clear the balance in 36 months, along with the total cost of that faster path and the savings against minimum-only.
That figure quantifies the payment-size decision — the large one. It is per-account, so it does not address sequencing at all.
The CFPB's 2025 report found that about 15 percent of general purpose cardholders made only the minimum payment, up from 13 percent in the prior report. For that group, the ordering question is not yet live.
Two things the comparison is often taken to mean
"The snowball costs nothing." It costs something, always, unless the balance order and the rate order happen to coincide. In the scenarios modeled here that cost ran from $131 to $1,136. Describing it as free is inaccurate in the other direction.
"The avalanche saves thousands." It saves thousands only when the rate spread is very wide — typically when a zero-percent promotional balance sits alongside a high-rate card. Across general purpose cards clustered within a few points, three figures is the realistic range.
Where This Doesn't Apply
The model holds several things constant that are not constant in practice. Each can reverse the conclusion.
Fixed-term installment debt behaves differently. Student, auto, and personal loans have scheduled payments that do not fall as the balance falls, so paying one down early does not free up a minimum the way a revolving account does. The roll-down mechanic both methods depend on works only partially, and a mixed portfolio does not produce the numbers above.
Promotional rates have expiration dates, and the ordering must account for them. A balance at 0 percent for another five months is not a low-rate debt after month five. Ranking by today's rate misprices it. Neither method, applied mechanically, handles a scheduled rate change.
Federal student loans sit outside this frame entirely. Income-driven repayment, forgiveness programs, and deferment provisions can make accelerated payoff the more expensive choice, and applying an ordering to them without first checking the program terms at the Department of Education's Federal Student Aid site can forfeit benefits worth more than any interest saved.
The behavioral question is not settled by interest arithmetic. If a method is abandoned in month 9, its theoretical cost is irrelevant. Nothing above measures the probability of continuing, and a $442 advantage over 43 months is not large enough to survive a difference in adherence. That is a claim about arithmetic magnitude, not a recommendation.
Delinquency changes the priority order completely. An account approaching charge-off, about to be reported 30 days late, or carrying a penalty APR is not competing on ordinary rate or balance grounds. Secured debt whose collateral is at risk sits outside both frameworks.
Tax-advantaged and employer-matched accounts are not in this comparison. Whether surplus money should go to debt payoff at all, rather than to a matched retirement contribution or an emergency reserve, is a question the framing does not address.
Numbers to Re-check
| Figure | As used here | Where to verify | Updated |
|---|---|---|---|
| Average rate, accounts assessed interest | 22.15% (2026 Q2) | Federal Reserve, G.19 Consumer Credit, Terms of Credit | Quarterly |
| Average rate, all credit card accounts | 20.94% (2026 Q2) | Federal Reserve, G.19 Consumer Credit | Quarterly |
| Revolving credit outstanding | $1,351.1 billion (June 2026) | Federal Reserve, G.19 Consumer Credit | Monthly |
| Average APR, general purpose cards | 25.2% (2024) | CFPB, Consumer Credit Card Market Report | Roughly every two years |
| Interest assessed on credit cards | $160 billion (2024) | CFPB, Consumer Credit Card Market Report | Roughly every two years |
| Share paying only the minimum | About 15% (2024) | CFPB, Consumer Credit Card Market Report | Roughly every two years |
| Repayment disclosure requirement | 36-month figure, required when minimum-only payoff runs past 3 years | 12 CFR 1026.7(b)(12), Regulation Z | On amendment |
| Your own minimum payment formula | 1% of balance after interest posts, + interest, $30 floor (assumed) | Your cardholder agreement | On account change |
Concrete framework
- Write down every balance, its APR, and its current minimum. The APR is on the periodic statement; the minimum formula is in the cardholder agreement.
- Sum the minimums, then sum the total available monthly. The difference is the only money either method controls.
- Compare the rate spread. If the highest and lowest APR sit within about three points, the ordering decision is worth low three figures over several years.
- Locate the 36-month payment figure on each statement, which 12 CFR 1026.7(b)(12) requires whenever minimum-only repayment would run past three years. Compare the sum of those figures to the total available. That comparison quantifies the larger decision.
- Separate out promotional balances by expiration date, installment loans by fixed schedule, and federal student loans entirely. None of the three fit the model.
- Re-check every figure in the table above at the source, since rates and thresholds change on the schedules listed.
This article explains how the rules are written. It is not tax, legal, or insurance advice, and it does not account for any individual situation. Amounts and thresholds change; verify the current figures at the source listed above before acting.
Comments
Post a Comment