Which Debt First? Debt Avalanche vs Snowball, Priced in Dollars and Months

Run $700 a month at the same four debts and ordering by rate saves $424.52 in interest, while ordering by balance closes the first account 17 months sooner.

Fix the total payment at $700 and leave the minimums alone

Here is the list this piece runs on: a $6,800 card, a $3,100 card, a $9,400 personal loan and a $2,150 car loan. That is $21,450. The four balances and the four minimum payments are assumptions, picked to look like an ordinary household rather than a hard case. The rates are not assumptions. They are commercial-bank averages from the Federal Reserve’s G.19 consumer credit release, 2026 Q2. The big card carries 22.15%, the rate on accounts assessed interest. The smaller card carries 20.94%, the stated average across all accounts, which includes cards paid off every month. The personal loan is at 11.86%, the 24-month bank average, and the car loan at 7.14%, the 60-month new-car average.

One more assumption, and it is the one worth flagging. Every minimum here is a fixed dollar amount that never falls. Real card minimums shrink as the balance does, but no universal formula exists to cite. The issuer sets it. What is not left to the issuer is the box on the statement. Federal law requires the issuer to print how long the minimum alone would take, and what it costs. It must also print the payment that clears the balance in 36 months. Check that box against the months below.

Debt (illustrative balance)	APR	Fixed minimum	Closes, rate order	Closes, balance order
Card A – $6,800	22.15%	$170	month 28	month 33
Card B – $3,100	20.94%	$78	month 33	month 20
Personal loan – $9,400	11.86%	$215	month 40	month 40
Car loan – $2,150	7.14%	$87	month 27	month 10
Total interest paid			$6,022.00	$6,446.52
Months to debt-free			40	40
First account closed			month 27	month 10
Final payment			$172.00	$596.52
Balances, minimums and the $150 extra are illustrative assumptions. APRs are commercial-bank averages from the Federal Reserve’s G.19 release, 2026 Q2, collected May 2026. Interest accrues monthly at APR/12 on the remaining balance; minimums are fixed and are paid on every open debt; the total monthly payment is held at $700 throughout.

Sort the four by rate and the interest bill lands at $6,022

The minimums add up to $550. Add $150 and the household sends $700 every month, forever, until the last balance clears. Under the rate ordering that extra $150 goes to the 22.15% card. Every other debt gets its minimum and nothing more. Total interest over the run: $6,022.00, and the last debt clears in month 40. Note what does not happen. Nothing closes for more than two years.

The first account to close is not even the one under attack. The car loan runs out in month 27 on its own fixed $87, one month ahead of the card that has been getting the extra money all along. Twenty-seven months is a long time to see no account disappear.

Send the same $150 to the $2,150 car loan instead

Now point the extra at the smallest balance. The car loan closes in month 10. Its $87 minimum rolls into the pile, the smaller card closes in month 20, the big card in month 33. The personal loan still clears in month 40. Total interest: $6,446.52. The debt-free month does not move at all. The difference is $424.52 of interest and a much smaller final cheque under the rate ordering, $172.00 against $596.52.

Price the $424.52 against a closed account in month 10

So the trade is legible. Rate-first costs 17 extra months before anything is gone. Balance-first costs $424.52, or 1.98% of the starting balance. Both finish in month 40. That is the whole decision, and it is not a math question any more once both numbers are on the table.

The case for buying month 10 is contested. Gal and McShane, writing in the Journal of Marketing Research in 2012, examined records from a debt settlement firm and found that closing accounts predicted getting out of debt regardless of the dollar balance closed. Once the fraction of accounts closed was controlled for, the dollar amount stopped predicting anything. The link is to the co-author’s posted copy of the published paper, not the journal’s. What that does not establish is more interesting. Those clients had stopped paying their creditors and were negotiating settlements. The finding is predictive, not causal, and nobody in it was making four ordinary payments a month.

This arithmetic ignores plenty. It assumes no annual fee, no late fee, no new charge on either card, no promotional rate expiring, and no variable rate moving with prime. It ignores any tax treatment of the interest. It holds the $150 constant for 40 straight months, which is its own assumption. That same $150 is also money not going into a retirement account, where which of the two account types takes it is its own piece of arithmetic. Change the minimums and every month in the table moves.

The comparison runs one illustrative four-debt list at a fixed total payment with the minimums held flat, so a real list’s rates, its issuers’ own minimum-payment formulas, any fees and any promotional period will change both the interest figure and the month the first account closes.