Debt Avalanche

The debt avalanche is a debt-repayment method in which the borrower lists every debt by interest rate from highest to lowest, pays the required minimum on all of them, and applies every additional dollar available to the highest-rate balance until it is gone. The payment freed by that account is then added to the amount attacking the next-highest rate. Balances play no part in the ordering.

It is the mirror image of the debt snowball, which orders by balance instead, and the two methods are identical in every other respect. Both pay minimums on everything, both concentrate all spare money on one target, and both roll each freed payment forward. So the choice between them is a choice about one variable, and the avalanche is the version that answers it with arithmetic.

Like the snowball, this is popular-finance vocabulary rather than a defined term. No federal agency, banking regulator, or standard-setting body defines it, so the details vary between sources; what is consistent is the ordering rule.

Why it wins, stated as the thing that actually determines the margin. Hold the total monthly payment fixed. Every dollar you move from a lower-rate balance to a higher-rate one saves the difference between those two rates for every month it stays there, and it saves nothing else. That makes the size of the avalanche's advantage the product of two figures and only two. The first is the spread between the rates involved. The second is how long the expensive balance would otherwise have sat, which is roughly how long the cheaper targets ahead of it take to clear. Where the spread is wide and the cheap balances are large, the gap is substantial. Where the rates are clustered inside a few points, the two orderings converge and the choice is mostly about which plan gets finished. Neither the number of debts nor their size enters into it except through those two terms.

The honest cost of the method, and it is not the usual point about motivation. The snowball orders by a figure that moves in one direction only. A balance you are attacking falls, and it does not spontaneously rise. The avalanche orders by a figure that can move either way, and Regulation Z contemplates three distinct ways it can. A temporary promotional rate ends on a date fixed in advance, and the rate that follows was disclosed when the promotion started (12 CFR 1026.55(b)(1)). A variable rate moves whenever its index moves, which is expressly permitted (1026.55(b)(2)). And an issuer may raise the rate on future transactions after 45 days' advance written notice (1026.55(b)(3)), subject to a protection worth knowing on its own: that route is unavailable during the first year after the account is opened. The consequence is operational. A rate-ordered plan is a plan that has to be redrawn whenever a rate moves or a promotion lapses, and a borrower who set the order once and stopped looking may be attacking what used to be the expensive balance.

Order by the annual percentage rate, and be careful comparing across kinds of credit. What the APR captures depends on the type of credit. On a closed-end loan it folds in certain financing costs beyond interest, while on a credit card it is essentially the annualized interest rate and does not reflect an annual fee. So a card APR and an installment-loan APR are not constructed the same way, and ranking them against each other on the printed figure alone can put them in the wrong order. Where interest on a debt is deductible the ranking figure is the after-tax rate rather than the nominal one, which currently reaches acquisition-debt mortgage interest under IRC 163(h)(3), student loan interest under IRC 221, and, for tax years 2025 through 2028, qualifying car loan interest under IRC 163(h)(4). A deduction someone cannot use because they do not itemize changes nothing, so this matters only where the deduction is actually claimed.

Three debts where the rate is the wrong ranking variable. A deferred-interest balance, the kind advertised as no interest if paid in full within a set number of months, is not priced by a rate you can rank at all. Interest is accruing throughout and is waived only if the whole balance clears in time, so what governs is a deadline, and missing it charges the accumulated interest retroactively to the purchase date. That balance belongs at the front of the queue regardless of the rate printed on the offer. A federal student loan being carried toward cancellation under an income-driven plan can be made worse by prepayment, because extra payments reduce the amount eventually canceled rather than saving interest. And a secured debt carries a consequence of default that its rate does not express: the lender can take the collateral. A car loan at a single-digit rate sits at the bottom of a rate-ordered list and at the top of the list of things whose non-payment removes your way of getting to work.

What the research says, which is narrower than the comparison it is usually cited in. The paper cited on both sides of the snowball-versus-avalanche argument is Kettle, Trudel, Blanchard and Häubl's "Repayment Concentration and Consumer Motivation to Get Out of Debt," in the Journal of Consumer Research 43(3) in 2016, and its subject is repayment concentration. The finding is that concentrating repayment on one account rather than dispersing extra money across several increases motivation to keep repaying, most strongly when the target is the smallest balance. Both named methods are concentrated, so the finding chiefly argues against the intuitive third option of paying a little extra on everything. It does not establish that either named ordering beats the other. A pair of six-month adherence percentages circulates in side-by-side comparisons and is attributed to that paper, which reports no adherence rates of that kind, so it is not repeated here.

One place the law already runs the avalanche for you. Where the debts are separate balances on a single credit card, at different rates, the ordering is not yours to choose. 12 CFR 1026.53(a) provides that when a consumer pays more than the required minimum on a credit card account, the issuer must allocate the excess "first to the balance with the highest annual percentage rate and any remaining portion to the other balances in descending order based on the applicable annual percentage rate." There is one significant exception, and it is the deferred-interest case above: under 1026.53(b)(1)(i), during the two billing cycles immediately before a deferred-interest period expires, the excess must go to that balance instead.

List every debt with its annual percentage rate, its balance and its required minimum. Add the minimums together and subtract the total from what you can put toward debt each month; the remainder is the amount that attacks one target. Point it at the highest rate. When that account reaches zero, add its former minimum to the attacking amount and move to the next-highest rate. Recheck the rates whenever a statement arrives, because the ordering depends on them.

A hypothetical example, built so the ordering is the only thing that differs. Interest is added once a month at the annual rate divided by twelve, on the balance at the start of the month, with the payment applied at the end and cents rounded to the nearest penny. Real card interest is computed against an average daily balance, so treat this as illustrative rather than exact.

Priya has two cards with identical $6,000 balances, one at 24% and one at 9%, each with a $120 minimum, and she can put $700 a month toward them. Minimums total $240, so the attacking amount is $460.

Because the balances are equal, a balance ordering cannot distinguish them and the rate is the only thing that can. Here is the whole argument in one line. Applying that $460 to the 24% card rather than the 9% card avoids $5.75 more interest in the following month ($460 × 0.15 ÷ 12, the 15-percentage-point spread), and it goes on doing so every month the money stays where it was put.

Run both orderings to the end and the gap compounds. Attacking the 24% card first clears both in 20 months for $1,454.95 of total interest. Attacking the 9% card first takes 21 months and $2,238.42, a difference of $783.47 on the same $700 a month.

Two things in that figure are worth separating. The spread did most of the work, and so did the fact that the ignored balance was large enough to keep the expensive card waiting for a long time. Change either input and the answer changes with it: two cards at 22% and 19% would produce a gap of a few tens of dollars over the same period, which is the case where finishing the plan matters more than choosing the right one.

Pros

• It minimizes total interest for a given monthly payment. That is arithmetic rather than a claim about behavior.
• It concentrates every spare dollar on one account, which is the feature the repayment research supports most clearly.
• It puts the most damaging balance at the front, which is usually the one growing fastest and the one whose minimum payment buys least.
• It shortens the plan slightly as well as cheapening it, because less of each payment is consumed by interest along the way.
• The advantage is largest exactly when it matters most, which is when one balance is priced far above the others.

Cons

• The ordering input can move. A promotional rate expires, a variable rate follows its index, and an issuer may reprice future purchases with notice, so the list needs rechecking rather than setting once.
• The first target can be the largest balance, so nothing closes for a long time and the plan shows little visible progress early on.
• Ranking on the printed rate can mislead across different kinds of credit, because a card APR and a closed-end loan APR are not built the same way.
• It says nothing about the consequence of default, so it ranks a secured debt below an unsecured one at a higher rate even though only one of them can cost you the collateral.
• Where rates are tightly clustered the advantage is small, and a method abandoned in month four is worse than either.
• Concentrating every spare dollar leaves no buffer, so an unexpected expense tends to land back on a card.