14

There are two main algorithms for paying off debt;

  • The snowball, where you put your extra money towards the smallest debt to knock them off the board faster
  • The avalanche, where you put the extra money towards the highest interest debt to reduce the total amount of interest paid.

Received wisdom is that the avalanche is superior in terms of total cost. However, a recent question of mine got an answer suggesting that the debt snowball works by using the freed up cash flow from the smaller debts to clear your bigger debts faster. I have read other answers that make the same claim. Further, I can't find a blog post that explicitly discusses the cash flow angle so I'm wondering if that has been accounted for.

Does this actually work out? Does freeing up your cash flow actually boost the speed of debt payment?

17
  • 9
  • 2
    Do you think that in the avalance method you don't use the freed up cash from the larger debts to pay off the smaller debts, once you've completed the larger debt?
    – Brondahl
    Jun 30, 2023 at 8:17
  • 2
    Ah, I see. Yes, it makes the maths more complex, but (as the other answers have detailed) yes it has been accounted for, and no it doesn't change the answer.
    – Brondahl
    Jun 30, 2023 at 9:32
  • 3
    "Received wisdom is that the avalanche is superior" No, simple math will tell you this.
    – Glen Yates
    Jun 30, 2023 at 14:21
  • 2
    Also, anyone following Dave Ramsey's advice should consider his potential conflicts of interest. I mean, are you really sure you want to take advice from someone who gets kickbacks from financial advisors and pushed back against those advisors being forced to act as fiduciaries for their clients?
    – JimmyJames
    Jun 30, 2023 at 18:03

11 Answers 11

20

Barring some edge-case complications (see below), debt size is a non-factor in the math of efficient debt repayment. Having a $100,000 loan at 5% is equivalent to having a hundred $1,000 loans at 5%.

Each dollar of debt at 5% costs you $0.05/year (simple interest). Each dollar of debt at 8% interest costs you $0.08/year. How much debt you have at each rate doesn't matter, all that matters is that for each extra dollar you can put to the 8% debt instead of the 5% debt you save $0.03 in interest/year. Conversely, each dollar extra you pay towards a 5% debt while you have an 8% debt outstanding costs you $0.03/year.

If you have extra money and you have a savings account (1% interest) and a high-yield savings account (4% interest) then you'd put your extra in the account that earns you more interest, because you want the most benefit for each extra dollar. With debt, you're the one paying the interest so the most benefit comes from paying off the highest rate first.

I have repaid a significant amount of debt and always put extra toward highest rate first, I don't feel the snowball approach offers significant psychological advantage but it's popular for a reason. I think viewing debt as one big pile and focusing on best use for each dollar, watching the pile decrease at a faster rate is better for anyone who has decent financial health/discipline.

Some common complications that may affect repayment approach:

  • Tax deductible interest (should calculate 'effective' interest rate after tax benefit for repayment decisions)
  • Pre-payment penalties
  • Variable rates
4
  • Keep in mind that tax-deductible interest might not matter in many situations (eg., the US has several scenarios in which actually deducting interest might result in a higher tax burden). Tax-deductibility is a factor, but be sure that you should deduct the interest given the rest of your tax scenario.
    – minnmass
    Jul 1, 2023 at 3:26
  • @minnmass Yep, that's why I said you'd calculate effective interest rate after tax benefit, if there's no tax benefit then the effective rate is the interest rate with nothing else to figure. Takes a good bit extra figuring since all the itemization up to the standard deduction only serves to enable deductions beyond that point. Kind of nice that it's a non-issue for most people now.
    – Hart CO
    Jul 1, 2023 at 4:29
  • The only thing I think I'd comment here is that snowball may be more effective when someone is not good at keeping track of payments - if you keep missing payments and getting charged fees, getting rid of small debts might be beneficial. Otherwise, yeah, highest interest first is always going to come out ahead
    – lupe
    Jul 2, 2023 at 20:54
  • @lupe Agreed, if you struggle to make required payments then there are bigger issues to address before efficient use of extra payments.
    – Hart CO
    Jul 3, 2023 at 14:54
42

Mathematically, if you apply the same total capital to debt (meaning you take the additional cashflow after paying off a debt and apply it to the next one), avalanche will always result in less interest paid because you're paying down higher-interest rate debt first.

But that requires that the total capital that goes to debt (total of principal and interest) is the same in both cases until the debt is paid off. The main argument I (and others) use for snowball is the psychological benefit of getting smaller balances out of the way and focusing more energy on ones that can be paid off in a shorter time frame. For some (not everyone), that momentum and sense of accomplishment can actually accelerate (or at least not slow down) the debt payment and can (again, in some cases, not all) reduce the amount of interest paid. Plus you risk making mistakes by having lots of loans lingering, like missing a payment, which adds late fees that could completely negate the interest savings, instead of just knocking out the small ones, freeing up the cash flow and getting them off your radar.

Yes, if you have the financial discipline to consistently keep chipping away at the highest rate debt, then you will pay less interest. But depending on the amount of debt, the difference in interest is not as significant as the amount of debt actually paid off.

Many here will argue that finances shouldn't be driven by feelings, and that's fine - but personal finance in my experience is MUCH more about behavior than mathematics, so it seems reasonable to me to focus more on behavior that works rather then the "optimal" path that may or may not lose momentum without some sense of progress along the way.

18
  • 1
    From a quick search the average credit card interest rate is about 24% currently and average credit card debt in the US is over $7k. If the interest rates don't vary too much then maybe it's worth paying off a lower rate first, but it's unconscionable to recommend the snowball method to someone with significant credit card debt.
    – Hart CO
    Jun 30, 2023 at 2:50
  • 6
    @HartCO If they have genuinely significant credit card debt, then you can roll all the high-interest debt into a lower interest loan, and pay that off instead, but that's changing the interest rate, which has all sorts of other traps, like... freeing up your credit for even more debt.
    – Nelson
    Jun 30, 2023 at 6:00
  • 2
    @HartCO That's very fair and I actually would suggest paying off higher-rate cards first in some cases if the interest savings were more significant. Usually what I see is people being pedantic the other way - focusing on a large 5% student loan for years while letting a small 4% retail loan linger just because "the math says so".
    – D Stanley
    Jun 30, 2023 at 14:10
  • 2
    @JimmyJames: From my experience: cash flow/payment safety. Those who are in problematic debt situations rarely have savings or financial fall backs in case of any issues arising, be they external or internal. And missing a payment can have significant draw backs for them (e.g. late fees, as well as additional interest2, or other penalties such as eviction or repossession), which can lead to even bigger issues. Closing out a small loan, and eliminating a mandatory payment can increase flexibility if done right (e.g. more money for other loans, effectively a small "transitive" emergency fund).
    – sharur
    Jun 30, 2023 at 21:18
  • 1
    @JimmyJames And that's a perfectly valid scenario. What worked for me was focusing on the "annoying" small debts by getting rid of them, cutting spending to the bone, and working on knocking the smallest debts first. None of them were high interest cards, so I didn't minimize interest but the difference was not significant compared to the momentum of "getting things done". If instead I focused on the largest rate debt (student loan by a small amount) and let the smaller rate debts linger, I'm not sure I would have had the same motivation.
    – D Stanley
    Jun 30, 2023 at 21:58
12

The theory of freeing up cash flow doesn't work, as well elaborated. There is however one small advantage to the snowball method.

Besides the interest cost, there's a transactional cost to every debt. The time you spend tracking it, budgeting for it, and making the payments.

The snowball approach provides the lowest transactional cost, i.e. the least time spent tracking and repaying the debts. Its monetary cost will however be at best the same (if all debts have the same rate) or higher.

In practice, the snowball approach isn't terrible, because the largest loans, like a home mortgage, tend to have lower interest rates. Smaller ones like payday loans tend to have the highest rates, since the lender also faces transactional costs and risks. (This applies to personal finance; in business lending, it can be reversed.) In some cases both approaches will give the same repayment schedule.

Snowball is never cheaper (unless there are transfer fees, which isn't normal). It's never faster for getting to zero debt. But it can save some personal time if there's a lot of small debts that have to be managed separately.

Even in that case, it's more sensible to mix the approaches, e.g. clear debts under some "nuisance level" like $100 right away, then sort larger debts in terms of highest APR paid first.

19
  • 5
    Perfect answer and the only one to bring up that pure dollar value isn't the only type of "cost" one can optimize for
    – Hobbamok
    Jun 30, 2023 at 13:11
  • 1
    I'm sorry, but this is nonsense. How much of a cost do you really think tracking the balance and payments on a single loan amounts to, particularly these days when you can set up an automatic payment every month? It is ludicrous to suggest that someone should pay real costs in extra interest to save on the intangible cost of clicking a "confirm payment" button one extra time per month.
    – Nobody
    Jun 30, 2023 at 18:16
  • 1
    @Nobody At $7.25/hour value of time, assuming 2 minutes/month to track and 2 minutes to pay, a 12-month $100 debt at 5% APR will cost $2.73 in interest and $5.8 in the time cost of tracking and making monthly payments.
    – Therac
    Jun 30, 2023 at 18:55
  • 2
    @Therac The cost of the interest scales with the size of the debt; the cost of your 4 minutes per month does not. Try the math again with a more realistic size for the principal.
    – Nobody
    Jun 30, 2023 at 19:31
  • 3
    @njzk2 The illiquidity of small amounts of time cuts both ways: suppose you're short a few minutes on the payment day, and slip the deadline by just a bit... The odds of that are higher with 100 debts than with 10.
    – Therac
    Jul 1, 2023 at 2:51
6

When someone reaches the stage of needing to have "a method" for paying off debt then that is usually because they have reached a state of being heavily indebted. Of having a debt repayment burden that is negatively affecting their quality of life.

The main advantage of the snowball method is that it more rapidly frees up income. That income can then be used to avoid needing to go further into debt by covering emergencies and daily living expenses. So you more rapidly reach the state of having comfortable levels of debt.

At the end of the process, yes, you'll not have as much money as in the avalanche method but it may take many years to clear your debt. For many people, the important thing is to get from the state where paying back debt is negatively affecting quality of life to the state where it isn't as quickly as possible.

If, on the other hand, you're already in the position of being financially comfortable but are just looking to manage your debts in a way that works out best in the long run then the avalanche method is always the right choice.

3
  • Good point, individual situations really drive appropriate strategy, debt snowball seems best suited for those that are in rough shape financially. I'd argue though, that anyone with debt should have a strategy/plan. Even if someone just has a mortgage they should consider if it's worth paying extra vs the minimum.
    – Hart CO
    Jun 30, 2023 at 16:30
  • @HartCO I guess it depends on how loosely you interpret the idea of having "a method" to pay off debt. I think people should look seriously at their finances from time to time, and certainly consider the long term impact of paying extra on a mortgage for example. But I wouldn't call that "a method". Jul 1, 2023 at 7:08
  • 1
    This is the best answer in this whole thread. It's the only answer which really convinced me that for some people, in some circumstances, snowball is absolutely the best approach mathematically (and not just physiologically), because they really need the extra income to stabilise their life financially and possibly prevent themselves going further into debt because minimum payments meant they didn't have enough left over to live.
    – niemiro
    Jul 1, 2023 at 21:52
4

The money "freed up" by the snowball method was already being used to pay down debts.

Ie, imagine you have a pair of debts:

10k at 1% per month; min payment 200$

1k at 0.1% per month; min payment 100$

Free cash flow of 100$ (and 300$ which you are using to pay off the two debts; if you fail to make min payment, you get penalized harshly).

Snowball says "pay off the 1k debt first, as it is the smallest". And it is true that after paying off the 1k debt, you now have 200$ in cash flow.

But that min payment on the 1k isn't "lost money", it is being used to pay off a debt already. You are forced (by the min payment rules) to put it against the lower interest debt, but putting more money into the lower interest debt isn't the optimal solution.

Instead, you throw the 100$ at the larger 10k debt. Every 100$ you throw at it reduces its interest by 1$ per month, while 100$ thrown at the 1k debt lowers its interest by 0.1$ per month.

While it takes longer to clear a debt this way, you end up with more money at the end of the process.

You can flip everything on its head. Imagine if instead of debt, you had 2 investment chances. One can soak up to 1k, and returns 0.1% per month in yield. The other can soak up to 10k, and returns 1% per month in yield.

The second is clearly a better investment option.

High interest debt is toxic.

There are real advantages to snowball beyond the psychological.

  1. Every debt requires effort to keep track off. Fewer debt, less tracking load.

  2. Free cash flow can help avoid adding more debt, and min payments aren't usually linear. By clearing a debt completely, you can maybe avoid having to borrow to handle unanticipated problems.

  3. If you do need to renegotiate or whatever, fewer counter parties makes such renegotiation easier. Like, if you owe 1k to 10 people, getting a 1 month delay on payments requires 10 different deals. If you owe 10k to 1 person, getting a 1 month delay on payments requires 1 deal.

and the psychological boost can be real (victory! instead of a battle of attrition).

And for most human endevours, motivation and psychology is insanely important.

Some people can be motivated by the math of the avalanche, and how they are smarter and doing it optimally than the "easy win of snowball". But often the snowball is pretty damn close to avalanche in costs, and doesn't require fancy tricks to convince yourself you are winning.

On the other hand, an annual 30% interest debt is insanely worse than a 5% interest debt. If the difference is 16% and 15%, go snowball the smaller one -- if the difference is huge, focus on the one that really matters, the big interest one.

1
  • For many people, the psychological advantage is the MOST important one.
    – barbecue
    Jul 1, 2023 at 17:48
2

First you check if there are any rules beyond just interest payments. Credit card debt in the Uk, you often have high interest rate if you just make minimum payments and a lower rate if you pay off more, so you want to pay enough to get the lower interest rate. Or shops give you “interest free credit” for two years with huge penalties if you don’t pay back in time. Make sure you don’t fall into any traps.

Apart from that, paying the highest interest loan first will be most effective. So the “snowball” method isn’t going to win. It’s psychological only. If you use snowball and continue paying back debt, that’s a lot better than using avalanche and giving up.

And use some common sense. If you owe $1000 at 15.1% and $20,000 at 15.3% then by all means pay back the $1000 first. If the $1000 are at 3% then don’t. Or pay back a $100 debt to get it out of the way, no matter what the interest rate is.

1

I actually wrote a python script (pastebin link) to simulate the two methods. The capital, minimum installment and interest are randomized independently and the interest is compounded monthly, but the results point to a fairly obvious answer.

On average, even when accounting for the freed up cash flow, the avalanche method is cheaper by around 1% of the capital, and the distribution is skewed towards avalanche being cheaper. There's hardly any difference in total payment time, with the average being about half a month in avalanche's favor.

With that, I feel confident in saying that the avalanche method is indeed superior barring edge cases.

10
  • 3
    No - if you have the same total capital to put towards debt, avalanche always minimizes the interest paid. Snowball doesn't change the math, it (in some cases) changes the behavior.
    – D Stanley
    Jun 29, 2023 at 18:41
  • 4
    @HAEM No, the premise is that you have $X each month to allocate to your debt however you want. Once you've paid off the low-interest snowball debt, you just switch to putting that $X to the higher-interest debt. Paying off the low interest debt doesn't give you any more disposable income to pay off your loans, you still only have $X to spend each month. You don't now have $X+$Y to spend on debt after the smaller debt is gone. Jun 29, 2023 at 18:56
  • 8
    @HAEM It should be clear that snowball cannot be cheaper (it's a psychological trick), so your script cannot work correctly. A quick glance suggested at least 2 potential bugs: 1) if sn_tot <= av_tot: counts snowball as a win even if they are just equal (e.g. if they accidently pay in the same order) and/or have a floating point difference (e.g. summing in the 12th digit may not be exact) 2) in simulate_paying, if remainder > 0 and last >= 0: seems to only pay off one remaining debt. If there's a remainder after that single payment, it won't get used, same for the initial lump sum.
    – Solarflare
    Jun 29, 2023 at 23:20
  • 2
    @Solarflare Thanks. Fixing those made the snowball wins go away.
    – HAEM
    Jun 30, 2023 at 6:09
  • 2
    @NuclearHoagie Most debts come with either a min payment or an amortized payment requirement. Clearing a debt clears that min payment/amortized payment, which is what HAEM is confused by. You have 100$ disposable and 300$ you are paying in debt payments (if you fail to keep up you get penalties). So you clear a debt and its 50$/month in required payments is freed up, giving you 150$. The error here is that the 50$/month was already being paid to a debt, it isn't "real" additional debt payments.
    – Yakk
    Jun 30, 2023 at 14:46
1

An entity acting on pure reason, such as an organization with a board of directors of competent investment banker types, will never choose "snowball" unless there is a business incentive to do so (e.g. something wacky about contract terms). Avalanche is the most prudent and dollar-efficient method.

The "snowball" method is an emotional tool. It is based on the premise that the smallest debt is probably smaller than the highest-interest debt, so you will sooner achieve the "contact high" and sense of accomplishment of crossing one off. That can be important to motivate some people to stay in it for the long haul.

I cannot quantify the value of that emotion.

I might say that making emotion based decisions in lieu of a financial education is what got a lot of debtors into trouble in the first place! So I don't think replacing "emotion" with "other emotion" is the right play, better to replace it with "financial education". E.G. Buying a house (on a mortgage) may be more rational than renting. Fueling and maintaining an old "beater" car may be less wise than financing an EV. Oh boy, here come the downvotes :)

Unfortunately, some financial advisors recommend snowball on a one-size-fits-all basis, believing the emotion benefits everyone equally, and that indebtedness only arises from lack of financial education. I wouldn't say so. Nor would I say lack of financial education can't be corrected. I might even call it a financial advisor's job :) I'm in trouble now :)

1
  • Humans are fundamentally emotional beings. Ignore that at your peril.
    – barbecue
    Jul 3, 2023 at 1:47
-1

Why does the snowball method work? The answer is behavioral, not mathematical. As much as we like to think we are driven solely by logic, very few of us are.

Here is a simple example that explains how debt snowball can be superior.

Lets say one has two debts:

  • 6K @ 14% interest
  • 300 @ 0% interest

You have an extra $350 this month after paying minimums on both debts. However, there is a sporting event you would like to go in a week's time. Ticket cost: $300.

If one were to use the avalanche method, they would pay on the larger loan. However, reducing the loan by a relatively small amount does not generate a lot of pleasure.

Going to the sporting event generates a lot of pleasure, and they would need the extra 50 (or more) for parking, drinks and food. I would argue that most avalanchers would tend to choose the sporting event. This would quite possibly leave them in a worse situation as they may spend more than 50 on other associated costs.

The snowballer is left with a choice. Go to the sporting event or pay off the debt and have $50 left over. A likely choice is for this person to pay off the debt, and use the 50 to go out and watch the game at a pub or something.

And this is really the crux of debt pay off: small choices that leads to big change. Paying of debt is not fun. It is hard to stay motivated. Seeing large loans decrease by small amounts is underwhelming. Seeing debts crossed off your list is much more motivating.

And I will ask the obvious question. If one is so logical, why now the need to pay down debt?

Even though I am a bit of a math guy, I made many bad choices that lead to excess spending and bad debt. These were emotional although I'd like to think they weren't. For me the snowball method worked, and I assert that it will work well for most people.

4
  • Needs evidence (not anecdotes) that people who reduce debt via snowball don’t actually reduce debt, but instead are actually crypto-spendthrifts.
    – RonJohn
    Jul 5, 2023 at 23:40
  • @RonJohn the evidence is pretty obvious. How many people actually transition from being into consumer debt to having none? I would assert very few. There are the disciplined that never get in consumer debt in the first place (or well within their income) and those that remain living from paycheck to paycheck. Actually getting out of debt is weird.
    – Pete B.
    Jul 6, 2023 at 14:08
  • We were $30K in CC debt (plus auto and mortgage) and dug our way out with alternating snowball and avalanche methods, so must be weird.
    – RonJohn
    Jul 6, 2023 at 17:59
  • Mathematically, if you have 300 @ 0% interest, forget this one and focus on the one with 14%. Depending on the nature and the amount of cash flow the person is expected to get, he/she can judge/justify unnecessary spending. To me. I'd rather pay the debt than go for pleasures that piles more debt :)
    – NoChance
    2 days ago
-3

I call my method the "zombie killing" method.

Each debt is a zombie, advancing towards you. Each zombie requires X bullets per unit time (say, per month) to keep it at bay. Each month, I have just so many bullets.

For each zombie, if X is the bullets required to keep it at bay this month, Y is the bullets needed to kill it so it doesn't get up ever again.

The more bullets I have each month, the easier it is to keep zombies at bay, the faster I kill them, the less stressed I am.

So I want the technique which frees up, or allows me to allocate, more and more bullets.

So I concentrate on the easiest-to-kill zombie, defined as the most X for the least Y bullets.

For example,

  • Zombie A needs 50 bullets per month to keep at bay, and 2000 bullets to kill = ratio 40
  • Zombie B needs 100 bullets per month to keep at bay, and 10000 bullets to kill = ratio 100
  • Zombie C needs 75 bullets per month to keep at bay, and 2500 bullets to kill = ratio 33.3

I'm going to concentrate on killing zombie C first. It's harder to kill than zombie A, but once it is dead, I get more bullets per month to kill the others.

I kill zombie A next. At the end I'm killing zombie B with 225 bullets per month, which is more than twice the kill rate, if I took it on first.

This method has the advantage of being far more exciting.

22
  • 2
    You just rephrased the snowball method
    – Hobbamok
    Jun 30, 2023 at 13:13
  • 1
    @Stewart Agreed for the most part, efficiency is mostly an issue when there are significant differences in interest rate. Watching a $10k credit card balance balloon due to crazy interest rate while knocking down the remainder of a smaller loan with a lower rate could easily derail someone. Congrats on your progress too!
    – Hart CO
    Jun 30, 2023 at 17:02
  • 1
    Some pretty dumb downvotes on this.
    – barbecue
    Jul 1, 2023 at 17:47
  • 1
    @HAEM Any method that ignores the psychological aspect is inadequate for the majority of people. Most people in debt are not math nerds.
    – barbecue
    Jul 3, 2023 at 1:46
  • 1
    Are the bullets the minimum payments?
    – RonJohn
    Jul 5, 2023 at 23:44
-4

None of them.

My method.

  1. check how much money can spend for pay off
  2. calculate intrest rate for every debt
  3. calculate commision refund if pay off debt
  4. calculate commision refund if use previous commision refund to pay off
  5. repeat 4 until commision refund is negligible
  6. repeat calculations from points 3-5 for all debts
  7. choose to repay this debt after wich my total monthly intrest is the lowest.

Example:

  1. 50k total taken, commision 6k, intrest 20%, 5y total 4y left
  2. 100k total taken, commision 20k, intrest 20%, 10y total 9y left
  3. 300k total taken, commision 30k, intrest 10%, 30y total 22y left

To pay of first is 2. - every 10k is 2k commision refund wich means 10k pays around 12,5k.

1
  • 9
    what is a commission refund? Jun 30, 2023 at 12:22

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .