Snowball vs avalanche: the fastest way to clear multiple loans in Nepal

A Kathmandu Post report from July 2026 quoted a former finance-company association president describing what happens when a household borrows from several cooperatives and microfinance lenders at once to keep its cash flow patched together: the moment those channels close simultaneously, "their cash cycles collapsed and repayment capacity disappeared." That is not a rare edge case. It is the mechanical description of what happens to a lot of Nepali households carrying a credit card, a personal loan, and a gold loan at the same time, each with its own due date and its own rate, and no order of attack.

Nepal's own numbers back this up. The banking system's average non-performing loan ratio climbed from 3.02% in mid-2023 to 5.42% by mid-February 2026; in microfinance specifically, bad loans jumped 74% year-on-year to Rs 54.08 billion by mid-April 2026, an 11.32% NPL ratio, up from 7% a year earlier. Multiple concurrent loans, each manageable alone, become unmanageable together the moment one income disruption hits.

What the two methods actually do

Both methods start the same way: pay the contractual minimum on every single debt, every month, no exceptions. The only difference is where any extra rupee beyond those minimums goes.

• Avalanche: extra money goes to whichever debt has the highest interest rate, regardless of its size. Once that's cleared, the next-highest rate gets the same treatment.
• Snowball: extra money goes to whichever debt has the smallest balance, regardless of its rate. Once that's cleared, the next-smallest balance gets the same treatment.

A real Nepal example: three loans, two orders

Take a household with a credit card, a gold loan, and a personal loan running at once, priced at realistic 2026 Nepali rates:

Debt	Balance	Rate	Monthly interest on current balance
Credit card	Rs 80,000	~24%/year (2%/month)	~Rs 1,600
Gold loan	Rs 300,000	~9%/year	~Rs 2,250
Personal loan	Rs 50,000	~8.5%/year	~Rs 354

The monthly interest column is simple arithmetic (balance × monthly rate), not a full amortization schedule, but it makes the point cleanly. Under avalanche, every extra rupee goes to the credit card, because 24% is nearly three times the gold loan's rate and almost triple the personal loan's, even though the card's balance is the smallest of the three by rupee amount. Under snowball, every extra rupee goes to the personal loan first, purely because Rs 50,000 is the smallest number on the table, even though the gold loan's 9% technically costs more per rupee outstanding than the personal loan's 8.5%.

That's the entire mechanical difference. Avalanche always chases the number in the rate column. Snowball always chases the number in the balance column. In this example, snowball would leave the 24% card sitting on minimum payments the longest, which is exactly the debt accruing the most interest per rupee outstanding.

The math says avalanche; the data says snowball often wins anyway

Avalanche's interest-savings advantage is not in dispute; it's arithmetic. Attacking the highest-rate balance first always minimizes total interest paid across the full payoff, for the same reason paying down the most expensive debt first always beats paying down a cheaper one first.

What the personal-finance industry (Ramsey Solutions being the most vocal proponent) argues is that raw interest math is not the only variable that decides whether someone actually finishes paying off their debt. The strongest empirical support for that argument comes from a 2012 study by Kellogg School of Management researchers Gal and McShane, published in the Journal of Marketing Research, analyzing 5,943 consumers. It found that people who paid off their smallest-balance debts first, independent of interest rate or dollar size, were about 14% more likely to become completely debt-free within a year than those who paid down debts in another order, an effect that grew further with more time. The finish-line effect of clearing an entire account, not just a chunk of a big one, appears to matter for follow-through.

Neither of these findings is Nepal-specific; both come from US-based analysis. But the underlying behavioural logic doesn't need a Nepal-specific study to be relevant here: a Kathmandu household staring at three, four, or five separate loan accounts is exactly the situation where momentum and motivation decide whether the plan gets followed for the next 12 months, not just whether the spreadsheet is optimal.

What Nepali loans actually cost in 2026, so you rank yours correctly

Before picking a method, know what you're actually ranking. These are realistic 2026 ranges, not the number printed on a five-year-old brochure:

Loan type	Typical 2026 rate
Credit card	~24%/year (about 2%/month), higher at some banks
Microfinance (laghubitta), new loans from Jul 2025	3-month average base rate + 3 percentage points (previously a flat 15% cap, which still applies to older loans)
Cooperative (sahakari)	No flat cap under the current NRB directive; loan rate floats with a spread capped at 6 points over the cooperative's own deposit rate
Gold loan	~5.4% to 15%, LTV capped at 75% of gold's market value
Home loan	~8% to 12% (base rate roughly 4.65-6.5% plus a 2-5 point premium)
Personal (unsecured) loan	~7% to 11%, market average around 8.5%
Auto/two-wheeler loan	~6.4% to 11%

The averaged commercial bank base rate itself fell below 5% for the first time in Nepal's banking history during the tail end of FY 2025/26, which is part of why personal and home loan rates have been drifting down across the board. That downward drift doesn't change which debt to attack first; it just means the exact numbers on your statement are worth re-checking rather than assumed from memory.

Yes, you can redirect extra payments without a penalty

Both methods depend on being able to throw a lump sum at one loan while paying only the minimum on the others, and Nepali retail borrowers can actually do this. An NRB circular bars banks and financial institutions from charging any prepayment penalty on loans below Rs 50 lakh (NPR 5 million), which covers the overwhelming majority of personal loans, credit cards, gold loans, and auto loans held by ordinary households. Above that threshold, a penalty can still apply, but only if the prepayment isn't tied to a rate or condition change. The same rule caps the variation in administrative and prepayment fees between similarly situated borrowers at 0.25 percentage points, and limits commitment fees to cases where a borrower has used less than 60% of an approved facility. In plain terms: for the loan sizes this post is actually about, nothing structurally stops you from executing either method.

Why payoff order matters beyond the interest math

A missed payment doesn't just cost interest. Once a loan crosses 90 days overdue past the reporting threshold, the lender is required to recommend the borrower for CIB blacklisting, and Nepal's blacklist numbers have been on a multi-year rising trend: more than 37,500 individuals were blacklisted in an eight-month stretch of FY 2080/81 (2023/24) alone. There's no confirmed rule that simply holding several current, on-time loans at once damages your CIB record. The real danger sits exactly where the Kathmandu Post reporting placed it: households juggling multiple loans across multiple institutions who lose one income source and watch every account go overdue in the same month, because none of them had a deliberate order of priority.

That's the actual argument for picking either snowball or avalanche and sticking to it, rather than paying whichever lender calls first or complains loudest. A deliberate order means minimums are always covered on everything, and any income shock hits a plan instead of a scramble.

Which method fits which household

• Pick avalanche if you're disciplined about following a spreadsheet regardless of how slow the early progress feels, and your rate spread is wide (a 24% credit card next to single-digit secured loans, as in the example above). The interest savings compound and the choice of what to attack first is usually obvious anyway.
• Pick snowball if you have three or more separate debts and have struggled to stick with a payoff plan before. Clearing an entire small account in month three, even if it wasn't the mathematically optimal target, is what keeps some households going through month twelve.
• Either way, never let "which method" become an excuse to skip minimums on the others. Both methods assume every non-target debt still gets its minimum payment, on time, every month. That's the part that protects your CIB record regardless of which target you're chasing.

The tier to avoid falling into

If a household is choosing between snowball, avalanche, or reaching for an instant digital microloan to cover a gap between paydays, the math changes entirely. Short-term digital lending products in Nepal have been seen pricing around Rs 40-90 per Rs 1,000 borrowed for terms of 7 to 15 days, which annualizes to a triple-digit effective rate, with an additional penalty charge layered on top for late repayment. Neither snowball nor avalanche is designed to compete with that tier; the point of both methods is to avoid ever needing it. If a gap is forcing that choice, it's a sign to revisit the emergency fund target before the next loan decision, not to add another loan to the stack.

Tracking it in Kharchapatra

• List every debt in one place with balance, rate, and minimum payment, and re-check it monthly. The method only works if the ranking is current, not the number from when the loan was disbursed.
• Tag your chosen target debt (Snowball Target or Avalanche Target) so every extra rupee you log against it is visible as progress, separate from the routine minimum payments on everything else.
• When a target debt clears, immediately re-tag the next one. The entire mechanism of both methods depends on rolling the freed-up payment forward without a gap month where it quietly gets spent instead.

What you actually need to know

1. Avalanche minimizes interest; snowball maximizes the odds you finish. Neither is wrong. The 24% Nepali credit card rate usually makes the "attack the card first" call an easy one regardless of which method you nominally follow.
2. You can legally redirect extra payments without a penalty on ordinary loan sizes. The Rs 50 lakh NRB threshold covers almost every personal loan, credit card, and gold loan in a typical household.
3. The real risk of multiple loans is a simultaneous cash-flow shock, not a paperwork penalty for holding several accounts. A written payoff order is what keeps every non-target debt current if one income source disappears.

Juggling a specific set of loans and want help ranking them? Email parjanya57@gmail.com.

This post is part of the Nepal Money Basics guide — the save-the-gap section.