Saving for your child's education in Nepal: the 18-year plan

A cousin had a baby in Magh and, two weeks later, sent me a voice note at midnight asking one question: how much does he need to put away every month so the kid can study whatever she wants in eighteen years. Not the sentimental version. The number.

The honest answer is that the number is smaller than he feared and the timing matters more than the number. A parent who starts the month the child is born can hit a serious education corpus on a modest monthly amount. A parent who starts when the child is in Grade 3 has to save almost three times as much for the same goal. The whole game is the eighteen years, and most people waste the first eight of them.

First, what are you actually saving for?

The target sets everything else, and the range is enormous. A quick map of where Nepali education money goes (the full birth-to-Class-12 breakdown is in the cost of raising a child post; here we care about the big-ticket end):

• A domestic bachelor's degree. Kathmandu University's BBA, engineering, and computer science programs each run about Rs 9,15,000 for the full program, per the KU fee structure. Tribhuvan University's public engineering (IOE) is far cheaper at around Rs 3,18,000. Private BBA colleges land Rs 5–10 lakh.
• MBBS. The Medical Education Commission caps it at Rs 40,23,250 inside the Kathmandu Valley and Rs 45,95,720 outside, per the MEC fee notice. This is tuition alone, and it is the single number most Nepali parents quietly save toward.
• Studying abroad. The big one. Nepalis spent Rs 138.48 billion on overseas education in FY 2024/25, up from Rs 100 billion two years earlier, per NRB data reported by the Kathmandu Post. A four-year Australian degree runs AUD 180,000–340,000 with living costs; the UK now requires proof of £1,171–1,529 a month in living funds on top of tuition, per GOV.UK. Converted at roughly Rs 138 to the US dollar, even a modest foreign degree clears Rs 60–80 lakh, and the well-known destinations run well past a crore. The shorter pre-departure version is in the study-abroad runway post, and the legal way to wire the fees is here.

So the target is not one number. It is a band: roughly Rs 10 lakh for a frugal domestic path, Rs 40–50 lakh for medicine or a budget foreign degree, and Rs 1 crore+ for the Australia/US/UK route. Pick the one your family is realistically aiming at, then build the plan around it.

The enemy is fee inflation, not the sticker price

Here is the trap. You see today's fee, you save toward today's fee, and eighteen years later the fee has tripled.

Nepali school fees have risen faster than general prices for a decade. Kathmandu Metropolitan City once approved hikes of 27% in a single year when headline inflation was 4.4%, and district committees have historically allowed 22–50% depending on the school grade. Those are older figures, and the 2026 news is more about municipalities capping fees than about hikes, so treat them as illustration rather than a current rate. The planning number the cost-of-raising post settles on, drawn from the NRB inflation reports, is 8–12% a year for education costs more than three years out.

Sit that next to what savings vehicles pay. A one-year individual fixed deposit currently earns about 4.84%, near a 13-year low, per NRB data reported by Fiscal Nepal. An FD growing at 5% against fees growing at 10% is going backwards in real terms every single year. That gap is the entire reason this post exists.

The vehicles, ranked for an 18-year goal

Vehicle	Typical return	Best used for
Child savings account	~3–5%	The everyday account, not the corpus
Fixed deposit / recurring deposit	~5%	The last 3–4 years before the money is needed
Mutual fund SIP (equity)	~8% (price-only NEPSE, 22-yr)	The first 10–15 years of growth
Child endowment insurance	well below the headline bonus	Protection, not investment

Child savings accounts ("Balbalika Bachat" and similar) exist at most banks and pay roughly 3–5%, with one or two products at the higher end. They are fine as the account the money flows through and useless as the place an 18-year corpus actually grows. Confirm the exact rate on the bank's own page before opening, since these move.

Fixed and recurring deposits are safe and predictable, which is exactly what you want at the end, not the start. A recurring deposit is the disciplined monthly version; an FD ladder keeps the money liquid. Park the corpus here in the final few years so a NEPSE dip the year before admission cannot derail you.

Mutual fund SIPs are the workhorse for the long stretch. The minimum is just Rs 1,000 a month, per Siddhartha Capital's SIP, and several AMCs (Siddhartha Capital, NIMB Ace, NMB Capital) run them. The return assumption is the honest hard part: the NEPSE index has compounded at about 8.69% a year over 22 years on price alone, before dividends, per Investopaper. It is also volatile, with single years down 18% or more. No Nepali fund publishes a reliable long-run CAGR, so the responsible move is to plan at 8% and treat dividends and anything higher as upside. The full setup is in the how to start a SIP post.

A note on CIT: the Citizen Investment Trust has no child-specific or education scheme. Its products are retirement and pension oriented, so it does not belong in this plan.

Why "child education" insurance usually loses

Every Nepali life insurer sells a child plan: Nepal Life's Baal Shikshya Beema, LIC Nepal's Baal Unnati, IME Life's Bal Ujjwal, and others. The structure is the same. The parent is the proposer and payer, the child is the life assured, and the standout feature is a premium waiver: if the parent dies, the policy continues and the full benefit is still paid at maturity. That protection is real and genuinely useful.

The investment side is where it falls apart. The maturity payout is the sum assured plus accrued bonus, and the bonus is declared per Rs 1,000 of sum assured, not on the money you actually paid. For FY 2080/81 those bonus rates ran roughly Rs 22 to Rs 90 per thousand across insurers, per Beema Post. Crucially, that bonus is simple, not compounding: it is calculated on the sum assured every year and does not earn a return on itself. So a headline that looks like 6–8% is nothing like a SIP compounding at 6–8%, and the effective return on your premiums ends up well below the headline once you account for the chunk of early premiums that vanished into commission.

The cleaner separation, the same one the health and life insurance basics post argues, is to buy protection and growth separately: a cheap term life policy so the family is covered if you die, and a separate SIP for the corpus. You get full cover and full compounding, instead of a watered-down version of each. It is telling that as bank rates fell below 6%, savers poured around Rs 7 billion into fixed-deposit-style insurance products chasing yield; the marketing works even when the math does not.

The actual 18-year math

This is the part the cousin wanted. The figures below are calculations, using the standard monthly-investment (SIP) future-value formula, with the return assumed and stated. They are not guarantees, and real returns will wobble year to year.

Monthly saving needed to hit a target in 18 years (216 months):

Target corpus	At 5% (FD-like)	At 8% (equity-tilted)
Rs 25 lakh	~Rs 7,150/month	~Rs 5,200/month
Rs 50 lakh	~Rs 14,300/month	~Rs 10,400/month
Rs 1 crore	~Rs 28,600/month	~Rs 20,800/month

Two things jump out. The 8% column needs over a quarter less every month than the 5% column for the same goal, which is the entire case for an equity SIP over an FD across a long horizon. And even a crore, the number that sounds impossible, is about Rs 20,800 a month from birth at 8%, less than many Kathmandu households spend on a car EMI.

Now the cost of waiting. The same Rs 50 lakh goal, at 8%, by how late you start:

Start when child is	Years to save	Monthly saving needed (8%)
Newborn	18	~Rs 10,400
Age 4	14	~Rs 16,200
Age 8	10	~Rs 27,300
Age 12	6	~Rs 54,300

Waiting until the child is eight pushes the monthly cost of the same goal to more than two and a half times the birth-start figure, from about Rs 10,400 to Rs 27,300. The first Rs 10 lakh is harder than the next post makes the same point in general: early money compounds longest, so the rupees you save in year one do far more work than the rupees you save in year fifteen. By the rule of 72, money at 8% doubles in nine years, so a lump saved at birth roughly quadruples by age 18; at 5% it barely doubles.

The glide path: equity early, safe late

Putting it together into a plan that actually survives a bad market year:

1. Years 0–13: grow it. Run a monthly SIP into an equity-tilted mutual fund at the amount your target needs. This is the long compounding stretch, and short-term NEPSE swings do not matter when you have a decade-plus to go.
2. Years 14–16: start de-risking. Begin moving the accumulated corpus out of equity and into FDs or recurring deposits. You are now protecting a number you have already built, not chasing growth.
3. Years 17–18: fully safe. The whole corpus sits in deposits maturing around when the fees are due. A market crash the year before admission should not be able to touch it.

This is the opposite of how most families do it, which is to keep everything in an FD the whole way (too slow) or to stay in shares right up to the admission month (too risky at the end). The glide path takes the growth when you can afford the volatility and the safety when you cannot.

The mistakes that quietly wreck an 18-year plan

The math above assumes the plan survives eighteen years. Most do not, and the reasons are predictable:

• Stopping the SIP in a bad year. NEPSE can fall 18% or more in a single year, and that is exactly when people panic and pause contributions. A long SIP relies on buying more units when prices are low; stopping then locks in the loss and skips the cheap buying. The discipline of paying it on salary day, before you can flinch, is half the strategy.
• Raiding the fund for something else. A car down-payment, a sudden medical bill, a plot that "won't last." An education fund only works if it is ring-fenced. Keep it in a separate account with the child's name on it, mentally and literally off-limits.
• Leaving everything in an FD "to be safe." Safe against volatility, unsafe against fee inflation. Over eighteen years the slow erosion of a 5% deposit against 10% fee growth does more damage than any single market crash.
• Not raising the contribution with your salary. The plan is set at today's income and never touched. Every increment should lift the monthly amount, or inflation eats the target from underneath.

None of these is a market problem. They are behaviour problems, and behaviour is the part you actually control.

What you actually need to know

• Start the month the child is born. Time is the single biggest lever; starting at age eight nearly triples the monthly cost of the same goal. If you are already late, the answer is to start now at the higher number, not to wait further.
• Use a SIP for the long stretch, FDs for the final years. An FD alone loses to fee inflation over 18 years. An equity SIP gives it a fair fight, and the glide path protects the corpus once you are close.
• Buy protection and growth separately. Skip the "child education" endowment plan as an investment; its simple bonus underperforms a compounding SIP. A term policy plus a SIP covers both jobs better.

Got a specific target and timeline you want the monthly number for? Email parjanya57@gmail.com.

This post is part of the Nepal Money Basics guide — the big-ticket-decisions section.