Two foreign investors put the equivalent of USD 2 million into two identical Nepali companies. Same sector, same margins, same profit, same year. One takes home materially more than the other.

Neither of them out-traded the other. The difference was decided years earlier, on the day each of them chose the line item their money would enter Nepal under.

Most foreign investment into Nepal arrives as a single block of equity, because equity is the default, the approval process is built around it, and nobody asked the question at the right time. It is a decision worth taking deliberately, because Nepal taxes the four ways money can leave the country at four different rates — and one of them is taxed twice.

The four exit channels, and what each one costs

Money leaves a Nepali company to a foreign investor through four doors:

The rates above are the standard domestic positions and are subject to change by the annual Finance Act, and to reduction by treaty — verify the applicable rate for the year of payment before you rely on it.

Read the second column and the third column together, because that is where the whole argument sits.

The arithmetic nobody runs before signing

Take NPR 100 of operating profit, before deciding how the investor is funded.

Funded entirely by equity. The company pays corporate tax — 25% for most businesses, 20% for special industries, 30% for banking, insurance and a few others. At 25%, that leaves NPR 75. A dividend of NPR 75 suffers 5% withholding, or NPR 3.75. The investor receives NPR 71.25.

Funded so that the same NPR 100 leaves as interest on an approved foreign loan. The interest is a deductible expense, so it is not exposed to corporate tax at all. Withholding at 15% takes NPR 15. The investor receives NPR 85.

Same business, same rupee, roughly nineteen percent more in the investor's hands. And that is before the treaty question.

Now apply a treaty. Nepal's domestic dividend withholding of 5% is lower than the dividend rate in essentially every treaty Nepal has signed, so a double taxation avoidance agreement gives the equity investor nothing on dividends. It is a fact that surprises almost every investor who structured through a treaty jurisdiction on the assumption that the treaty was the point. On interest, by contrast, treaties frequently do reduce the rate below the domestic 15%, on production of a current-year tax residency certificate. The gap between the two structures widens further.

If you take nothing else from this article: in Nepal, treaty planning is largely wasted on equity and genuinely valuable on debt, royalty and capital gains.

Why, then, isn't everything funded by debt?

Because a foreign loan into Nepal is not a private arrangement between a parent and its subsidiary. It is a regulated instrument with four hard constraints.

1. Eligibility. Foreign borrowing is not open to every Nepali company. The framework has been built around industries that already have foreign investment, and access has historically been narrower than investors expect. NRB's bylaw does contemplate lending from parent companies, holding companies and group entities within its definition of a foreign lender, but eligibility, permitted purpose and documentation must be confirmed for the specific borrower before any term sheet is drafted. This is the first question to ask, not the last.

2. NRB approval at entry — and its consequences at exit. A foreign loan must be approved and recorded when it comes in. If it was not, the interest and principal cannot be repatriated later. The parallel with equity is exact: unrecorded money in cannot become approved money out, and no amount of good performance in between fixes it. This is the same failure I described in the first article of this series, wearing different clothes.

3. The interest rate is capped. NRB sets the benchmark against which foreign loan pricing is measured — SOFR for US dollar loans, SONIA for sterling, SARON for Swiss francs, TONA for yen — with a permitted spread above it. A rate agreed commercially between related parties that exceeds the permitted range will not clear for repatriation. The cap must be checked as at the date of the loan agreement, because it moves.

4. Debt does not flex, and it is a currency bet. Interest falls due whether or not the company had a profitable year; a dividend does not. And a hard-currency loan in a Nepali company is a short position on the Nepali rupee, which has depreciated against the US dollar over most long horizons. The rupee cost of servicing that loan rises even when the business performs exactly as modelled. Equity carries no such exposure — the dividend is whatever the rupee profit supports, converted at the rate of the day.

That last point is under-modelled in nearly every FDI projection I see. A structure that wins on tax by nineteen percent can lose more than that to currency over a seven-year tenor if the debt is priced and denominated carelessly.

The third channel: royalties and service fees

Between pure equity and pure debt sits the technology transfer route — royalty, technical service, management and franchise fees paid to the foreign parent. Deductible at company level, 15% at the border, often reducible by treaty.

It is legitimate and it is frequently the right answer, particularly where the foreign investor genuinely supplies brand, technology, systems or people. But three conditions decide whether it survives review:

• The agreement must be approved. Technology transfer arrangements go through the Department of Industry, and the fee cannot be invented after the fact.

• The rate must sit inside the permitted range. Royalty rates above the applicable cap are not cleared for repatriation, which converts a tax plan into trapped cash.

• The service must be real and documented. A management fee with no evidence of management is a transfer pricing adjustment waiting to happen. Nepal's Income Tax Act contains arm's length provisions and the Inland Revenue Department applies them to related-party charges.

The test I apply is simple: if the foreign parent charged this fee to an unrelated Nepali company, would that company pay it? If the answer is no, do not build the structure on it.

So what does a good structure actually look like?

There is no universal ratio. There is a set of questions, and the answers point to a mix:

• How certain is the cash flow? Contracted, predictable revenue can service debt. Early-stage or cyclical revenue cannot, and equity is the honest answer.

• What is the tenor and the exit? Debt has a repayment schedule that can be timed to the investor's horizon. Equity exits require a buyer, a valuation and an approval — as covered in the second article of this series.

• Is the investor in a treaty jurisdiction with substance there? If yes, the interest and royalty channels are worth more. If the entity is a shell, treaty benefits are exposed on beneficial ownership grounds and should not be assumed.

• What does the sector allow? Corporate rates differ by sector, ownership caps differ by sector, and royalty caps differ by sector. The answer for a manufacturing joint venture is not the answer for an IT services company.

• How much does the investor need to be able to withdraw in a bad year? Debt service is unconditional. That is either the discipline the project needs or the thing that kills it in year three.

• What happens on default? Security over Nepali assets, enforcement and the position of a foreign lender in an insolvency are materially harder questions than the interest rate. They belong in the term sheet.

Preference shares and other hybrid instruments come up regularly as a way to get debt-like returns with equity-like flexibility. They can work, but their characterisation for both tax and foreign investment purposes has to be confirmed in advance rather than assumed from the label on the instrument — an instrument that is equity for FITTA and debt for tax, or the reverse, creates problems in both directions.

The sequence that keeps the structure intact

1. Model the project's cash flows before deciding the funding mix, not after.

2. Confirm foreign borrowing eligibility for this borrower and this sector.

3. Fix the equity-to-debt split, and price the debt against the applicable NRB benchmark and permitted spread.

4. Confirm the treaty position and whether the investor can actually produce a tax residency certificate each year.

5. Approve and record every instrument at entry — equity approval, loan approval, technology transfer agreement — before a single rupee moves.

6. Document the related-party charges to a standard that survives a transfer pricing review.

7. Model the currency exposure across the full tenor, not at the spot rate.

8. Build the repatriation file from the first transaction, so that year one looks like year seven.

Frequently asked questions

Can a foreign parent lend to its Nepali subsidiary? NRB's framework recognises parent and group companies as foreign lenders, but eligibility depends on the borrower, the sector and the purpose, and the loan must be approved and recorded at entry. Confirm the specific position before drafting the loan agreement.

Is interest on a shareholder loan cheaper than a dividend in Nepal? Usually, yes. Interest is deductible at the company level and suffers withholding once at the border; a dividend is paid out of profit that has already borne corporate tax and then suffers a further withholding. The gap is significant, but it is only available if the loan is properly approved and priced.

Does a tax treaty reduce the tax on dividends from Nepal? Generally not. Nepal's domestic dividend withholding sits below the dividend rate in the treaties Nepal has signed, so the domestic rate applies. Treaties matter for interest, royalties, technical service fees and capital gains.

Is there a limit on how much I can lend rather than invest? The regulator assesses the funding structure when approving a foreign loan, and the interest rate is capped against a published benchmark. A structure that is overwhelmingly debt with minimal equity attracts scrutiny both at approval and, later, at the tax office. Confirm the current expectation before fixing the ratio.

Can I charge a management fee to my Nepali subsidiary instead? Yes, where the arrangement is approved, the rate sits within the permitted range, and the services are genuinely provided and documented. A fee that fails any of those three tests is likely to be disallowed for tax and blocked for repatriation.

The point

Nepal does not tax foreign investors harshly. It taxes different channels differently, publishes the rules, and then lets investors choose the expensive route by default.

The choice between equity, debt and fee income is not a tax trick. It is a financing decision with tax, currency, regulatory and exit consequences, and it is almost impossible to reverse once the money is in and the approvals are issued. It costs a few weeks of modelling before the term sheet, and it compounds for the life of the investment.

Nishant and Team advises foreign investors and Nepali promoters on FDI structuring and funding mix, Department of Industry and NRB approvals, technology transfer and royalty arrangements, business valuation for entry and exit, and repatriation planning.

If you are structuring an investment into Nepal, or you already hold one that was funded entirely by equity without the question being asked, canishantha(Whatsapp). Send us the shareholding structure, the funding history and the last audited accounts, and we will model what the current structure costs you against the alternatives.

This article reflects the position as verified on 17 August 2026. Withholding rates, benchmark spreads, royalty caps and eligibility conditions change frequently, often by gazette notification, NRB circular or annual Finance Act. It is general information, not advice on any specific transaction.