The FDI Illusion: Why Nepal’s Rising Approvals Hide a Falling Investment Story

Nepal Rastra Bank’s latest Foreign Direct Investment survey, released in August 2026 for fiscal year 2024/25, will be reported in most places with a reassuring line: FDI stock crossed Rs. 340 billion, gross inflows reached Rs. 12 billion, and Nepal now hosts investment from 60 economies.

Read past the summary and a far less comfortable picture emerges. On almost every measure that matters to an investor or a policymaker, Nepal moved backwards in a year when the rest of South Asia moved sharply forward. This article sets out five uncomfortable findings, and one problem with the report itself.

One: Nepal went backwards while South Asia boomed

The report reproduces UNCTAD data that deserves far more attention than it has received. Foreign direct investment into South Asia rose 35.4 percent in 2025, from USD 34.1 billion to USD 46.1 billion. India rose 43.6 percent to USD 38.9 billion. Bangladesh rose 44.6 percent. Sri Lanka rose 37.5 percent.

Nepal fell 22.8 percent.

In UNCTAD’s own table, Nepal’s inflow rounds to USD 0.0 billion. Pakistan, an economy in visible macroeconomic distress, still attracted USD 1.9 billion. The Maldives, with a population smaller than Lalitpur district, attracted USD 0.9 billion. Nepal did not just underperform; it contracted during a regional boom. That combination removes the usual defense that global conditions were unfavorable. They were favorable. Nepal did not capture them.

Two: The FDI stock shrank in real terms

FDI stock grew 2.1 percent, from Rs. 332.97 billion to Rs. 339.85 billion, an increase of Rs. 6.9 billion. With consumer inflation running above that rate through the review year, the real value of Nepal’s foreign investment stock declined.

The composition of that growth is more revealing than the total. Paid-up capital rose 9.2 percent, and reserves rose 2.6 percent, but intercompany loans collapsed 28.1 percent, from Rs. 41.9 billion to Rs. 30.1 billion. Roughly Rs. 11.8 billion of parent company debt left the system in a single year and was only partly replaced.

The official framing is that this represents a healthy shift towards equity financing. That reading is defensible, but it is not the only one. A 28 percent contraction in parent company lending in twelve months is more often a sign that parents are pulling liquidity back to headquarters than a sign of deliberate capital restructuring. When a subsidiary is performing well, and the host market is attractive, parent debt usually grows alongside equity, not in place of it.

Three: The two sectors that matter most both contracted

Figures in Rs. million. Source: NRB FDI Survey 2024/25, Table 8.

Hydropower and manufacturing together hold 55.6 percent of Nepal’s FDI stock, and both declined. The single largest contributor to growth was financial and insurance services, which rose by Rs. 8.0 billion.

That distinction matters more than it appears. Growth in banking and insurance FDI is largely the arithmetic of retained profits accumulating in institutions that entered Nepal years ago. It is not new capital arriving to build capacity. Hydropower and manufacturing, by contrast, are where FDI creates plant, employment, exports, and technology transfer. Nepal’s FDI stock grew in the sector that recycles domestic savings and shrank in the sectors that build productive assets.

Agriculture, employing the largest share of Nepal’s workforce, holds Rs. 453 million of FDI stock. That is 0.1 percent of the total, and it is roughly the cost of a single mid-sized hotel.

Four: Chinese investment in Nepal is, in aggregate, loss-making

This is the finding most likely to be overlooked, and it is unambiguous in the data.

Source: NRB FDI Survey 2024/25, Tables 9, 10 and 11.

China has committed Rs. 30.6 billion of paid-up capital into Nepal and carries accumulated losses of Rs. 9.99 billion against it. Roughly one third of Chinese equity in Nepal has been eroded by operating losses. India, by contrast, shows reserves of Rs. 49.9 billion against paid-up capital of Rs. 57.8 billion, meaning Indian invested enterprises have very nearly doubled their original capital through retained profit.

The contrast holds within the numbers rather than across them. Both countries invest heavily in the same two sectors. Indian enterprises are profitable in them. Chinese enterprises are not. Any serious conversation about attracting further Chinese investment has to begin with why the existing stock has performed this way, whether the issue lies in project selection, cost structures, transfer pricing arrangements or the regulatory experience on the ground. The survey does not ask that question. It should.

A related figure deserves scrutiny. Transport and storage show a negative FDI stock of Rs. 7.0 billion, meaning accumulated losses in that sector exceed all capital ever brought in. A negative line item of that size in a national statistical publication needs an explanatory note, and there is none.

Five: Eleven paisa of every rupee approved actually arrived

In 2024/25, approved FDI was Rs. 64.97 billion. Net realized inflow was Rs. 7.31 billion. The realization rate was 11.25 percent, the second lowest in three decades of data, marginally above 2023/24.

Over the full period from 1995/96 to 2024/25, Rs. 577.0 billion was approved, and Rs. 170.9 billion arrived, a realization rate of 29.6 percent. But the recent collapse is the story: 60.6 percent in 2020/21, 34.3 percent in 2021/22, 20.1 percent in 2022/23, 12.0 percent in 2023/24, and 11.25 percent in 2024/25.

The report attributes this to gestation lags and staged capital deployment. Those factors are real, but they cannot explain a fivefold deterioration in four years, because gestation lags do not suddenly lengthen. What has changed is the willingness of approved investors to actually remit. The survey itself lists the reasons: infrastructure bottlenecks, regulatory and procedural delays, and land acquisition difficulty.

There is a further consequence that is rarely stated. So long as approvals are announced as achievements while realization is buried in an appendix, the incentive across the system is to generate approvals rather than to close projects. Nepal has become efficient at issuing letters and inefficient at receiving money.

Six: The money going out now exceeds the money coming in

Foreign invested enterprises repatriated dividends of Rs. 33.9 billion in 2024/25, against Rs. 12.6 billion the previous year. Manufacturing alone repatriated Rs. 29.0 billion. Gross FDI inflow in the same year was Rs. 12.0 billion, and net inflow Rs. 7.3 billion. Disinvestment took a further Rs. 4.7 billion, equal to 39.2 percent of gross inflows.

Dividend repatriation is a current account item and FDI inflow is a financial account item, so these are not directly offsetting entries in the balance of payments. But for a country managing foreign exchange, the direction of travel is what counts. On the FDI account taken as a whole, Nepal was a net exporter of foreign currency in 2024/25 by a wide margin.

This should not be read as an argument against repatriation. Unrestricted repatriation is one of the few genuinely strong features of Nepal’s investment regime, and restricting it would be self-defeating. The point is different: Nepal’s FDI stock has matured into a dividend-paying base while the replacement pipeline has failed, and that is a structural condition that will persist unless realisation improves.

One further exposure sits alongside this. Outstanding third-party foreign loans of FDI enterprises rose 26.6 percent to Rs. 75.8 billion, with hydropower accounting for Rs. 52.8 billion of it. These are foreign currency obligations serviced from rupee-denominated power purchase agreements. In an environment of exchange rate movement, that mismatch is a genuine sectoral risk, and it is growing.

A word on the report itself

A statistical publication of this importance should be internally consistent, and this one is not. Paragraph 6.2 of the summary states that gross FDI inflows increased 6.2 percent to Rs. 8.5 billion. The executive summary, Chapter 4, and Appendix I all state Rs. 12.0 billion. Paragraph 6.5 puts outstanding foreign borrowing at Rs. 56.7 billion, while Table 13 puts it at Rs. 75.8 billion. Appendix II shows total paid-up capital for mid July 2024 as Rs. 68,755.7 million where Table 6 shows Rs. 168,755.7 million. The table of contents dates the international investment position appendix to mid July 2025, while the appendix itself is dated mid July 2026.

These are not trivial. Rs. 8.5 billion and Rs. 12.0 billion are different stories, and a reader quoting the summary would be wrong. Analysts, rating agencies and prospective investors read summaries.

The methodology also deserves scrutiny. Of 576 enterprises sampled, 262 responded, a response rate of 45.5 percent. Among small enterprises, which are 809 of the 931 approved firms, only 170 responded, a rate of 21.0 percent. The remainder is estimated by applying sector and size ratios derived from respondents. That is a legitimate technique, but it means a substantial part of the published stock figure is extrapolation rather than observation, and the confidence around it is correspondingly wider than a figure quoted to one decimal place suggests.

One further data point should be treated with caution. Capacity utilisation is reported to have risen from 47.5 percent to 59.4 percent in a single year. A twelve-percentage-point improvement in manufacturing utilisation, in a year when sales fell 0.7 percent, and the manufacturing FDI stock contracted 3.1 percent, is difficult to reconcile. It more likely reflects a change in the respondent mix than a genuine improvement across the sector.

What would actually change the numbers?

1. Publish realization, not approvals. Nepal Rastra Bank, the Department of Industry, and the Investment Board should jointly publish a quarterly approval-to-realization dashboard by sector and province. What is measured publicly is what gets managed.
2. Introduce a validity and lapse regime for approvals. Approvals that generate no capital inflow within a defined period should lapse, so that the approval pipeline reflects live intent rather than accumulated paper.
3. Fix the three named constraints with service standards. Land acquisition, procedural delay, and infrastructure are identified in the survey itself. Each needs a statutory timeline with consequences for breach, not another policy statement.
4. Treat aftercare as the primary acquisition channel. Reinvested earnings are 36.9 percent of FDI stock. Existing investors are already Nepal’s largest source of new capital and are far cheaper to retain than new ones are to attract.
5. Confront the underperformance of specific source countries. A review of why Chinese-invested enterprises carry Rs. 10 billion of accumulated losses would be more valuable than another investment summit.
6. Address the hydropower currency mismatch. Rs. 52.8 billion of foreign currency debt against rupee revenue needs a hedging framework, whether through indexed tariffs or a market-based instrument.

The bottom lines

Nepal is not failing to attract investor interest. Nine hundred and thirty-one approved projects and 60 source economies show otherwise. Nepal is failing to convert interest into capital, and failing to make the capital that does arrive productive enough to reinvest.

Rs. 65 billion approved. Rs. 7.3 billion received. Rs. 33.9 billion repatriated. Those three numbers, taken together, are the honest summary of Nepal’s foreign investment year, and no amount of reporting on the Rs. 340 billion stock figure changes them.

The next survey will tell us whether 2024/25 was a trough or a trend. On current evidence, the burden of proof sits with those who believe it was a trough.

Article By: CA Tej Prakash Dixit

The author is a Chartered Accountant. Views expressed are personal and do not constitute professional advice on any specific transaction. All figures are drawn from Nepal Rastra Bank, Economic Research Department, Foreign Direct Investment (FDI) in Nepal (2024/25): A Survey Report, August 2026, and from UNCTAD data reproduced therein.