Nepal Rastra Bank’s latest real estate report puts hard numbers on a market most people only feel anecdotally. Behind the transaction data sits a bigger story for the economy: a banking system now carrying more than Rs 2 trillion in property-linked exposure, a government revenue line that has not recovered, and a growth engine quietly shifting out of the Kathmandu Valley.
Analysis · BiznesVue Markets Desk | Primary data: Nepal Rastra Bank, Economic Research Department; Department of Land Management and Archives (DOLMA); Inland Revenue Department
Key Takeaways
- Nepal’s banks hold over Rs 2 trillion (Rs 2,070 billion) in combined real estate and home-loan exposure as of FY 2081/82 — an all-time high for the banking system.
- Property transactions hit a record 1,14,769 in Q4 FY 2081/82, but year-on-year growth has cooled to single digits: the market is normalizing, not booming.
- Declared transaction value more than doubled in three years to Rs 127.61 billion in Q4, partly driven by rising government valuation rates rather than pure price appreciation.
- Madhesh leads in transaction volume while Bagmati holds the value — Kathmandu Valley land is worth seven to eight times Terai rates, making real estate risk highly geographic.
- Government revenue from property (about Rs 49 billion in FY 2081/82) remains far below its FY 2078/79 peak despite doubled declared values, signalling a collection or rate gap.
The Headline: A Rs 2 Trillion Bet on Property
Strip away the seasonal noise and the provincial detail, and the single most important figure in Nepal Rastra Bank’s new real estate report is this: the banking system’s combined exposure to real estate and residential home loans stood at roughly Rs 2,070 billion, or Rs 20 kharba 70 arba, at the close of FY 2081/82. That is not a housing-market statistic. It is a financial-stability statistic. A market correction of any real size would not stay contained inside the property sector; it would land on bank balance sheets, and from there on the wider economy.
That is why this report deserves reading by anyone whose decisions depend on the health of Nepal’s financial system, not just those buying or selling land. Nepal Rastra Bank’s Economic Research Department pulled data from 135 Land Revenue Offices across three full fiscal years, 2079/80 through 2081/82, focusing on “rajinama” transactions, the official category for voluntary property transfers and the closest available proxy for genuine buy-sell activity. What the data shows is a market that has stopped booming, stopped falling, and settled into something more consequential for policymakers: a large, systemically important asset class growing at a normalizing pace.
Transaction Volume: Records Set, Momentum Fading
Nepal’s land market runs on a predictable seasonal clock. The first quarter (Shrawan to Ashwin) is always the quietest, as the monsoon slows site visits and paperwork at Land Revenue Offices; activity then builds through the year and peaks in the fourth quarter. That rhythm held across all three years without exception, which matters for anyone tempted to read a single quarter as a trend.
The more telling number is the record. Q4 of FY 2081/82 logged 1,14,769 rajinama transactions, the highest single-quarter figure in the dataset, up 16.8 percent from the 98,270 recorded in the same quarter of FY 2079/80. On volume alone, the market has not merely recovered from its post-boom slump; it has pushed past prior highs.
But volume and momentum are telling different stories. Year-on-year growth peaked at 26.71 percent in Q1 of FY 2080/81, then decelerated steadily, dipping to minus 9.3 percent in Q1 of FY 2081/82 before recovering to a modest 8.15 percent by Q4. The signal for markets and policy is clear: the headline totals look strong, but the rate of expansion is cooling. This is growth on a glide path, not a takeoff.
The number that matters: Quarterly transactions averaged 87,354 over three years. Any quarter well below that (the Q1 troughs) reflects seasonality, not structural weakness — a distinction lenders and analysts pegging decisions to quarterly snapshots routinely get wrong.
Declared Value: The Money Doubled — But Why?
Transaction counts tell you how often property changed hands. Declared value, “thaili” in official terms, tells you how much money moved, and the two do not always move together. Total declared value under rajinama transactions climbed from Rs 58.66 billion in Q1 of FY 2079/80 to Rs 127.61 billion in Q4 of FY 2081/82, more than doubling in three years. The twelve-quarter average was Rs 86.74 billion.
The Q4 spike is where the analysis gets interesting. Declared value jumped 27.52 percent in a single quarter, from Rs 100.07 billion in Q3 to Rs 127.61 billion in Q4. A jump that steep has at least three plausible drivers, and they carry very different implications. It could be genuine demand. It could be year-end bunching, as parties rush to close before the fiscal year ends. Or it could be an artefact of rising minimum government valuation rates (nyunatam mulyankan dar), which push reported figures up even when underlying prices have not moved proportionally.
For the financial sector, distinguishing between these is not academic. If declared values are rising on valuation-rate adjustments rather than real price appreciation, a gap opens between the book value of collateral and its realizable value. That gap is precisely the kind of quiet mispricing that looks harmless until a downturn forces the question.
What Is Being Bought: A Market Anchored in the Middle
NRB splits transactions into four land-size bands, and the composition has barely shifted in three years, itself a useful finding for a market often assumed to be volatile. Mid-sized plots of 2.5 to 10 aana dominate at 43 to 45 percent of all transactions. Plots above 20 aana come second at 29 to 31 percent, the 10-to-20 aana band holds near 18 percent, and the smallest plots, under 2.5 aana, make up just 8 to 9 percent.
| Land Size Category | FY 2079/80 (Annual) | FY 2080/81 (Annual) | FY 2081/82 (Annual) | 3-Year Avg Share |
|---|---|---|---|---|
| < 2.5 aana | 26,986 | 27,624 | 28,915 | ~8% |
| 2.5 to 10 aana | 1,34,616 | 1,54,836 | 1,64,418 | ~43% |
| 10 to 20 aana | 56,344 | 67,219 | 66,294 | ~18% |
| > 20 aana | 79,351 | 1,14,002 | 1,07,647 | ~30% |
| Total | 2,97,297 | 3,63,681 | 3,67,274 | 100% |
The 2.5-to-10 aana band is the residential sweet spot, the size a typical household targets to build a home, especially in semi-urban and suburban belts. The large-plot share above 20 aana is inflated almost entirely by Madhesh Province, where agricultural and industrial parcels are common; exclude Madhesh and that share falls sharply. The concentration point for the financial sector is straightforward: a loan book weighted toward small urban plots under 2.5 aana is exposed to a segment that is less than a tenth of national volume, a thin market where finding a buyer at auction after default is harder and slower.
The Provincial Split: Madhesh Moves the Volume, Bagmati Holds the Money
The starkest pattern in the report is the disconnect between where deals happen and where the value sits. Madhesh Province led transaction volume across all three years, logging over 1,04,027 transactions in FY 2081/82, ahead of Lumbini at 78,248 and Koshi at 74,638. Bagmati, home to the Kathmandu Valley, recorded just 57,036.
On value, the map inverts. Bagmati alone declared Rs 51.37 billion in Q4 of FY 2081/82, more in a single quarter than most provinces generate in a year. The mechanism is no mystery: Kathmandu Valley land carries valuations that are multiples of Terai or hill-district rates, as any scan of current Kathmandu listings makes plain. Fewer deals, far more money per deal.
| Province | Q4 2081/82 Transactions | Q4 2081/82 Declared Value (Rs. Bn) | Q4 2081/82 Area (Mn sq.m.) | Implied Value per sq.m. (Rs.) |
|---|---|---|---|---|
| Koshi | 22,850 | 18.66 | 26.40 | ~707 |
| Madhesh | 34,001 | 20.63 | 31.33 | ~658 |
| Bagmati | 17,027 | 51.37 | 10.13 | ~5,072 |
| Gandaki | 5,430 | 6.24 | 4.47 | ~1,396 |
| Lumbini | 24,269 | 20.63 | 22.03 | ~936 |
| Karnali | 4,234 | 2.34 | 3.42 | ~684 |
| Sudurpashchim | 6,958 | 7.74 | 6.54 | ~1,183 |
Implied value per square metre in Bagmati runs roughly seven to eight times that of Madhesh or Karnali. The consequence for banks with national footprints is that a “real estate loan” is not one asset class but several. Collateral in Bagmati and collateral in Karnali differ fundamentally in liquidity, market depth, and price discovery, even at identical loan amounts. Geographic exposure, not just sector exposure, is the risk to model.
Thin-market risk: Karnali and Sudurpashchim together account for under 5 percent of national transaction volume and under 4 percent of declared value. Collateral-backed lending there faces genuine liquidity risk — on default, a sale at the appraised value is far from guaranteed.
Where Urban Demand Concentrates — and the Kathmandu Surprise
Nepal’s six metropolitan cities (Kathmandu, Lalitpur, Biratnagar, Birgunj, Bharatpur, and Pokhara) and eleven sub-metropolitan cities capture a small slice of national volume but an outsized slice of value. Metros accounted for about 4.1 percent of transactions but roughly 13 percent of declared value; that gap is the urban premium expressed in a single ratio. Sub-metros ran closer to the national average, at 9.85 percent of transactions and 9.28 percent of value.
Bharatpur, Not Kathmandu, Leads Metro Transactions
Among metros, Bharatpur recorded the most rajinama transactions, ahead of Birgunj and Pokhara. Kathmandu and Lalitpur, the priciest markets, saw fewer, exactly what you expect where plots are small, prices are high, and owners hold rather than trade. In expanding cities like Bharatpur and Birgunj, turnover is naturally higher. Metro declared value more than doubled, from Rs 7.45 billion in Q1 of FY 2079/80 to Rs 19.09 billion in Q4 of FY 2081/82, with Kathmandu leading on value despite trailing on count. In the sub-metro segment, Q4 of 2081/82 totalled Rs 12.07 billion, and its Y-o-Y recovery has outpaced the metros, hinting that the next leg of price appreciation may build in these mid-tier centres rather than the established capitals.
The Lending Line: The Sector’s Systemic Weight
For the financial sector this is the operative section. NRB tracks two buckets: “Real Estate Loan” (broad property credit) and “Residential Home Loan” (homes specifically, capped at Rs 20 million in this reporting, since revised to Rs 30 million).
Between FY 2077/78 and 2081/82, real estate loans grew 72.41 percent, from Rs 484.4 billion to Rs 835.2 billion. Residential home loans grew 61.55 percent, from Rs 764.8 billion to Rs 1,235.5 billion. The banking system’s exposure to property, in all forms, has expanded substantially and steadily.
| Fiscal Year / Quarter | Real Estate Loan (Rs. Bn) | Residential Home Loan (Rs. Bn) | Combined Exposure (Rs. Bn) |
|---|---|---|---|
| FY 2077/78 Q1 | 484.4 | 764.8 | 1,249.2 |
| FY 2078/79 Q4 | 672.1 | 1,015.1 | 1,687.2 |
| FY 2079/80 Q4 | 738.2 | 1,046.0 | 1,784.2 |
| FY 2080/81 Q4 | 777.3 | 1,161.8 | 1,939.1 |
| FY 2081/82 Q4 | 835.2 | 1,235.5 | 2,070.7 |
The growth curve, though, has flattened. Real estate loan Y-o-Y growth peaked at 24.5 percent in Q1 of FY 2078/79, during the post-COVID property boom, then moderated to just 2.9 percent by Q4 of FY 2081/82. Home loan growth has been steadier but is also easing, from above 23 percent to roughly 6.35 percent in the latest quarter. Credit is still expanding, but the appetite that defined the boom years has clearly cooled.
The Rs 2,070 billion combined figure is the one for bank boards and regulators to sit with. Home loans carry lower risk thanks to end-user occupancy, but the broader real estate loan bucket, which folds in speculative and developer lending, is where a property correction would bite first. Individual delinquency can stay manageable while system-level exposure quietly becomes the more important variable.
Worth watching: NRB notes no definite relationship between the growth of real estate lending and the growth of declared values. Rising declared values have not translated into proportionally higher lending — a sign of either tighter underwriting or a larger share of transactions being funded outside the formal banking system.
Government Revenue: The Barometer That Never Bounced Back
The state earns from property through registration fees and capital gains tax, and that revenue line doubles as a market-health barometer. It peaked at Rs 23.7 billion in Q1 of FY 2078/79, the height of the boom, then collapsed to Rs 8.2 billion by Q2 of FY 2080/81. FY 2081/82 averaged about Rs 12.3 billion per quarter, roughly Rs 49 billion for the year: a recovery from the trough, but nowhere near the peak.
Here is the fiscal puzzle. Declared values and revenue correlate only loosely (0.43). Revenue depends not just on transaction value but on applied rates and collection efficiency. So the fact that declared values have surged past FY 2079/80 levels while revenue has lagged points to one of two things: effective tax rates have slipped, or the mix of transactions generating revenue has structurally changed. Either way, a doubling of declared value that fails to lift the revenue line is a gap worth a hard look from fiscal authorities.
The Takeaways for Nepal’s Economy
Read together, the transaction, value, lending, and revenue data point to five conclusions worth carrying forward.
The market is normalizing, not crashing or booming. The explosive growth of FY 2078/79 is over. Volumes sit at or near record highs while growth runs in single digits. The crash narrative is not supported by the data; neither is a return to the boom.
Provincial disparity is wide and widening. Bagmati’s share of declared value keeps climbing relative to the rest. Real estate risk is not evenly spread across a national branch network; a single large default in Bagmati can dwarf dozens elsewhere.
Banking exposure is at an all-time high. Over Rs 2 trillion in combined property lending is not a figure to be complacent about, and the speculative end of it warrants continued regulatory attention.
The government is leaving revenue on the table. Declared values more than doubled while revenue barely recovered. A collection gap or a rate gap of that size is fiscally significant in a country with tight budgets.
Growth is migrating out of the capital. Bharatpur outpacing Kathmandu on metro transactions, sub-metros recovering faster than metros, and Madhesh leading on volume all point the same way. The next decade of Nepal’s property market may be defined less by what happens inside Kathmandu’s Ring Road and more by the emerging urban corridors of the Terai and the fast-growing towns only now formalizing their land markets.
Frequently Asked Questions
Is Nepal’s real estate market growing or slowing?
It is normalizing. Transaction volumes hit record highs in FY 2081/82, but year-on-year growth has cooled to single digits — the market has neither crashed nor returned to its boom-era pace.
How exposed are Nepal’s banks to real estate?
Combined real estate and residential home-loan exposure reached about Rs 2,070 billion (Rs 20 kharba 70 arba) by the end of FY 2081/82, an all-time high for the banking system.
Which province has the most property transactions in Nepal?
Madhesh Province records the highest transaction volume, but Bagmati holds the most declared value — Kathmandu Valley land is worth roughly seven to eight times Terai or hill-district rates.
Why hasn’t government revenue from real estate recovered?
Declared transaction values have more than doubled since FY 2079/80, yet revenue remains far below its FY 2078/79 peak, pointing to a collection gap or a tax-rate gap.
What is a rajinama transaction?
Rajinama is the official category for voluntary property transfers in Nepal, used as the closest proxy for genuine buy-sell activity in the land market.
Data source and disclaimer: All figures are drawn from “A Report on the Status of Real Estate Market in Nepal,” published by Nepal Rastra Bank‘s Economic Research Department, using data from the Department of Land Management and Archives (DOLMA), the Inland Revenue Department, and NRB Banking and Financial Statistics. For rajinama transaction data, DOLMA applies a minimum threshold of Rs 1,000 on declared value and excludes records without valid 17/18-digit registration numbers; roughly 10 percent of rajinama records were excluded during data cleaning. This article is analytical in nature and does not constitute financial or investment advice.
