The heaviest line item in the upfront cost of a hotel development is not the building but the land. With commercial land prices near stations still rising, the alternative to buying land that is actually used in practice is the fixed-term business land lease (jigyō-yō teiki shakuchiken) under Article 23 of Japan’s Act on Land and Building Leases. In this article we first establish station-front commercial land prices in 15 cities using the Ministry of Land, Infrastructure, Transport and Tourism (MLIT) official land price data, then model a 100-room limited-service hotel under two structures — “buy the land + own the building” and “fixed-term business land lease + own the building” — and back-solve the required ADR for each. We also examine how the cost of returning the site as vacant land at lease expiry and the use of the floor-area ratio affect feasibility.

Conclusion — The Higher the Station-Front Land Price, the More a Land Lease Lowers Required ADR

The bottom line first. Under mid-range assumptions of a 5.0% target NOI yield and a 4.5% ground-rent yield, the fixed-term business land lease lowered the required ADR in all 15 cities. The size of the effect, however, tracks the land price level closely. At Osaka Station (Kita Ward, Osaka City), the required ADR is ¥70,892 under ownership versus ¥56,359 under leasehold — ¥14,533 lower — while at Takamatsu Station the gap is just ¥185, effectively no difference.

This gap is structural, not coincidental. Under ownership, the land carries a burden equivalent to “5.0% target yield + 1.2% fixed asset and city planning tax on land = 6.2%”, whereas under leasehold the land burden is simply the ground-rent yield. As long as the ground-rent yield stays below roughly 5.3–5.9% (the break-even after factoring in the opportunity cost of the security deposit and the demolition reserve), leasehold wins, and the gap grows in proportion to the land price. Most of the 3–6% ground-rent range cited in practice sits below this break-even, so the relative advantage of leasehold shows up most clearly in cities with expensive station-front land.

The Legal Framework — Fixed-Term Business Land Leases Come as a Package: No Renewal, No Buyout Right, Return as Vacant Land

Let us first lay out the legal framework. The Act on Land and Building Leases provides three types of fixed-term land lease, and the one used for hotel development is the fixed-term business land lease under Article 23. It cannot be used to own residential buildings, but in exchange it offers more flexibility in setting the lease term. Hotels and ryokan are treated as business buildings for lodging, so they can use this type.

There are three practical points. First, an Article 23 contract has no effect unless executed by notarial deed (paragraph 3 of the same article). Second, the long-term type (paragraph 1) must explicitly exclude renewal and the building buyout right by special agreement, whereas under the short-term type (paragraph 2) these simply do not apply by law. Third, because the lessee is obliged, as a rule, to remove the building and return the land vacant at expiry, demolition costs must be reserved during the project term. This third point works to narrow the gap with the ownership case in the feasibility calculation below.

If demolition is to be avoided, the Article 24 land lease with a building transfer agreement is an option, but it requires the landowner to pay for the building at expiry. That makes it hard to close when the landowner is a municipality, or when the landowner has no intention of acquiring a lodging facility more than 30 years old. This practical constraint is why Article 23 tends to be chosen for hotel development.

Three Typical Situations Where Hotel Developers Choose a Fixed-Term Business Land Lease

One example of a public land tender is the Iiyama Station Front City-Owned Land Lodging Facility Development Promotion Project, run by Iiyama City in Nagano Prefecture on city-owned land in front of Iiyama Station on the Hokuriku Shinkansen. The city published the proposal tender guidelines in April 2021 and adopted a framework in which it would sign a basic agreement and a fixed-term business land lease contract with the top-ranked bidder. According to the city, the lease contract was to be concluded by March 31, 2022, and the conditions were that the lodging facility open within three years of the contract and continue operating for at least 20 years. Fixing the opening deadline and minimum operating period at the tender-guideline stage is a design principle common to public land tenders. For the route by which public land comes to market for sale rather than lease, we have separately counted the individual lots offered in Japan’s state-owned land sales, finding 128 lots, of which only 4 could fit a 100-room hotel.

Station-Front Commercial Land Prices — Medians Across 15 Cities Span More Than 40x, from ¥200,000 to ¥8.58M per m²

As the basis for back-solving ground rent, we first establish land price levels. From the official land price data in MLIT’s Real Estate Information Library, we extracted standard sites classified as “commercial” within roughly 700m of each city’s main station, and compiled their median, maximum, year-on-year change and floor-area ratio. The reference date is January 1, 2026.

By median, the highest is Osaka Station at ¥8,580,000/m² and the lowest is Takamatsu Station at ¥203,000/m², a gap of more than 40x. Year on year, Kyoto Station (+15.2%), Yokohama Kannai (+15.9%) and Sendai Station (+12.0%) posted double-digit gains, while Takamatsu (+2.0%) and Kagoshima (+1.6%) were steady — each city is clearly in a different phase of the land price upcycle. A decision to buy land is also a decision about how to price in this rate of appreciation.

Model Assumptions — A 100-Room Limited-Service Hotel, 3,500m² Floor Area, 1,500m² Site

We build a model to compare the two cases. All assumptions are disclosed in the table below. The unit building cost of ¥590,571/m² is derived from our own aggregation of hotel construction plan data (construction cost of ¥20.67M per room, 35.0m² gross floor area per room). Converted per tsubo, this is about ¥1.95M, nearly identical to the 2025 all-structure average (¥1.952M/tsubo) based on MLIT’s Building Construction Starts statistics. The ¥20.67M per room and 35.0m² per room figures come from 31 hotel construction plans in our database.

The structure of upfront investment and annual costs in the two cases is shown below, using Sapporo Station (Chuo Ward, Sapporo; median official price ¥1,180,000/m²) as the representative example.

Note the two depreciation rows. The statutory useful life for tax purposes is 39 years, but with a 30-year lease term the building still has an undepreciated balance of ¥477,000,000 at expiry. This is written off as a loss on disposal together with demolition, so the mismatch between the project term and the useful life, while invisible in cash flow, cannot be overlooked in the investment decision. Taking a longer lease term of 40 or 50 years largely resolves the issue.

Back-Solving Required ADR — Leasehold Helps Most Where Land Is Expensive; the Gap to Market Rates Varies Widely by City

With the assumptions set, we back-solve by city the ADR each case needs to meet the 5.0% target NOI yield. For comparison, we show the estimated contracted ADR of business hotels in each city’s central ward (average of the last 12 months, September 2025–August 2026).

The difference (A−B) grows almost in proportion to the land price. It is ¥14,533 at Osaka Station, ¥6,791 at Nagoya Station and ¥5,670 at Hakata Station, versus only ¥518 at Kanazawa Station and ¥185 at Takamatsu Station. This reflects a simple relationship — the land-burden gap = land price × (target yield + land tax rate − ground-rent yield) — so the relative advantage of leasehold is structurally larger in cities with higher land prices.

On the other hand, the ratio of required ADR to market ADR differs widely by city. In Sapporo (A/Market 1.28x, B/Market 1.14x), Takamatsu (1.33x / 1.31x) and Kanazawa (1.36x / 1.30x), a new build is within reach if it can command a rate 20–30% above the existing stock average. By contrast, Osaka Station and Nagoya Station show multiples of 3x to 6x; at the median price of commercial land right by the station, a low-density standalone hotel of 100 rooms and 3,500m² cannot reach the assumed yield.

Two caveats apply to this reading. First, market ADR here is the median estimated contracted ADR of the entire existing stock, and newly opened properties generally launch at rates above the market average. The multiple should therefore be read as a question: how much new-build premium can be priced in? Second, the standard sites used are commercial land within 700m of the main station, whereas actual hotel sites are often found on the outer edge of this zone. In cities where the multiple exceeds 3x, the choice is between looking for a site one step back from the prime station-front location, switching to a design that uses the full floor-area ratio as discussed in the next section, or leasing the land rather than buying it.

End-of-Lease Costs and Sensitivity Analysis — Demolition Matters Little; Ground-Rent Yield Is What Counts

Next we factor in the cost specific to the leasehold case: returning the site as vacant land. For 3,500m² of floor area at a demolition unit cost of ¥12,000–25,000/m², total demolition cost is ¥42M–¥87.5M; spread over the 30-year lease term, the annual reserve is ¥1.4M–¥2.92M. Per occupied room-night, that is ¥51–¥106.

Even doubling the demolition unit cost moves the required ADR by only about ¥120. It is a fixed cost spread over 30 years and divided by a large denominator of rooms × occupancy × 365. What decides the economics of the leasehold case is not demolition cost but the ground-rent yield.

The calculations so far hold occupancy fixed at a single national average. The table below shows, on two axes combined with the ground-rent yield, how required ADR moves when demand shifts. The base row matches the figures in the text.

Lowering the assumed occupancy by 5.3pt, from 75.3% to 70.0%, raises the required ADR in the leasehold case (4.5% rent) from ¥15,757 to ¥16,951, an increase of ¥1,194. That is about 55% of the impact of moving the ground-rent yield by 1.5pt (¥2,158). Because a demand-side downside carries weight comparable to 1.5pt of ground-rent negotiation, the rent level and the assumed occupancy must be tested together. The model does not reflect occupancy differences between cities, so for an actual project the figures should be adjusted to the demand level of the target city.

The 100% line in the chart marks the break-even. At Osaka Station, with a 3.0% ground-rent yield the leasehold required ADR falls to 57.4% of the ownership case, reaching parity around 5.9%. At Takamatsu Station the gap is small even at 3.0% (94.7%), and the order reverses around 5.3%. The break-evens for all 15 cities fall between 5.25% and 5.89%, so regardless of land price the practical decision line is whether the ground-rent yield exceeds the mid-5% range. Ground rent is negotiable, and whether a developer can approach the landowner with this benchmark in mind determines whether a leasehold scheme succeeds. Our analysis Leasehold Hotel Development: 1.38x Land Price Reach Across 9 Markets, which puts ground rent itself at the center and back-solves the maximum rent a project can bear in each market, is also useful for narrowing the negotiating range.

Zoning and Floor-Area Ratio — The Same Site Can Hold Anywhere from 90 to 300 Rooms

Finally, we add site utilization as a variable. All the standard sites in the 15 cities are zoned as commercial districts, but the median designated floor-area ratio ranges from 300% at Kyoto Station and Naha to 700% at Nagoya Station. Some sites reach 1,300% at Nagoya Station, 800% at Osaka Station and 900% at Hiroshima Station. Even on the same 1,500m² site, the buildable floor area and room count when the median FAR is fully used change as follows.

At Nagoya Station with a 700% FAR, 300 rooms fit in theory; at 600% (Osaka Station, Sapporo Station, Kobe Sannomiya and Kannai), 257 rooms. Because land cost is divided by the number of rooms, higher density lowers the land burden per room. In the Osaka Station leasehold case, the required ADR of ¥56,359 for the 100-room model falls to ¥27,606 in the 257-room model that uses the full FAR. In other words, the main way to make the numbers work at expensive station-front sites is not the leasehold scheme alone but the combination of leasehold and high-density design.

Conversely, at Kyoto Station and Naha, where the FAR is 300%, the ceiling is 128 rooms, leaving little room to adjust through density. Kyoto Station shows the highest land price growth of the 15 cities at +15.2% year on year, yet with its FAR held down, there is limited scope to dilute land acquisition cost across more rooms. In such areas, the case for leasing rather than owning the land carries relatively more weight.

Note that the FARs used here are the values designated for the standard sites; actual plans are also subject to limits based on front-road width, shadow regulations, and bonuses under the comprehensive design system or high-level use districts. Site-specific verification is essential at the master planning stage.

Summary — “Buy or Lease” Can Be Decided at a Ground-Rent Yield in the Mid-5% Range

A fixed-term business land lease removes land acquisition from the upfront investment in exchange for annual ground rent and the obligation to return the land vacant at expiry. What our calculations show is that whether this trade-off pays off depends not on the land price level itself but on how far the ground-rent yield falls below the sum of the target yield and the cost of holding land. And the more expensive the station-front land, the larger the monetary benefit from the same yield gap.

That said, as the ratio of required to market ADR shows, only a limited number of cities can support a standalone business hotel at the median price of commercial land right by the station. In markets such as Sapporo, Kanazawa and Takamatsu, where the multiple is 1.1–1.4x, a new-build premium is enough to close the gap; at the station fronts of Osaka, Nagoya and Hakata, one of several design decisions is a prerequisite — moving one step off the prime location, reworking the plan as a high-density scheme that uses the full FAR, or positioning lodging within a mixed-use development. “Buy or lease” is a question that should be considered together with those design decisions.

Related Articles