Conclusion — Sapporo and Fukuoka clear at today’s rates; in central Tokyo, site selection decides the outcome

The conclusion first. Assuming a newly developed 100-room limited-service hotel and standard terms of a 3.00% borrowing rate, 60% LTV and 1.3x DSCR, the required NOI yield is 4.48%. Setting the “break-even ADR” implied by land and construction cost at that yield against each submarket’s prevailing estimated settled ADR, four submarkets are already in feasible territory at today’s rates: Sapporo Odori–Susukino (prevailing rate 45% above the break-even line), Fukuoka Tenjin (30%), Fukuoka Hakata Station (26%) and Osaka Shinsaibashi–Hommachi (18%).

Nagoya Meieki, Nagoya Sakae and Osaka Umeda, by contrast, sit almost exactly on the line — a few more percent of rate growth and the arithmetic works. Central Tokyo (the Marunouchi, Ginza and Nihonbashi area) comes out at a break-even ADR of roughly ¥23,600 if land is acquired at the submarket median price, against a prevailing rate of about ¥17,300 — a required upside of 36%. That, however, assumes the submarket’s standard land price. Within the same submarket, a site at the bottom-quartile level (about ¥4.61 million per square meter) brings the break-even ADR down to roughly ¥15,800, putting the prevailing rate 10% above the line. In central Tokyo, “choosing the site within the submarket” matters more to feasibility than “choosing the submarket.”

Financing assumptions — across three reference-rate cases, the required NOI yield runs from 4.11% to 4.99%

Start by quantifying the financing-side constraint. Per the Bank of Japan, the short-term prime rate at major banks (modal value) has been 2.125% since February 9, 2026, and the long-term prime rate is 3.15% as of the June 10, 2026 revision. Non-recourse loans for hotel development and acquisition add a spread over these reference rates, sized to sponsor credit and asset risk. Three cases are used here.

The required NOI yield is given by DSCR × loan constant × LTV. At 1.3x DSCR and 60% LTV, that is 4.11% in Case A, 4.48% in Case B and 4.99% in Case C. Even when the rate moves 1.75 points, from 2.25% to 4.00%, the required NOI yield widens by only 0.88 points. Raising LTV from 60% to 70% lifts the required NOI yield by roughly 0.7 points at the same DSCR. The obvious structure — the more leverage applied, the more earning power the asset must deliver — shows up cleanly in the numbers.

Construction cost — ¥1.95m per tsubo in 2024, and ¥2.38m if the build lands in 2028

Next, the project-cost side. Taking accommodation-use buildings from e-Stat’s “Statistics on Building Construction Starts” and dividing planned construction cost by floor area gives a 2024 national average of ¥1.950 million per tsubo. That is up 40% in two years from ¥1.390 million in 2022. Construction starts numbered 2,836 buildings in 2024 across roughly 1.76 million square meters of floor area, a sharp recovery from 2023 (2,222 buildings, roughly 900,000 square meters).

Because development projects take several years from groundbreaking to opening, using the actual figure as-is would miss cost escalation during the build. Here we assume a 2029 opening (construction 2027–2029, cost midpoint in 2028) and compound in construction-cost inflation of +5.6% in 2025, +5.3% in 2026, +5.0% in 2027 and +4.5% in 2028. That is +22.0% versus 2024, giving ¥2.379 million per tsubo.

Rising construction cost is a burden on the business plan and, at the same time, a brake on new supply. Seen from the existing stock, a higher cost of entry translates into tighter medium-term supply and demand — an upside factor. That supply-side spillover is mapped against scheduled openings in Construction +41% & Labor Crunch Erase Japan’s 2027-2029 Hotel Supply.

Two business models — break-even ADR formulas for a 100-room limited-service and a 200-room full-service hotel

Once project cost and required NOI yield are fixed, the break-even ADR follows uniquely. Two models with different formats are used here. The occupancy assumptions reference recent operating results published by listed hotel REITs: Invincible Investment Corporation’s 101 domestic hotels ran 82.7% occupancy in June 2026; Japan Hotel REIT Investment Corporation’s 29 hotels under variable-rent and similar structures ran 81.2%; Ichigo Hotel REIT Investment Corporation’s total portfolio ran 80.6%. On that basis, 82% is applied to the limited-service model and 80% to the full-service model.

In Model A, at a project cost of ¥30 million per room, the required ADR swings from ¥9,700 to ¥15,900 depending on financing terms — ¥9,700 in rate Case A at 60% LTV and 1.3x DSCR, ¥15,900 in rate Case C at 70% LTV and 1.5x DSCR. In other words, for the identical asset, how the financing is structured alone moves the break-even ADR by a factor of 1.6.

Operating variables matter too. At the same ¥30 million per room and rate Case B, setting full-year occupancy at 75% puts the required ADR at ¥11,600; at 85% it is ¥10,200 and at 88% it is ¥9,900. At a 36% GOP margin it is ¥12,500; at 48% it is ¥9,200. Adding 3 points of occupancy and improving the GOP margin by 3 points therefore lower the break-even ADR by ¥300 and ¥900 respectively. Revenue management and operating efficiency are powerful levers for lifting feasibility in a rising-rate environment.

Submarket analysis — matching published land prices against estimated settled ADR

This is the heart of the exercise. Commercial land prices for eight submarkets nationwide were taken from the 2026 Official Land Price Publication via the MLIT Real Estate Information Library. The catchment is roughly a 1.5 km radius from each submarket center, covering commercial-use standard sites. Alongside them sit the estimated settled ADR for each ward compiled by MetroEngines Research (12-month average, July 2025 to June 2026). The mirror image of this exercise — solving from the land-price side for the maximum price a given ADR can support — produces the same feasibility boundary approached from the opposite direction.

Sapporo has the widest cushion for a straightforward reason: at ¥1.105 million per square meter, its median commercial land price is the lowest of the eight, while its estimated settled ADR of ¥13,900 is on par with Fukuoka Hakata Station (¥14,000). Land costs one-eighth of central Tokyo’s, yet rates hold at 80% of the level. Fukuoka’s two submarkets share the same structure — high rate levels relative to land cost.

Osaka’s two submarkets sit in opposite positions. Shinsaibashi–Hommachi runs 18% above the break-even line, while Umeda falls 5% short. Umeda’s median land price of ¥2.390 million per square meter exceeds Shinsaibashi–Hommachi’s ¥2.030 million, yet the estimated settled ADR for limited-service inverts the ranking: ¥11,300 in Umeda against ¥13,300 in Shinsaibashi–Hommachi. Umeda is a submarket where land cost has run ahead, and the phase in which rates catch up is the upside. That land prices in both submarkets are growing around +16% year on year — the highest in the country — is itself the market pricing in future rate growth.

The upside range — site selection moves the break-even ADR by 30%

The submarket median is the price of “standard land in that submarket,” and actual hotel sites are often cheaper parcels. So the break-even ADR was also calculated for each submarket assuming land acquired at the bottom-quartile price level.

On this assumption, all eight submarkets fall into feasible territory. Central Tokyo in particular sees its break-even ADR drop ¥7,800, from ¥23,600 to ¥15,800, moving the prevailing ¥17,300 into a 10% cushion. Even within the same Chiyoda and Chuo Cities, land prices in prime Marunouchi and Ginza run close to double those in secondary locations such as Kanda, Hatchobori and Ningyocho. Site selection for a limited-service hotel is precisely the work of capturing that gap.

The full-service threshold — the 200-room model nearly doubles project cost

The 200-room full-service model assumes 55 m² of gross floor area per room and ¥39.58 million of construction cost, putting project cost at nearly twice the limited-service level. Because its NOI coefficient of ADR × 105.4 is lower than the limited-service model’s, the break-even ADR rises by more than the increase in project cost alone.

A full-service hotel in central Tokyo comes to more than ¥120 million per room at the submarket median land price, for a break-even ADR of ¥52,100 — a required upside of 47% against the prevailing ¥35,400. That is a conservative assumption, applying a 600% floor-area ratio uniformly; actual large-scale central-Tokyo projects benefit from sites with ratios above 1,000% and from land-cost allocation across mixed-use development with offices and retail. Property-level acquisition records at listed hotel REITs show a median acquisition price of ¥63.27 million per room for holdings in Minato City, Tokyo (7 properties) — actual transactions, including mixed-use development and acquisitions of existing buildings, land below this calculation.

Supply pressure — the confirmed pipeline thins after 2027, but the count will grow

Future supply cannot be ignored in any discussion of break-even ADR. Aggregating building-plan data from MLIT’s Building Starts Statistics Survey within the scope MetroEngines Research tracks, hotel plans on a building-confirmation basis number 64 projects and roughly 14,500 rooms. By year: 22 projects and 5,304 rooms in 2025; 21 and 4,967 in 2026; 13 and 2,922 in 2027; 7 and 1,228 in 2028; and 1 project with 100 rooms in 2029. Limiting to projects of 100 rooms or more gives 57 projects and roughly 14,100 rooms — a mix dominated by large developments.

One caveat matters here: because building confirmation is typically filed one to two years before opening, counts for 2028 and beyond are structurally undercounted. These figures are a floor for the confirmed pipeline as of now, and they will grow as further applications are filed. Reading the thin post-2027 years as “supply stops” is not appropriate.

By region, Tokyo leads with 17 projects, followed by Kyoto, Chiba, Hokkaido and Kanagawa at 5 each and Osaka at 4. Plans are moving even in submarkets with ample feasibility cushion such as Sapporo and Fukuoka, which puts the question of how much first-mover advantage can be secured squarely on the table.

Where estimated settled ADR stands — seasonal patterns differ by submarket

Finally, a look at the prevailing rates that form the denominator of the feasibility test. Estimated settled ADR for limited-service hotels in the five major submarkets is plotted monthly from July 2025 to June 2026. Central Tokyo (Chiyoda City) peaked at ¥20,400 in April 2026 and stood at ¥12,900 in June 2026. Fukuoka Hakata Station held a stable ¥12,500–15,700 range across the year, the narrowest swing of the group. Sapporo’s Chuo Ward shows the strongest seasonality: a winter-event peak of ¥20,300 in February 2026 against a drop to ¥8,300 in April.

Break-even ADR is judged on a full-year average, but the more volatile a submarket’s rates, the more room revenue management has to lift the annual figure. In a submarket like Sapporo, with a 2.4x monthly swing, measures that raise low-season rates carry a correspondingly large impact on annual ADR.

Summary — a 1.00% policy rate does not mean “infeasible”; it means the conditions are now explicit

Many Bank of Japan board members are reported to share the view that a 1.00% policy rate still represents accommodative financial conditions. The calculations here bear that out: under standard terms of a 3.00% borrowing rate, 60% LTV and 1.3x DSCR, the required NOI yield is only 4.48%, and four of the eight submarkets clear at today’s rates even on submarket-standard land prices. Assuming a bottom-quartile site, all eight fall into feasible territory.

What matters more is that with the interest rate now an explicit cost in the formula, feasibility can be judged precisely. A 1.75-point move in the rate widens the required NOI yield by 0.88 points. Against that, a 3-point gain in occupancy cuts the break-even ADR by ¥300, and a 3-point gain in GOP margin cuts it by ¥700–900. Halving the land price cuts the break-even ADR by 30%. The structure in which site selection and operating quality outweigh interest rates has not changed.

Construction cost rose 40% from 2022 to 2024, and pricing in inflation during the build period puts ¥2.38 million per tsubo within sight for 2028. That lifts project cost, but it also raises the bar for new entry and increases the scarcity value of existing stock. Together with a confirmed pipeline that looks thin from 2027 onward, the medium-term picture is one in which tighter supply and demand support rate upside.