The Back-Calculation — Hotels from ADR, Condos from Sale Price, Both per 1 m² of GFA

For both uses, the maximum land price is obtained by the residual method. What remains after subtracting building costs from the total project cost the business can justify is the most that can be paid for land. The only difference is the input figure.

Because both are calculated per 1 m² of GFA, markets with different floor-area ratios can be compared directly. If you need the amount per tsubo of site area, multiply the per-m²-of-GFA figure by “floor-area ratio × effective ratio × 3.3058”.

The assumptions are set as three scenarios rather than single values. The standard construction cost is based on MLIT’s “Statistics on Building Construction Starts” for 2024, with ±15% as the conservative/optimistic range. The key fact here is that hotels cost 1.71 times as much to build as condominiums. In 2024 construction starts, the planned construction cost of accommodation buildings was ¥591,000 per m² of floor (¥1,955,000/tsubo, 2,836 buildings), versus ¥345,000 (¥1,141,000/tsubo, 86,294 units) for for-sale reinforced-concrete apartment buildings. Hotels, with plumbing in every room and a high share of corridors, stairs and common areas, cost 1.7 times as much as housing to build the same floor area. This structurally depresses the hotel side’s bidding power.

Inputs for the Six Markets — Settled ADR ¥10,200–14,400, Condo Prices ¥660,000–1,360,000/m²

The six markets are one each in Tokyo, Osaka, Nagoya, Fukuoka and Sapporo, plus Kanazawa as a regional hub for comparison. Each market is defined at the city or ward level, and the hotel ADR is the average estimated settled ADR over the 12 closed months (September 2025 to August 2026).

The residential sale price is the median per-m² transaction price, by market, of condominiums completed in 2019 or later (near-new) and sold in 2024–2025, taken from MLIT’s “Real Estate Transaction Price Information”. New-build sale prices themselves are not published at the city/ward level, so the near-new transaction price — which can be aligned across all six markets from the same public source — is used instead. For reference, the Real Estate Economic Institute’s “Kinki Region New Condominium Market Trends, H1 2026” puts the new-build price in the city of Osaka at ¥1,159,000/m², broadly consistent with this article’s near-new transaction price of ¥1,232,000 for Osaka Chuo Ward. The same institute’s “Greater Tokyo” report for H1 2026 gives ¥2,226,000 for Tokyo’s 23 wards, but that is a 23-ward average including the six central wards (¥3,436,000) and differs from the level of Taito Ward alone.

Maximum Land Price Compared — Hotels Pay 3.71x Condos in Sapporo, 0.26x in Osaka

Plugging in the assumptions and back-calculating gives the following amounts that can be paid for land per 1 m² of GFA. The figures discussed are for the standard scenario.

Hotels come out ahead in two markets: Sapporo Chuo Ward (3.71x) and Fukuoka Hakata Ward (3.31x). In both, estimated settled ADR is around ¥14,300–14,400, near the top of the six markets, while near-new condo prices stay in the ¥700,000s per m². A high ceiling on room rates combined with relatively low residential sale prices is what lifts hotels’ bidding power.

Conversely, in Tokyo Taito Ward (0.50x) and Osaka Chuo Ward (0.26x) the residential side comes out ahead. Taito’s estimated settled ADR of ¥14,400 is almost the same as Sapporo’s, but its near-new condo price of ¥1,356,000/m² is 1.9 times Sapporo’s. Osaka Chuo Ward has a lower ADR of ¥12,300 and a condo price of ¥1,232,000/m². When a lower ADR is combined with a high condo price, the gap widens sharply.

In Nagoya Nakamura Ward and Kanazawa, the hotel’s maximum land price turned negative under the standard scenario. Their ADRs of ¥10,400 and ¥10,200 fall below the break-even ADR discussed below (¥10,800 in the standard scenario). This is a matter that comes before any comparison between uses: under standard specifications and yield assumptions, construction and design/supervision costs use up almost all of the total project budget. However, in the optimistic scenario (construction cost −15%, assumed OCC 88% (full-year average), GOP margin 50%, target yield 5.0%), both markets turn positive, so the picture can change depending on how specifications, operating efficiency and financing terms are worked out.

Checked Against Official Land Prices — Hotels Can Reach the Bottom Quartile to the Median of Commercial Land

Next, we look at where the back-calculated maximum land prices sit relative to actual land prices. For the commercial land in each market, we collected the survey points of the 2026 Official Land Price Publication (as of January 1, 2026) and the 2026 Prefectural Land Price Survey (as of July 1, published September 2026), used the newer prefectural survey where the same site appears in both, and took the distribution by quartile.

In Sapporo Chuo Ward, the hotel maximum of ¥231,000 per m² of GFA compares with ¥228,000 for the commercial-land median converted to GFA terms. The numbers work on land up to the median. In Fukuoka Hakata Ward, the hotel maximum of ¥224,000 compares with ¥146,000 for the bottom quartile and ¥373,000 for the median, so land from the lower to the middle part of the distribution is within reach.

In Tokyo Taito Ward, the hotel maximum of ¥230,000 is roughly level with the bottom quartile at ¥235,000 — the numbers only just work in the cheapest quarter of commercial land. The residential maximum of ¥462,000, on the other hand, exceeds the median of ¥341,000. If both bid for the same site in Taito Ward, the hotel can only compete for bottom-quartile land, while the condominium developer can take land above the median. In Osaka Chuo Ward the gap is even wider: the hotel’s ¥98,000 against a bottom quartile of ¥318,000. Under standard-scenario assumptions, acquiring vacant commercial land in Shinsaibashi/Honmachi to build a new hotel is hard to make pencil out.

Floor-Area Ratio Does Not Decide the Winner — Payment Capacity per m² of Floor Does

In competition for land, it is sometimes said that “the higher the floor-area ratio, the more housing is favored”. Checked against this article’s formula, that intuition is only half right.

For both hotels and housing, the maximum land price takes the form “payment capacity per 1 m² of GFA × floor-area ratio × effective ratio”. Because the floor-area ratio multiplies both by the same factor, the hotel-to-housing ratio does not change with the floor-area ratio. Whether the site has a FAR of 200% or 1,000%, in Sapporo Chuo Ward a hotel can pay 3.71 times what a condominium can. What the FAR changes is not who wins, but the absolute amount the winner can pay and the sale proceeds the landowner receives.

So what does decide the winner? The payment capacity per 1 m² of GFA itself — ADR for hotels, and the sale price for condominiums. Starting from there, the break-even points at which each use “works even with zero land cost” are as follows. These break-even points also do not depend on the floor-area ratio.

Under the standard scenario, the break-even point is an estimated settled ADR of ¥10,800 for hotels and a sale price of ¥606,000/m² for condominiums. The six markets’ ADRs range from ¥10,200 to ¥14,400, with Nagoya Nakamura Ward (¥10,400) and Kanazawa (¥10,200) below the line. On the residential side, all six markets exceed ¥606,000, with the distance from break-even ranging from 2.2 times in Taito Ward to 1.1 times in Kanazawa.

Sensitivity — ADR ±20% Moves the Maximum Land Price by ±79%

To see how strongly each assumption sways the result, we tested sensitivities against Tokyo Taito Ward’s standard scenario (maximum land price ¥229,860 per m² of GFA).

ADR has the largest effect: a ±20% swing moves the maximum land price by ±79.2% (+¥182,000 / −¥182,000). Next is GOP margin, where ±5pt means ±58.2%. Reducing GFA per room from 30 m² to 27 m² adds 44.0%, while enlarging it to 33 m² subtracts 36.0%. Construction cost ±15% moves it by ∓44.4%; lowering the target yield from 5.5% to 5.0% adds 39.6%, and raising it to 6.0% subtracts 33.0%. Assumed OCC ±5pt moves it by ±23.6%.

There are two takeaways. First, if you want to win in competition for land, ADR is the lever to pull first. Choosing a market that can command ADR at the site-selection stage, and raising the achievable ADR through room mix, brand and ancillary design at the planning stage, translate into bidding power more directly than any other variable. Second, the design of GFA per room is a lever almost as powerful as construction cost. Cutting 30 m² to 27 m² has nearly the same effect as lowering the construction unit cost by 15% (measured GFA per room in practice ranges from 13 to 196 m²).

Cross-Check on the Supply Side — Plans Pile Up Where Land Can Be Won

The bidding-power calculation is a desk exercise, but whether land is actually being secured shows up in supply. Below, planned projects from MLIT’s “Building Dynamics Statistics Survey” (based on building-permit applications) are shown alongside 2026 new openings confirmed via listings on major booking sites.

The population is small, so no conclusion can be drawn from the counts alone. Still, 4 planned projects totaling 714 rooms were found in Sapporo Chuo Ward, where hotels’ bidding power exceeded housing, while none were extracted in Nagoya Nakamura Ward or Kanazawa, where the hotel’s maximum land price was negative — consistent with the direction of the back-calculation. Both of the 2 projects (340 rooms) in Osaka Chuo Ward list hotel as their principal use.

At the prefecture level, 2026 new openings are led by Hokkaido with 135 properties and 2,702 rooms, followed by Tokyo with 80 properties and 2,324 rooms and Fukuoka with 56 properties and 1,372 rooms. Ishikawa had only 11 properties and 257 rooms, and Aichi 18 properties and 672 rooms.

When Hotels Can Win Land — Three Checkpoints

From the back-calculation across six markets, whether a hotel can go after a site alongside a competing use can be judged on the following three points.

From the land seller’s perspective, this structure is a guide to “which use to approach”. In central districts where condo prices exceed ¥1,000,000/m², residential developers tend to bid high; in markets where condo prices are sluggish but room rates hold up, hotel operators are competitive. Sapporo Chuo Ward and Fukuoka Hakata Ward are, at present, among the markets where the hotel side can put the highest price on land.

For non-hotel developers considering a hotel project, simply adding checkpoints 1 and 2 to the initial screening of markets goes a long way toward ordering the evaluation. Beyond that, concentrating planning effort on the three factors with the greatest sensitivity — ADR, GOP margin and GFA per room — is a rational use of limited time. It is also worth checking the financing ceiling (how much can be borrowed) and the carrying costs of land after acquisition.