At its earnings briefing on November 13, 2026, Isetan Mitsukoshi Holdings unveiled a plan to rebuild five office buildings it owns in Shinjuku 5-chome as a single project and to bring in a hotel brand making its first entry into Japan. Room count, investment amount, gross floor area and contract structure have not been disclosed, but according to press reports the company is aiming for an investment return that exceeds the profit it could realize by selling the land. In this article, we convert the president’s reference to “100,000 guests a year” into a room count, then use surrounding land prices and Shinjuku Ward’s rate levels to quantify, as an editorial estimate, the conditions under which developing the site while keeping it in-house beats the land-sale benchmark.

Room stock in this article covers lodging facilities with 10 or more rooms (excluding leisure hotels) within a 1,000m radius of the area around Shinjuku 5-17. “Operating” means properties with sale prices confirmed on major booking sites in the most recent 3 months (September 2026–November 2026) (as of November 14, 2026, compiled by MetroEngines Research).

All room counts, development costs, NOI and required ADR figures in this article are editorial estimates, not Isetan Mitsukoshi Holdings’ plan figures. Assumptions are stated in each table.

The bottom line: even on company-owned land with no acquisition cost, once the option to sell is priced in as an opportunity cost, the investment break-even sits at more than three times the ward’s typical rate. Shinjuku Ward’s estimated settled ADR averaged ¥16,200 over the latest 12 months (November 2025–October 2026), whereas the ADR needed to recover the land-sale benchmark plus development cost at a 4.0% cap rate is estimated at ¥51,100–¥66,000 at 80% occupancy. Working backward from this condition, the choice of an upper-tier brand making its Japan debut looks consistent.

The Plan So Far — What Has Been Announced and What Has Not

First, the facts that can be confirmed from press coverage. According to reports by Nikkei, WWDJAPAN and FASHIONSNAP, the site is in Shinjuku 5-chome, a few minutes’ walk from the Isetan Shinjuku flagship store, on a triangular block bounded by Meiji-dori and Gyoen-dori. The plan is to redevelop five company-owned office buildings, including the Shinjuku Hanazono MJ Building, as a single project, with an exterior linked to the flagship store’s Art Deco style. The target opening is 2032, and the brand to be brought in will be making its first entry into Japan; its name has not been disclosed because negotiations are ongoing.

The numbers that will determine feasibility, however, have not yet been released. Room count, investment amount, gross floor area and contract structure (management contract, franchise or lease) are all undisclosed. Press wording also ranges from “entering” to “considering entering,” so this article treats the project as being at the stage where “a plan was presented at the earnings briefing.”

Existing department-store-related hotel cases mostly convert a site after the store closes (the economics of such conversions are covered in Department Store Sites: 6 of 14 Go Hotel). This plan is different: the flagship store keeps operating while adjacent company-owned offices are replaced with a hotel, so no land acquisition cost arises. In exchange, the company gives up the options of selling the land or continuing to lease it as offices, which puts opportunity cost at the center of the investment decision. The sections below translate this into numbers step by step.

Converting 100,000 Guests into Rooms — 190 to 430 Rooms, Centered Around 300

For 100,000 guests a year, the required room count more than doubles depending on the assumed average length of stay and occupancy. We therefore fixed guests per room at 1.7 and ran nine combinations: average stays of 1, 1.5 and 2 nights against occupancy of 75%, 80% and 85%. If 100,000 refers to guest-nights, the 1-night row applies; if it refers to unique guests, the 1.5-night and 2-night rows apply.

The results fall between 190 and 430 rooms, with 302 rooms at a 1.5-night average and 80% occupancy. Upper-tier brands with a high share of international visitors tend to see longer stays, so if 100,000 means unique guests the guide is 280–430 rooms, and if it means guest-nights, 190–215 rooms. For the aggregate estimates that follow, we use 300 rooms, the middle of the range, as a provisional scale. Layered onto the 12,461 operating rooms within 1km, 300 rooms is a +2.4% increase, and even across the 190–430 range it stays at +1.5% to +3.5%.

Site Assessment — Commercial Zone, 900% FAR, 250,000 Workers Within 1km

Because the exact site boundaries have not been disclosed, we centered our analysis on the area around Shinjuku 5-17 (35.6928°N, 139.7057°E). According to the MLIT Real Estate Information Library, this point is zoned commercial with a designated floor area ratio of 900%, a building coverage ratio of 80%, and a fire-prevention district designation. It is not within a flood inundation zone (maximum assumed scale), so flood-related constraints appear limited.

About 250,000 people work within a 1km radius, making this a location where retail and business districts overlap. Room stock is deep at 89 operating properties and 12,461 rooms, but most of it is in the business hotel segment, and the deluxe hotel segment has only 1 property with 65 rooms. The picture is one of placing an upper-tier brand, connected on foot to a flagship department store, in a trade area with few upper-tier rooms, which reads as a positioning that is unlikely to compete head-on with existing properties.

Surrounding Land Prices — Up 1.4–1.8x in 9 Years; Sale Benchmark of ¥5.12M–¥9.88M per m²

Next, as a yardstick for the value of selling the land, we look at nearby official land prices (Land Price Publication standard sites) and the Prefectural Land Price Survey (benchmark sites). Of the commercial survey points within 250m of the center point, we lined up four on the Shinjuku 5-chome side and around Meiji-dori. Levels vary with floor area ratio and road frontage. The benchmark site “Shinjuku 5-10” (Shinjuku 3-5-4) on Shinjuku-dori is a notch higher at ¥23.2M per m²; because the street character differs, it is excluded from the reference range.

The benchmark site “Shinjuku 5-20,” inside Shinjuku 5-chome itself, reached ¥5.12M per m² as of July 2026, up 16.4% year on year and 81.6% versus 2017. The 900%-FAR standard site “Shinjuku 5-1” stands at ¥9.88M, up 41.1% versus 2017. All four points share the same pattern: a pause around 2021, followed by wider gains from 2024 onward. The unit price of the project site itself has not been disclosed, but given that it is a 900%-FAR commercial block facing Meiji-dori, this article sets ¥5.12M–¥9.88M as the reference range for the sale case. That works out to roughly ¥5.1bn–¥9.9bn per 1,000m² of land.

Shinjuku Ward Rate Levels — ¥19,300 in October, +13.0% YoY

The baseline for the revenue side is Shinjuku Ward’s estimated settled ADR. According to MetroEngines Research data, it was ¥19,300 in October 2026 (N=102), up 13.0% from ¥17,000 (N=91) a year earlier. The latest 12-month average (November 2025–October 2026) is ¥16,200, with a clear seasonal pattern of peaks in spring and autumn and a dip in summer. The November estimate, based on listed prices at the time of the survey, is ¥22,800 (N=97), +21.3% against ¥18,800 (N=84) a year earlier, so the autumn high is stepping up further.

Compared with the other central wards, Shinjuku’s October level is about the same as Chiyoda’s, while Minato and Chuo sit in the ¥22,000–¥23,000 range and Shibuya in the ¥29,000 range. Year on year, Shinjuku’s +13.0% is second only to Minato, so the direction of rates is strong. Note, however, that ward figures are the median of covered properties, i.e. the level of a typical property, and should be considered separately from the rate band an upper-tier brand making its Japan debut would target.

Rate Distribution Within 2km — 3 Properties and 786 Rooms Above ¥50,000; None Overlap the Assumed Scale

We therefore widened the radius to 2km, including Nishi-Shinjuku (127 operating properties, 20,538 rooms), and examined the distribution of 81 properties with 15,154 rooms in the business hotel and city hotel segments for which a latest 12-month estimated settled ADR could be calculated. The median is ¥16,100, the top-25% threshold ¥27,000 and the top-10% threshold ¥38,100. Six properties with 1,958 rooms reach ¥40,000 or more, narrowing to 3 properties with 786 rooms at ¥50,000 or more.

In Kabukicho and Nishi-Shinjuku, properties opened in the 2020s achieve rates of ¥50,000 to over ¥120,000, which already shows that an upper tier can work in Shinjuku. No property, however, captures ¥50,000+ at a mid-size scale of 190–430 rooms. This article’s required ADR in the ¥50,000s–¥60,000s sits inside the rate range of existing upper-tier properties, but represents a new position in terms of scale. The upper-tier line-up around Kabukicho and the investment issues for new brands are also covered in our analysis of Noted Collection Shinjuku’s 207 rooms.

Transaction prices offer another check on levels. Dividing acquisition price by room count for properties held by listed hotel REITs around Shinjuku and in central Tokyo produces a staircase: around ¥52M per room for lodging-focused hotels, ¥130M for an apartment-style hotel and ¥250M for luxury.

Land Sale vs In-House Development — Break-Even at NOI of ¥5.8M–¥7.5M per Room

Now to the core question. When developing company-owned land, there is no outlay for the land, so yields look high if only development cost is used as the denominator. But selling the same land would bring in cash at the benchmark value, so the rational test for choosing development is whether the asset value of the completed hotel minus development cost (the residual land value) exceeds the sale benchmark. Our editorial assumptions are as follows.

Under these assumptions we set three combinations of ADR and occupancy. A is ¥50,000, close to the lower end of existing upper-tier properties; B is ¥80,000 for an upper-tier brand making its Japan debut; and C is ¥120,000 for the luxury band. For reference, we ran the same calculation at the ward’s typical level (latest 12-month average ¥16,200, 85% occupancy).

Leaving land out of the denominator, the yield on development cost reaches 5.7% even in A and 8.5% in B. Add the sale benchmark, however, and A drops to 3.0–3.9%, with a residual land value of ¥4.76M per m², just short of the ¥5.12M floor. In B the residual value is ¥12.77M, above the ¥9.88M ceiling. The break-even therefore lies between A and B: NOI of roughly ¥5.8M–¥7.5M per room is the level at which selling and developing are equivalent.

The next table converts this break-even into required ADR, for each combination of land benchmark and development cost, assuming 80% occupancy and a 4.0% cap rate.

In the ¥100M-per-room development cost column, the required ADR is ¥51,100–¥66,000, and ¥58,500 at the midpoint land price of ¥7.5M. That is 3.1–4.1x Shinjuku Ward’s latest 12-month average, out of reach for an extension of the ward’s typical property, yet within the range that upper-tier properties within 2km already achieve. At a 3.5% cap rate it would be ¥51,200, at 4.5% ¥65,900, and lowering occupancy to 75% moves it to ¥62,400. Assuming 300 rooms, development cost would be about ¥30bn, the corresponding site about 2,667m² with a sale benchmark of about ¥13.7bn–¥26.3bn, and annual NOI ¥1.71bn in A, ¥2.56bn in B and ¥3.59bn in C.

Sensitivity Analysis — Even with ADR −10% and Occupancy −5pt, Residual Value Stays Above the Benchmark Floor

For the base Scenario B (ADR ¥80,000, 75% occupancy), we check what happens if rate and occupancy come in lower.

Even when ADR falls 10% and occupancy falls 5pt, the residual land value is ¥8.93M per m², remaining in the upper part of the ¥5.12M–¥9.88M reference range. At a 25% NOI margin it is ¥8.77M, and at 35% ¥16.78M, so differences in operating efficiency move the result about as much as differences in rate. This is a simplified estimate, but it suggests that a brand able to command around ¥80,000 would have a certain cushion against the sale benchmark.

The two-way table below shows residual land value when rate and occupancy move together (fixed at development cost of ¥100M per room, 30% NOI margin and a 4.0% cap rate).

At ¥50,000, residual value clears the ¥5.12M floor only at 85% occupancy; at ¥60,000 it does so at 70% occupancy or higher, and at ¥70,000 at every occupancy level in the table. Clearing the ¥9.88M ceiling as well requires roughly 80%+ occupancy at ¥70,000 or 70%+ at ¥80,000. A ¥10,000 difference in rate moves residual value by ¥2.6M–¥3.4M per m² at the same occupancy.

Four Layers of Income — What Stands on the Hotel Alone and What Depends on Cross-Referrals

According to press reports, the company describes the business’s income in four layers. The estimates so far cover only the first layer; the remaining three are best viewed separately as potential upside.

The first layer is income the hotel generates on its own once the brand and operator are decided, and it is what lenders and asset managers find easiest to evaluate. NOI of about ¥2.56bn a year for 300 rooms in Scenario B equals 8.5% of development cost of about ¥30bn. The third layer’s “tens of billions of yen,” by contrast, is department store sales, not NOI; per hotel guest it equates to purchases of ¥200,000 at ¥20bn or ¥300,000 at ¥30bn. Sales and profit are measured on different scales and cannot simply be compared, but if even part of it materializes, it would be a thick layer of upside on top of the break-even measured on the first layer alone.

In terms of decision sequence, the robust structure is to secure, on the first layer alone, a level that matches the sale benchmark, and then stack layers two through four as upside. In that order, the asset value is underpinned regardless of how far cross-referral effects materialize, and synergies with the department store business can be evaluated as pure upside. How the contract structure affects customer data and the share of referral benefits is something to revisit once the brand is announced.

Risk Factors and What to Watch Next

There are three things to watch. First, the brand and contract structure, which will determine the NOI margin and the handling of customer data. Second, the room count and gross floor area, which will show how the floor area ratio is used and the share of F&B and retail. Third, the level of construction costs: depending on unit costs at the time construction begins, required ADR can move by several thousand yen. In each case, we will compare the company’s eventual disclosures with this article’s assumptions. For how far construction cost overruns push up required ADR, see our sensitivity estimate for hotel construction cost overruns.

Summary — For Owned-Land Development, Set the Rate That Matches the Sale Benchmark First

To sum up: Isetan Mitsukoshi Holdings’ Shinjuku 5-chome plan implies a scale of 190–430 rooms from the reference to 100,000 guests, in an environment where surrounding commercial land is priced at ¥5.12M–¥9.88M per m² and Shinjuku Ward’s estimated settled ADR stood at ¥19,300 in October. Once the option of selling the land is counted as an opportunity cost, the ADR needed for development to beat the sale benchmark is ¥51,100–¥66,000 (80% occupancy, 4.0% cap rate, development cost of ¥100M per room, editorial estimate), and the choice of an upper-tier brand making its Japan debut fits this condition.

The same procedure works for department stores, railways and real estate companies weighing how to use land they own. Set the sale benchmark from surrounding land prices, derive site area per room from the floor area ratio, and back out the rate that recovers the total including development cost at the cap rate. Then evaluate cross-referrals with the core business as upside, and the hotel’s standalone feasibility can be judged without being conflated with its spillover into the core business.

FAQ

Q. When and where will the Isetan Mitsukoshi hotel open?
A. According to press reports, the company will redevelop five company-owned office buildings in Shinjuku 5-chome as a single project, targeting an opening in 2032. The brand will be making its first entry into Japan, and its name has not been disclosed.

Q. How many rooms will it have?
A. The room count has not been disclosed. Converting the 100,000 guests a year the president mentioned at 1.7 guests per room, a 1–2 night average stay and 75–85% occupancy gives 190–430 rooms (editorial estimate).

Q. Under what conditions is developing better than selling the land?
A. In our editorial estimate, assuming 80% occupancy, a 4.0% cap rate and development cost of ¥100M per room, an ADR above ¥51,100–¥66,000 makes the completed asset value exceed the land-sale benchmark plus development cost.

Related Articles