In front of Kobe’s Sannomiya station, a 32-storey, roughly 163m-tall mixed-use tower is rising. It is the Kobe Sannomiya Kumoidori 5-chome District Type 1 Urban Redevelopment Project. A bus terminal, a hall and a library occupy the lower floors, offices the middle floors, and lodging everything from the 24th floor up. This “the higher you go, the more it becomes guestrooms” cross-section is a pattern that keeps reappearing in station-front redevelopments across Japan. This article uses the actual figures in the project plan published by the City of Kobe, together with room-rate data for the Sannomiya area, to show that the arrangement is not the outcome of design but the consequence of a funding plan.

56.6% of the funding is floor sale proceeds — breaking down the ¥92.2bn in the project plan

An urban redevelopment project is an area-based development method under the Urban Renewal Act. Rights to the former land and buildings are converted into floor space in the new building (rights-conversion floor space), and the surplus floor space created by intensifying land use becomes reserved floor space that is sold to third parties. The proceeds, together with public subsidies, cover the project cost. Whether the reserved floor space sells at the assumed price therefore all but determines whether the project stands up.

In the case of the Kumoidori 5-chome district, the project plan published by the implementing entity, Kumoidori 5-chome Redevelopment Co., Ltd., with the City of Kobe’s approval (initial approval March 2021; second revised approval 22 January 2025), sets out the funding plan down to the exact amounts. The total is ¥92,246m. The revenue breakdown is as follows.

Of the just under ¥92.1bn in revenue, public money accounts for ¥40.065bn — ¥37.0bn in subsidies plus ¥3.065bn in public facility administrator contributions — or 43.4%. The remaining ¥52.2bn, fully 56.6%, is proceeds from selling floor space. The subsidy cost-sharing ratios are also published: according to Hyogo Prefecture’s investment project evaluation report, land preparation costs and joint facility development costs are split 4.5/10 national, 2/10 prefectural, 2.5/10 municipal and 1/10 developer, while survey, design and planning costs are split 1/3 national, 1/6 prefectural, 1/6 municipal and 1/3 developer.

What this shows is a simple fact. Subsidies fill four-tenths of the project cost, but six-tenths has to be recovered by selling floor space. The design philosophy of the project is therefore governed from the outset by the question of how much floor space that sells at a high price it can create. When discussing funding routes for hotel development, the reason redevelopment schemes are so often chosen is precisely this distinctive monetisation device called reserved floor space.

Who holds which floors — breaking down 99,000㎡ by use

The project plan discloses the roughly 99,130㎡ of gross floor area by use and by floor level. Looking at it makes clear why the building’s cross-section runs in the order it does.

The uses of the lower floors are all but fixed by policy. The bus terminal is a national direct-control road project under the “Route 2 and Kobe Sannomiya Station Front Space Development Project Plan” formulated by the Ministry of Land, Infrastructure, Transport and Tourism and the City of Kobe, while the hall (a main hall of roughly 1,800 seats plus multipurpose space) and the library are civic facilities positioned within the City of Kobe’s own plans. Together they occupy 41,700㎡, or 42.1% of the gross floor area. That four-tenths-plus is not floor space intended to be sold high at market prices.

That leaves retail, office, lodging and parking to carry most of the ¥52.2bn in reserved floor sale proceeds. Of those, retail sits low in the building from B1F to 3F: the floorplates are large but the ceiling on unit value is tied to street-level retail rent levels. The office component spans 15 levels from 9F to 23F for 24,700㎡, with column-free floorplates of about 350 tsubo — among the largest in the Sannomiya area — as its selling point. And the lodging facility is placed at the very top, on floors 23 to 32.

What matters here is that different uses value floor height with different sensitivity. Office rents within the same building typically differ between low and high floors by only a few percent to the low teens. For a hotel, by contrast, the view feeds directly into the room rate. In a city like Kobe, hemmed in between sea and mountains with unambiguous scenic assets, that gap widens further. For the same square metre of floor space, the use that can place the highest marginal value on the top of the building is a hotel. That is why the upper floors become a hotel.

Building 1,019% on a 700% FAR site — the uplift creates the reserved floor space

The other precondition is floor area ratio (FAR). Querying the zoning data in the MLIT Real Estate Information Library for the coordinates of the project district shows the site is a commercial zone with a designated FAR of 700% and a building coverage ratio of 80% (immediately around Sannomiya Station it is 800%). The planned figures in the project plan, meanwhile, are a building site area of approximately 8,230㎡ against a FAR-counted floor area of approximately 83,870㎡, a FAR of approximately 1,019% and a building coverage ratio of approximately 93%.

At the designated FAR the FAR-counted area would stop at roughly 57,600㎡. The plan exceeds that by about 26,300㎡. That uplift is the source of the floor space created beyond the rights-conversion floor space — in other words, of the reserved floor space. The building coverage ratio is likewise approximately 93% against a designated 80%, but that is a level explicable by stacking up the allowances for fire-resistant buildings in fire prevention districts and for corner lots. An urban redevelopment project can equally be described as a device for institutionally securing an uplift in the FAR allowed under city planning and then monetising it in the form of floor space.

This is the point at which analysis of zoning, FAR and land prices — reading development headroom from the regulatory side — connects with analysis of the redevelopment scheme. The former asks where and how much can be built; a redevelopment project asks how to use the system to move the ceiling itself, and how to monetise the amount by which it moved.

The locational case — land prices up 41% in seven years, 133,000 workers within 1km

The grounds for expecting the reserved floor space to sell at the assumed price lie in the location. Querying the published land price survey points around the project district in the MLIT Real Estate Information Library shows that the two nearest points were both up 9.0% year on year as of 1 January 2026, with growth accelerating.

The survey point at Miyuki-dori 5-chome, 470m from Sannomiya Station, rose from ¥705,000/㎡ in 2019 to ¥995,000/㎡ in 2026, up 41.1% over seven years. The Yahata-dori 3-chome point, 580m away, rose from ¥755,000/㎡ to ¥1,010,000/㎡, up 33.8%, crossing the ¥1m/㎡ mark in 2026. After dipping into negative territory in 2021-2022, the acceleration from 2024 onward is unmistakable. Both are commercial-zone sites with a designated FAR of 600%, a building coverage ratio of 80% and fire prevention district status.

The scale of the transport interchange is also worth confirming. According to station passenger data in the National Land Numerical Information, daily boardings and alightings at JR Sannomiya Station peaked recently at 250,268 in 2019, fell to 183,956 in 2020, and had recovered to 231,870 by 2023 — 92.6% of the 2019 level.

Layered on top of that are Hankyu Kobe-Sannomiya (89,189 in 2023), Hanshin Kobe-Sannomiya (107,270), Kobe Municipal Subway Sannomiya (120,008), Sannomiya-Hanadokeimae on the same network (16,166) and Port Liner Sannomiya (66,406). The simple sum across the six stations is about 631,000 per day. The project plan cites the dispersal of stations and medium- and long-distance bus stops, and the resulting poor convenience of transferring between them, as a condition of the district — the picture being that consolidating the bus terminal is meant to improve it.

40 properties and 6,850 rooms within 1.2km — median estimated ADR ¥11,800, and a void in the ¥20,000s

So which price band will the lodging floor space placed there have to target? Extracting, from the hotels and ryokan with confirmed operations within a 1.2km radius of the project district, those with at least six months of estimated settled ADR over the most recent 12 months (August 2025 to July 2026) yields N=40 properties and 6,850 rooms in total. Their distribution is shown below.

The median estimated ADR is ¥11,800. Not a single property sits between ¥18,500 and ¥43,600 — the entire ¥20,000s band is empty. The top two are a 30-room boutique hotel on the Kitano slope (estimated ADR approximately ¥57,700) and a 116-room property towards the Former Foreign Settlement (approximately ¥43,700). In terms of category mix and rates across major cities, Sannomiya has the classic regional-hub composition of a thick economy layer plus a very thin high-rate layer.

This void does not mean the market will not tolerate high rates. Estimated settled ADR in Chuo-ku, Kobe has been rising steadily.

Comparing January-July averages: approximately ¥9,100 in 2024, approximately ¥10,500 in 2025 and approximately ¥11,000 in 2026. That is +4.6% year on year in 2026 and +21.0% against the same period in 2024. Supply and demand have also tightened. Estimated occupancy across Hyogo Prefecture in July 2026 was 89.3% for city hotels (42 properties, 8,072 rooms) and 90.9% for business hotels (145 properties, 16,690 rooms). For where central Kobe sits among the prefecture’s rate tiers, an analysis of Hyogo’s rate structure across five layers gives the full picture.

Demand is strong and rates are rising. That there is still zero inventory in the ¥20,000s is because much of the existing stock consists of limited-service properties built between the 1980s and the 2000s, and renovation alone cannot change either room size or the view. The high-rate band is a layer that does not exist unless it is built. The upper floors of a redevelopment are among the few pieces of floor space that can fill that void.

The ¥7.3bn the lodging floors carry — a back-calculated required ADR of about ¥44,000

Putting these figures side by side reveals the economic hurdle imposed on the lodging floor space. Allocating the ¥52.2bn in reserved floor sale proceeds pro rata by gross floor area, the 13,940㎡ lodging facility (14.1% of the gross floor area) bears approximately ¥7.34bn.

According to the press release, this lodging floor space is planned for 60-70 guestrooms (standard rooms of approximately 35㎡). Dividing by the midpoint of 65 rooms gives approximately ¥113m per room — by no means a low level as a per-room acquisition price for an urban luxury hotel.

Working backwards to the room rate this floor space needs in order to stand up as an investment: the assumptions are a capitalisation rate (NOI cap rate) of 5.0%, a GOP margin of 30% (the level of a full-service city hotel), a room-revenue share of total revenue of 50-70%, estimated occupancy of 70-75% and 65 rooms. On that setting, the required NOI is approximately ¥367m a year and required total revenue approximately ¥1.223bn.

Swinging the cap rate, in the case of a 60% room-revenue share and 70% occupancy (a full-year calculation basis), gives approximately ¥39,800 at 4.5%, approximately ¥44,200 at 5.0% and approximately ¥48,600 at 5.5%. Even at the extremes, the required ADR stays within a range of ¥34,400 to ¥51,500.

That is 3.4-4.1x the ¥11,800 median estimated ADR within 1.2km of Sannomiya. This multiple explains in figures why a hotel has to go on the upper floors. A limited-service property selling at prevailing local rates cannot recover its pro-rata share of the reserved floor sale proceeds. Selling the floor space high enough to make the project work leaves no option but to create product in a price band that does not exist nearby — and that requires the view and the room size, the differentiators that can only be built at the top of the tower.

In fact, the published plan comprises 35㎡ standard rooms, a 30m-wide infinity pool and spa on the top floor, restaurants and bars exploiting the view, plus a chapel and banquet space, using the 13,940㎡ of lodging floor space for 60-70 rooms. Gross floor area per room works out at approximately 200-230㎡ (199㎡ at 70 rooms, 232㎡ at 60 rooms), showing that a substantial share of the area is allocated to facilities other than guestrooms. Combining weddings, banquets and F&B to lower the room-revenue share moves the required ADR toward the upper rows of the table. The mix of uses is itself designed to bring the required ADR down.

How far does this pattern generalise — four metrics for reading redevelopment hotel floors

The Kumoidori 5-chome figures yield a set of practical metrics for assessing hotel floor space embedded in a redevelopment project.

The fourth multiple is the most practical of all. From an operator’s site-development perspective, it is a yardstick for the distance between the rate the redevelopment demands and the rate the operator’s own brand can achieve in that catchment. At a multiple of around 1.5x an existing format can be extended to fit; above 3x, both the brand and the product specification have to be imported from outside the catchment. This structure is why brands introduced into large station-front mixed-use developments tend to be pitched higher than the local market.

The risk factors are also worth setting out.

On the project timeline: the implementing entity was established in June 2018, the city planning decision was announced in March 2020, implementation was approved in March 2021 (revised approval March 2022), the rights conversion plan was approved in May 2022, and demolition began in June of that year, bringing the project to where it stands now. The project plan sets the implementation period through the end of December 2028, while the press release puts completion at around fiscal 2027. Viewed through the distance band over which proximity to a station affects price, a position roughly 250m from JR Sannomiya Station ranks among the very best conditions.

Conclusion

A hotel occupying the upper floors of a redevelopment tower is not an aesthetic choice. It is the result of working backwards to the mix of uses that maximises recovery, given that ¥52.2bn of the ¥92.2bn total project cost has to be recovered by selling floor space. Policy fixes the uses of the lower floors, floorplate economics select offices for the middle, and at the top a hotel — able to convert a view premium into rate — delivers the highest marginal value.

The lodging floor space in this project carries ¥7.34bn on a pro-rata basis, or approximately ¥113m per room. The required ADR is approximately ¥44,000, 3.7x the median of the 40 nearby properties. Because it cannot work at prevailing local rates, the answer is to build product the area does not have. The conclusion of “luxury on the upper floors” follows mechanically from the size of the reserved floor sale proceeds.

Anyone assessing a redevelopment project should start by opening the funding plan in the project documents and checking the share of reserved floor sale proceeds and the share of public floor space. Those two figures alone predict, with considerable accuracy, how high a rate the project will demand of its lodging product.

Related reading

• Hotel Development Capacity 2026: Zoning, FAR & Land Price in 5 Markets
• Kaihin-Makuhari 1,000-Room ¥30bn Mixed-Use: Breakeven ADR ¥13,200
• Own a Hotel from ¥100,000 — Security Tokens Open a Third Funding Route
• Station Premium Ends at an 8-Minute Walk: 8 Stations, 1,012 Hotels
• Hotel Category Mix Explains Half of City ADR: 34 Municipalities
• Hyogo’s Hotel Market in 5 Layers 2026: ADR from ¥10,700 to ¥29,600
• Hotel Build Times: 2.80-Year Median, 1.96 Solo vs 4.00 Mixed-Use
• Hotel vs Condo Land Rivalry: ADR-Implied Feasible Land Price Map