The Land Underneath Is Compounding: How Tokyo's Skyline Rebuild Quietly Reprices Central Ground

TL;DR: In Japan the building loses value toward zero on a fixed schedule, but the land underneath does the compounding. The 2026 official land prices confirm that central Tokyo’s land is repricing faster: residential land in the five central wards rose 13.0% to an average ¥2,138,700/㎡, against 8.5% (¥692,000/㎡) in the other 18 wards. The same data also qualifies the story. Minato and Chuo carry the central average, while Chiyoda, Shibuya and Shinjuku rose more slowly than inner-ring wards such as Taito, Shinagawa and Bunkyo; on commercial land the other 18 wards actually rose faster. For a foreign buyer, the asset that appreciates is the land share embedded in the unit, and the right filter is “core plus inner ring,” not “central five” as a label.

---

The one rule that governs every Japanese property

Almost every mistake foreigners make in this market traces back to a single misread: they price the building and ignore the dirt. In Japan it works the other way around. The structure is a depreciating asset on a legislated clock, written down toward zero across its tax life (how depreciation works for foreign owners). The land it sits on is not depreciated at all, because land does not wear out. Your condo is really two assets stapled together, one melting and one compounding, and over a long enough hold the land share is the part that decides whether you made money.

This is why “the building is old” is a weak objection to a central purchase and a strong one to a suburban one. In a central ward, the land share embedded in your unit can be the dominant portion of value, so structural depreciation is small against ground that keeps repricing. Out in the periphery, where land is cheaper and more abundant, the depreciating box is most of what you bought. The 2026 survey puts numbers on that divide: average residential land in the central five wards is about 3.1 times the average in the other 18 (computed: ¥2,138,700 ÷ ¥692,000). If you want one mental model for Tokyo, it is this: you are buying a land position with a building attached, not a building with some land thrown in.

From the desk — In the deals I’ve watched over the years, the foreign buyers who lose money are almost always the ones who fixated on how old the building was while ignoring how scarce the dirt under it had become. I keep telling buyers near a funded redevelopment that they’re not buying a condo, they’re buying a land position the city is actively making harder to reproduce, and the ones who hold that ground rather than flipping it are the ones who quietly come out ahead.

What the 2026 official land prices actually show

The evidence for (and against) the “central land reprices faster” claim is the official land price survey (令和8年地価公示, chika koji), conducted nationally by the Ministry of Land, Infrastructure, Transport and Tourism with a base date of January 1, 2026; the Tokyo ward tables were published on March 17, 2026 by the Tokyo Metropolitan Government Bureau of Finance. Here are the five central wards, the fastest and slowest movers, and the aggregates.

Ward	Group	Residential ¥/㎡	YoY	Rank of 23	Commercial ¥/㎡	YoY
Minato	Central 5	3,013,700	+16.6%	1	6,991,800	+14.0%
Chuo	Central 5	1,930,000	+13.8%	4=	11,385,400	+13.5%
Shinjuku	Central 5	1,131,600	+11.2%	12	6,159,400	+13.8%
Shibuya	Central 5	1,874,100	+11.0%	13	9,297,600	+13.8%
Chiyoda	Central 5	3,631,400	+10.7%	14	7,356,500	+13.1%
Taito	Inner ring	1,320,800	+14.2%	2	2,512,900	+19.1%
Shinagawa	Inner ring	1,244,100	+13.9%	3	2,470,400	+16.3%
Bunkyo	Inner ring	1,448,100	+13.8%	4=	2,276,300	+17.8%
Setagaya	Outer	779,900	+7.0%	19	1,334,800	+9.4%
Adachi	Outer	410,800	+6.3%	20	847,300	+8.5%
Nerima	Outer	480,500	+6.2%	21	1,002,000	+8.9%
Edogawa	Outer	441,100	+5.7%	22	857,400	+8.7%
Katsushika	Outer	387,500	+5.6%	23	720,100	+11.1%
Central 5 wards	Aggregate	2,138,700	+13.0%		8,211,500	+13.6%
Other 18 wards	Aggregate	692,000	+8.5%		1,709,400	+14.0%
Tokyo 23 wards	Aggregate	856,400	+9.0%		4,015,100	+13.8%

Source: 令和8年地価公示 (2026 official land prices), Tokyo Metropolitan Government Bureau of Finance, zaimu.metro.tokyo.lg.jp. Prices are ward averages of official standard sites; YoY is the average change at continuing sites; ranks computed by us from the 23-ward residential YoY column. All 23 wards: Tokyo 23 Wards Property Data 2026. Land prices are not condo prices.

Four findings come out of that table.

1. The central premium is real, and larger in yen than in percent. Central residential land rose 13.0% against 8.5% elsewhere, a gap of 4.5 points, or about 1.5 times the rate (computed). It is the second year running: in 2025 the central five rose roughly 12.0% against about 7.4% for the other 18 (directional, as previously reported here). The yen gap is wider than the percentage gap because the base is three times higher. Applied to the 2026 averages, the central move is worth roughly ¥246,000 per ㎡ against roughly ¥54,000 per ㎡ in the other 18 wards, about 4.5 times as much new value per square meter (computed approximately from the averages and YoY rates; the official YoY is a site-level average, so treat this as an order of magnitude).

2. “Central” is really Minato, Chuo and an inner ring. Minato’s +16.6% was the fastest residential rise of all 23 wards, and Chuo tied for fourth. The other three central wards did not lead: Shinjuku ranked 12th, Shibuya 13th and Chiyoda 14th. Nine non-central wards rose faster than all three of them: Taito, Shinagawa, Bunkyo, Meguro, Sumida, Toshima, Koto, Arakawa and Kita. Chiyoda, the most expensive residential land in Tokyo at roughly ¥12.0M per tsubo (computed: ¥3,631,400 × 3.3058), rose the least of the five. The highest price is not the fastest compounding.

3. Commercial land tells a different story. Across the 23 wards commercial land rose 13.8% against 9.0% for residential, which fits the use-competition force described below. But in commercial land the other 18 wards (+14.0%) outran the central five (+13.6%), led by Taito (+19.1%), Bunkyo (+17.8%), Nakano and Suginami (+17.5% each). Among the central five, only Minato beat the 23-ward commercial average; Shinjuku and Shibuya matched it; Chuo and Chiyoda trailed it. And in Minato (+16.6% residential vs +14.0% commercial) and Chuo (+13.8% vs +13.5%), residential land rose faster than commercial land. In the two wards that carry the central average, housing is now bidding for land at least as hard as offices and hotels.

4. The slow end is the large outer residential wards, not simply the cheap ones. Katsushika (+5.6%), Edogawa (+5.7%), Nerima (+6.2%), Adachi (+6.3%) and Setagaya (+7.0%) were the five slowest, and the first four rose more slowly than the Tokyo Metropolis residential average of +6.5%. Cheapness alone did not produce catch-up. Sumida, at ¥613,300/㎡ (one of the cheaper wards), rose 13.2%. What separates Sumida from Katsushika is proximity to the core, not the price level.

Three forces squeezing central land at once

The data above says where land is repricing. Three mechanisms explain why the core and the inner ring keep pulling the same finite plots in the same direction.

First, trophy redevelopment. Projects like Azabudai, Toranomon and Shibuya do not just add floor space; they re-rate an entire district. Azabudai Hills is a roughly 640bn yen (about 5.3bn dollars) project packing around 861,500 square meters of office, residential and hotel onto roughly 8.1 hectares (as of 2023 completion). A development like that flips a neighborhood’s identity from “place you work” to “place you want to live,” and the surrounding land gets repriced to the new identity. Minato, home to Azabudai and Toranomon, posted the fastest residential rise in 2026, which is consistent with that mechanism, but a ward average cannot isolate one project’s effect. The node-by-node map is in Tokyo’s redevelopment supercycle.

Second, use competition. The highest-value uses (office, hotel and luxury mixed-use) bid for the exact plots a condo developer needs. Across the 23 wards in 2026, commercial land rose 13.8% against 9.0% for residential (official survey, above). When offices and hotels outbid housing for dirt, new condo supply is choked before a single unit is drawn. The twist in Minato and Chuo, where residential land outran commercial, suggests the luxury-housing bid has caught up in the very core.

Third, conversion. Older, less competitive office buildings are increasingly being turned into housing. That sounds like new supply, and at the unit level it is, but it also confirms that central land is too valuable to leave underused, which reinforces the floor under prices.

Honest caveat: these three forces are reinforcing today, but they are policy- and rate-sensitive. A sharp move in interest rates or a zoning shift could slow the pace. The survey is also a snapshot as of January 1, 2026; since then the resale condo market has flattened (below).

The gap between core and periphery: what compounding does to it

Compounding gaps are unforgiving. A 1.5x faster annual rate, sustained, does not produce a 1.5x difference in outcomes; it pulls apart over a ten- or fifteen-year hold. As a pure arithmetic illustration (not a forecast), if the 2026 rates repeated for ten years, central land would multiply by about 3.4 times and other-ward land by about 2.3 times (computed: 1.130¹⁰ and 1.085¹⁰). Rates this high are not a sensible base case for a decade; the point is the shape. A persistent spread widens the gap in both percentage and yen terms, while the depreciating building does its slow, predictable melt underneath you.

Two corrections to the simple version: the spread has held for two consecutive surveys, not across every cycle, and it is concentrated. You do not need to time anything clever, but you do need to be on the fast side of the spread, and in 2026 that side was Minato, Chuo and inner-ring wards such as Taito, Shinagawa and Bunkyo rather than “the central five” as a block.

Redevelopment proximity is a measurable land-value mechanism

The link between a nearby rebuild and your land value is concrete. Minato-ku resale prices in the Azabudai and Toranomon orbit rose roughly 3x over 2016 to 2025 (directional). That is the tower pulling the whole neighborhood, not just its own units.

Futako-Tamagawa is a cleaner natural experiment outside the core. Land within roughly 200 meters of that station-area redevelopment rose about 44.9 percent, with a local population bump of around 6 percent (directional, illustrative). Set that against Setagaya ward as a whole, whose residential land average rose only 7.0% in 2026, 19th of 23 wards. A pocket around a station rebuild can diverge sharply from its ward average, which is why proximity is something you locate on a map.

And the hotspots cluster. Chuo-ku held 5 of the top-10 highest land-price-increase locations in 2025 (directional), tied to the Ginza, Tsukiji and Shintomicho hotel and condo pipelines. Redevelopment does not lift everything evenly; it creates localized compounding pockets that a buyer can target by buying near a confirmed, funded, station-adjacent project rather than chasing a ward average.

Caveat worth saying plainly: “near a redevelopment” only pays if the project is real and financed. A rumored plan is not a catalyst. Verify the pipeline before you pay for proximity.

The supply side is structurally choked, which favors today’s owners

The most bullish fact for anyone who already holds central ground is the one developers complain about most: they increasingly cannot secure central sites at all (as of writing). Greater Tokyo new-condo supply in 2025 was 21,962 units, the lowest since 1973, while the 23-ward average new-condo price rose 21.8% to ¥136.13M (Real Estate Economic Institute). Office, hotel and condo uses are fighting over the same dirt, so the plots needed for new central housing are scarce and expensive before construction even begins.

For a foreign buyer this is the whole thesis in one sentence. The supply of new central housing is throttled by the city’s own rebuild cycle, while demand to live in those re-rated districts keeps rising. Constrained supply meeting rising demand on a fixed quantity of land is the textbook setup for the land share to keep compounding. You are buying a slice of a resource the market is actively making harder to reproduce.

The honest counterweight has three parts. This is a multi-year thesis, not a flip: transaction costs in Japan are real (brokerage capped at 3% + ¥60,000 + tax, registration and acquisition taxes among them), and the building does melt. Land prices are not condo prices. And the resale market has cooled since the survey date: in August 2026, 23-ward resale condo contracts fell 20.6% year on year and the average contract price per ㎡ slipped 0.3%, the first year-on-year decline in 76 months (REINS Market Watch, Aug 2026; 23-ward average, so mix effects apply). Long land is a hold-through-the-cycle position, and a flat resale year is part of the cycle.

How to actually act on this

Stop shopping for buildings and start shopping for land positions. Concretely:

• Weight the core and the inner ring. Minato and Chuo led the central five in 2026; Taito, Shinagawa and Bunkyo rose faster than Chiyoda, Shibuya and Shinjuku. Use the ward pages and the 23-ward data table to check each ward’s 2026 figures before you narrow.
• Buy proximity to confirmed redevelopment. Target units near funded, station-adjacent projects, not rumored ones. The Futako-Tamagawa and Azabudai reads show proximity is measurable; make it a filter, not an afterthought.
• Estimate the land share, not the sticker price. On any unit you like, work out roughly how much of the price is land versus structure. A high land share in a fast-repricing ward is the asset you want; a high structure share in a slow outer ward is the one you do not. The buyer tools help you frame the math, and compare puts two candidates side by side.
• Underwrite for a long hold. Price in transaction costs and building depreciation up front, underwrite at flat rents, then let the land do its job. New to the vocabulary? Start with the glossary so the contracts read clearly.

The skyline going up across central Tokyo is not just new buildings. It is the city making its own scarce land scarcer, plot by plot, and repricing the dirt under nearby condos as it goes. The 2026 numbers say that machine is running hardest in Minato, Chuo and the inner ring. Own a land position on the fast side of that spread, and buy the ground before the next tower confirms what it is worth.

Related reading: Tokyo Is Building Its Own Manhattan — One Trophy Tower at a Time.