September 2026: Property tax with the residential-land special, Japan's land rules stated precisely, and one click to walk to the station

April was about data, May about personalization, June about judgment, July about risk and location, August about precision. September was about the rules. A Japanese property investment is governed by a dense layer of tax law, land law and advertising practice, and a figure that looks precise is only useful if it follows those rules. So this month we taught the analysis how Japan actually taxes residential land, rewrote every land-acquisition notice against the statutes themselves, and brought the Maisoku flyer in line with the disclosure and advertising rules an agent works under. Here's what shipped.

1. Property tax that reflects the residential-land special

Residential land in Japan is almost never taxed at the headline rate. Under the 住宅用地特例 (residential-land special), the taxable base for 固定資産税 (fixed-asset tax) falls to one-sixth of the assessed value for the first 200 m² per dwelling and one-third beyond that, and the base for 都市計画税 (city planning tax) falls to one-third and two-thirds respectively. Analyses previously applied the flat rates, which overstated the holding cost of almost any residential property.

When you don't supply a tax figure, the analysis now estimates both taxes with the special applied to the residential land, and the same calculation feeds the cash-flow figures, the tables and the narrative, so they agree. The Property Tax box on the analysis form now starts empty and marked Auto, so the estimate runs by default. If you have the actual figure from the tax notice, type it in and your figure is used instead.

The minpaku analysis now covers the other side of the same rule. Converting residential floor area to short-stay lodging can reduce or remove the special on the land, and the analysis flags that property-tax risk, applying the mixed-use thresholds where they're relevant: a residential share of at least half keeps the special on all of the land, a share between a quarter and a half keeps it on half, and below a quarter it is lost.

Source: 地方税法 (Local Tax Act) residential-land special provisions for 固定資産税 and 都市計画税

2. Japan's land-acquisition rules, stated precisely

Three different regimes can require a filing when land changes hands in Japan, and none of them is a government approval. Every notice in the analysis has been rewritten against the statutes:

国土法 (National Land Use Planning Act). The filing is made after the contract, within two weeks, and it applies to every buyer regardless of nationality. The area thresholds are now applied exactly as written: 2,000 m² or more in an urbanization area (市街化区域), 5,000 m² or more elsewhere in a city planning area, and 10,000 m² or more outside one. A parcel of exactly 2,000 m² is now correctly flagged.
外為法 (Foreign Exchange and Foreign Trade Act). Buyers who are non-residents of Japan file a post-acquisition report. This depends on residency, not nationality, and the analysis now says so.
重要土地等調査法 (the law on land near important facilities and border islands). In specially designated zones, buyer and seller give notice before the contract. It is a notification, not an approval: the filing does not screen or block the purchase.

This matters most on exactly the properties where a buyer from outside Japan is already most cautious. A notice that implies the state can refuse a sale inflates a risk that does not exist, and the analysis now describes each obligation as it actually is.

Source: 国土利用計画法 Art. 23; 重要土地等調査法 and the Cabinet Office FAQ; 外為法

3. Single seniors, considered as a tenant segment

Single-person households are 38% of Japanese households (2020 Census) and growing, and single seniors are a substantial renter segment for exactly the older studio and 1K stock many investors analyze. Rental analyses for single-occupancy units now consider them as a tenant segment: what they look for (ground-floor or elevator access, nearby clinics, guarantor-company acceptance) and what they tend to offer a landlord in return, which is long tenancies and very low turnover.

The analysis also prices the uncommon tail scenario that makes some landlords hesitate, a 孤立死 (solitary death) that goes undiscovered for a period, and sets out how professionals mitigate it: monitoring services (見守り), a solitary-death insurance rider on the landlord's fire policy, the government's model contract clauses for handling belongings (残置物モデル契約条項), the 終身建物賃貸借 (lifetime building lease), and certified rent-guarantee and housing-support providers. It also states plainly that, under MLIT's guideline, a natural death discovered promptly normally creates no disclosure obligation.

The framing is deliberate. The risk is priced and mitigated, never avoided: the analysis will not recommend declining tenants on the basis of age. Family-sized units are unaffected and do not grow this section.

Source: 2020 Census; MLIT guideline on disclosure of deaths in real estate transactions (2021); MLIT and Ministry of Justice model contract clauses

4. Maisoku flyers: seismic information and advertising consent

The Maisoku (マイソク) flyer now answers both halves of the question a buyer asks about earthquakes. The hazard band shows earthquake as the sixth hazard type alongside flood, tsunami, storm surge, landslide and inland flooding. Where the earthquake hazard has not yet been assessed for a property, the flyer says Not yet assessed rather than printing "None", because no location in Japan has zero seismic probability. The building block gains an 耐震基準 (seismic standard) row: a value recorded for the property prints as recorded, and a value derived from the year built is marked "(inferred from year built)" so it is never read as certified compliance. The flyer states the era as fact and leaves the risk discussion to the analysis, in line with Japanese advertising practice.

For agents working with listings from another brokerage, Tatemono IQ now records the listing's REINS 広告転載区分 (advertising reprint permission), read from the REINS sheet at import where it is present, and lets you attach a consent record from the listing broker (元付): the date, the media the consent covers, a note and a supporting file. The export advisory about another firm's listing now appears only when it applies, and stays out of the way when permission is on record.

Two flyer fixes round this out: the location map is now centred on the property's own area, and the approximate-location label renders correctly in Japanese.

Source: J-SHIS seismic hazard model, NIED; REINS 利用ガイドライン (usage guidelines)

5. One click from the property page to the walk to the station

Every property page now has a Walk to Station button alongside Analysis, Maisoku PDF Export and Edit. One click opens the walking route to the nearest major station, with the distance, the estimated walking time, the station's name and its lines, and the Virtual Street Tour ready to start. It is the same experience the map panel offers, now reachable directly from the property itself.

The neighbourhood data behind it got more accurate too:

Walking times in dense areas are routed. Properties with many stations nearby, typically in central city districts, were falling back to straight-line distances. New and refreshed lookups now measure the actual walking route.
Nearby places are classified more accurately. Shinto shrines are recognized as shrines rather than churches, and a courier company's service centres are no longer listed as post offices.
Cemeteries are no longer read as an amenity. Unnamed cemetery fragments no longer inflate the count of nearby places, and cemeteries no longer count toward a location's religious-facility rating.
Addresses are matched more precisely. Imported addresses are now classified as 住居表示 (residence-indication numbering) or 地番 (registry lot numbering) using the Digital Agency's Address Base Registry, so lot-numbered addresses are located in the right form.

Source: Digital Agency Address Base Registry (アドレス・ベース・レジストリ)

6. Reports that read correctly in every format and language

Every Japanese character, in every PDF. The analysis PDF now renders Japanese-specific characters in full, whatever the export language, so cost-panel labels and terms such as 旧耐震基準 (old seismic standard) print in full.
The Investment Assessment is in the PDF. The assessment label and its rationale now appear in the exported calculations panel, including the Indeterminate state, explained rather than presented as a negative verdict.
Japanese exports read as Japanese. The yield methodology card in a Japanese PDF carries the Japanese formula, and the acquisition-cost total reads as a single, correctly labelled percentage of the asking price.
Your language, throughout. The assessment rating in Japanese, Korean and Chinese analyses now appears in that language instead of English. Missing information is named in plain words rather than as internal field names, and structure codes such as RC are spelled out (Reinforced Concrete) on first use in list analyses.
Shared links look like the app. Analyses you share now render their tables, scorecard and investor questions properly in the browser, in every language.

7. When a trial ends, you can still buy what you need

When a free trial ended without a subscription, an account used to sit in a paused state where you could neither import nor analyze, and buying more required subscribing again. That state is gone. An account without a subscription is now Pay as you go: buy a property or analysis bundle whenever you need one, and your remaining balance is shown on Usage & Quota. Subscribing again is the ordinary plan checkout, whenever you're ready.

The bundle checkout itself is more forgiving too: a stray click outside the payment window no longer closes it and discards what you've typed.

Also improved in September

Calculators start from the right numbers. Opened from a property, the Cash Flow and Pro Forma calculators now prefill annual rent from the stored monthly rent and vacancy from the stored occupancy rate in the correct units.
Exports carry no internal identifiers. The list-analysis PDF names the list and the number of properties analyzed, with no internal record numbers.
One place card at a time on the map. Clicking a second nearby-place marker now replaces the open card instead of stacking another on top.
Plan descriptions match the trial. Plan descriptions on the pricing page no longer quote a trial length that contradicts the trial shown on the card, in any of the five languages.
Seismic data is credited correctly. Analyses that use earthquake data now credit J-SHIS and NIED as its source.