Methodology & Data Sources

JRE Analytics is built on official, primary-source data. This page explains exactly where our numbers come from, how we process them, and — just as importantly — what public data cannot tell you. Transparency about limits is part of how we keep the analysis honest.

Primary data source: MLIT

Our core dataset is the official real estate transaction-price information published by Japan's Ministry of Land, Infrastructure, Transport and Tourism (MLIT). These are actual recorded transaction prices — what buyers and sellers reported paying — not listing or asking prices. We also use MLIT / National Tax Agency official land price figures (kōji chika / rosenka) where relevant.

  • Transaction prices (取引価格情報): condominium, house, and land transactions by area, period, and property type.
  • Official land prices (地価公示・路線価): published annual land valuations and their year-over-year change rates.

How we obtain the data

We retrieve transaction records from MLIT's public data, then store and aggregate them in our database. Data is refreshed as MLIT publishes new quarterly/annual releases. The data vintage shown on a location page reflects the most recent release we have processed for that area.

How we process it

  1. Grouping: raw records are mapped to investor-friendly areas and grouped by property type (Condo / House / Land) and time period (by quarter).
  2. Aggregation: for each area, type, and period we compute summary statistics such as the median price per square meter, transaction counts, and ranges. We prefer the median because it is less distorted by a few unusually large or small deals.
  3. Trend & index: we derive year-over-year change and a relative “Investment Index” that combines price trend, transaction volume, and price level versus the broader market. The index is a comparison aid on our own scale, not a valuation or a recommendation.
  4. Currency conversion: USD figures are indicative conversions from JPY using a periodically updated exchange rate, provided only for convenience.

Confidence levels

Each area is labeled with a confidence level (High / Medium / Low) that reflects how much transaction history is available. Areas with few recorded transactions are marked lower-confidence: the sample is thin, so figures should be read as directional rather than definitive.

How we write about numbers (hedging policy)

Because public data has real limits, we deliberately hedge our language. When a figure is derived, approximate, or representative rather than exact, we say so — using words like “estimated”, “typically”, “approximately”, or “around”. For example, listing-vs-transaction gaps are described as a typical range (commonly ~5–15%), not a guaranteed discount. We attribute figures to their source (“according to MLIT…”) and avoid stating estimates as hard facts.

Limitations of public data

  • Closed transactions, not listings: MLIT data reflects deals that already closed, so it lags the live market and does not show current supply.
  • Uneven coverage: some areas and property types have many records; others have very few, which reduces reliability.
  • Reported & rounded: source values can be rounded or banded, and reporting is not guaranteed to be complete.
  • No property-specific detail: aggregates cannot capture the condition, exact location, or quality of any single property.
  • Timing: official releases are periodic; the newest market moves may not appear yet.

Corrections

Accuracy matters to us. If you spot a figure that looks wrong, please tell us via the contact page and we will review it against the source. For more on our editorial approach, see the About page.

Not advice

This analysis is informational only and is not legal, tax, financial, or investment advice. See our Terms of Use.