You upgrade a Japanese equity data subscription to five years. Surely that means five years of prices immediately after an earnings announcement?
Then the specification sheet lands its punch: five years of daily bars, but only two years of minute bars and ticks.
As of October 30, 2026, this review found no verified public source offering decades of downloadable, free, legally reusable one-minute data covering a broad universe of individual Japanese stocks. There is an intriguing exception for a different instrument: Nikkei 225 futures minute data stretching back to 2001.
1. Why a five-year plan still gives you two years of minute data
J-Quants API lists four monthly base plans, tax included: Free at ¥0, Light at ¥1,650, Standard at ¥3,300, and Premium at ¥16,500. The historical windows for core datasets such as daily share prices are roughly two years with a 12-week delay on Free, five years on Light, ten on Standard and up to twenty on Premium. [1][2]
Equity minute bars and transaction-level ticks are a separate ¥5,500 monthly add-on. Their published historical window is two years regardless of the underlying plan. Premium does not transform the minute add-on into a twenty-year archive. [1][3]
Light plus the minute/tick add-on therefore costs ¥7,150 a month, tax included. Five years of TDnet disclosure documents are yet another paid add-on, not a hidden bonus bundled with Light. [1][4]
Think of buying five volumes of market history, only to discover that the high-resolution security camera recorded two volumes. Same market, different time machine.
2. But a free archive reaches back to 2001
225Labo advertises free downloads, after member registration, of Nikkei 225 futures bars from 2001 onward. One-minute, three-minute, five-minute and other intervals are listed; long historical series also exist for other index futures. [5]
That can be useful for researching volatile sessions, overnight moves and futures-market behavior. It does not answer what an individual Tokyo-listed stock traded at one minute after its company issued a disclosure.
Imagine an earnings upgrade at 2:00 p.m. Your research question concerns the issuing company's price at 2:01, 2:02 and 2:05. Twenty-five years of Nikkei futures history will not reconstruct those trades.
Finding decades of free minute bars and discovering they are futures is rather like ordering a fish and receiving an entire fishing boat. Impressive. Still not the order.
225Labo also restricts redistribution and resale of its downloads. Free access should not be confused with permission to republish the underlying data. [5]
3. What about free APIs, brokers and purchased history?
DataCedar covers multiple exchanges including Japanese stocks, but its documented free Explorer tier offers only seven days of one-minute history. A vendor's growing archive and the amount a free user can retrieve are different questions. [6]
When evaluating any GitHub project, charting service or third-party vendor, check six dimensions: individual-stock coverage, genuine one-minute resolution, years of history, bulk export, delisted and halted securities, and permitted use. A chart showing today's candles is not a reproducible multi-year research dataset.
For older paid archives, JPX's J-Quants DataCube sells transaction records as one-off products. Confirm the precise instrument, dates, price and license before purchasing. Tick records can be aggregated into minute bars, but that adds storage, processing and cost. [7]
4. How to test trading one minute after a disclosure
A five-year daily chart cannot establish the executable price one minute after a corporate announcement. Substituting a closing price for a minute-one entry price can make a backtest look much better than a realistic simulation.
Use two complementary tests. For the most recent two years, align available minute prices with publication timestamps and test entries after one, two and five minutes. Account for trading halts, auctions, liquidity, fees and price impact. For the wider five-year window, study next-day and later reactions using daily prices and appropriately sourced historical disclosures. Access to five years of disclosure files may require the separate TDnet product. [1][4]
J-Quants minute data is distributed daily for historical analysis, not as a real-time execution feed. [3]
An older sample helps measure whether results survive different market regimes. It does not magically supply the missing intraday prices.
5. Is Premium worth buying?
For the narrow objective of individual Japanese-stock prices immediately after disclosure, upgrading the base plan to Premium does not extend the minute archive beyond two years. Preserve the available data with timestamps, definitions and adjustment rules, and test it carefully before paying for more history. Consider DataCube's one-off historical products only when the additional period would materially change the study.
If the subject is index futures instead, 225Labo's long free archive is a genuine lead. First decide which market the strategy actually trades.
“Five-year plan. Five years of minute bars?”
“Two.”
“Premium, then?”
“Still two.”
The subscription rises faster than the minute-bar time machine.
Sources, checked October 30, 2026: J-Quants history windows [1]; JPX base-plan prices [2]; JPX minute/tick add-on [3]; JPX TDnet add-on [4]; 225Labo futures archive [5]; DataCedar documentation [6]; J-Quants DataCube [7].
