Asian economic releases can arrive during the European night, but their timestamps vary by publisher and series. Convert the official dated release time into your own zone before planning an observation. An unexpected number does not imply a fixed currency move.

Japan's share does not measure the Tokyo session

Desk location and trade time are separate questions. In a fictional USD 100 million sample, Japan sales desks have 12 million in our Tokyo window and 8 million outside it; other countries' desks have 18 and 62 million respectively. Each item is classified once by country and once by time interval.

Scroll sideways to see the full diagram

Desk location is not trade timeIn the hypothetical matrix, Japan desks have twelve million dollars in the chosen Tokyo window and eight million outside it. Other desks have eighteen million inside and sixty-two outside. The country total is twenty million; the window total is thirty million. Mosaic areas are proportional to amounts.Fictional turnover · USD 100m totalJapan desks: 20128Other desks: 801862Tokyo window: 12 + 18 = 30Outside: 8 + 62 = 70
Four cell areas are proportional to amounts. Row heights represent 20% and 80% shares; width subdivisions differ between rows.

Japan's row sums to 12 + 8 = USD 20 million, or 20% of the total. Green cells sum to 12 + 18 = 30 million, or 30% in the Tokyo window. The overlap is only 12 million. To calculate the second share you need timestamps; country shares alone do not provide them. BIS describes survey geography by sales desk location. Do not turn this into an hourly ranking or liquidity conditions for your own instrument.

Assumptions: fictional amounts in USD millions for one chosen period. The Tokyo window is defined as in the article: 09:00–18:00 Asia/Tokyo. Each record has one desk location and one timestamp; this example does not model the double-counting adjustments used in BIS surveys. Official material: BIS: FX turnover geography methodology.

What the Tokyo session is and when it runs

This article defines Tokyo as 09:00–18:00 Asia/Tokyo, or 00:00–09:00 UTC. Japan Standard Time stays UTC+09:00; this is a comparison convention, not exchange hours, a broker schedule or a guaranteed peak. For an example observer in Warsaw, the window is 01:00–10:00 on 27 March and 26 October 2026, but 02:00–11:00 on 30 March and 23 October. Poland’s clock changes fall on 29 March and 25 October that year; Tokyo’s offset does not change. Select named zones and a date, then separately check instrument sessions and holidays.

BIS geographic FX turnover is assigned by sales-desk location. Japan’s country share does not measure turnover within the Tokyo time window. Other countries’ desks can trade during that window, while Japan desks can trade outside it. The fictional diagram distinguishes these two dimensions; neither its country share nor its time-window share describes live liquidity available to your order.

Hypothetical USD/JPY conversion: equal JPY results can differ in USD

Buy USD 10,000 at 150 JPY/USD. Exclude spread, commission and funding and, solely for this model, convert the price result to USD at the exit rate without a fee. Exit at 151 gives +JPY 10,000 ÷ 151 ≈ +USD 66.23; exit at 149 gives −JPY 10,000 ÷ 149 ≈ −USD 67.11. Each move is 1 JPY per USD, but the conversion denominator changes. Using the entry rate for both would give an apparent ±USD 66.67. This is an explicit calculation convention, not a broker’s conversion policy or proof that opposite trading directions have different success rates. Preserve the result currency and conversion rate; the real specification and platform account-currency calculation need their own check.

Compare pair-specific costs and events

Yen pairs such as USD/JPY, EUR/JPY and GBP/JPY are candidates for observing Japanese releases and order flow. Check the actual quotes and execution records for your account. Neither the presence of yen nor Tokyo hours proves the tightest spread or cleanest chart.

Australian-dollar pairs may react to Asian developments because Australia’s trade and commodity exposure matter for the currency. That relationship is conditional: a Chinese release does not act exactly like an Australian one or dictate the size and direction of an AUD move.

EUR/USD and GBP/USD can also trade during the defined window when their instrument schedules allow it. Session names do not rank their costs. In a hypothetical USD/JPY long, assume a JPY account, a pip of 0.01 JPY/USD, purchase of USD 10,000 at Ask 150.020, entry Bid 150.000 and total commission JPY 100 for both sides. Assume full fills and zero funding, with no conversion needed. The entry spread is 2 pips; one pip is JPY 100 for this size. Selling at Bid 150.120 gives (150.120−150.020) × 10,000 = +JPY 1,000 before commission, or +JPY 900 net. Selling instead at 150.000 gives −JPY 200 before commission, or −JPY 300 net. Spread is already in the executed Ask/Bid difference, so do not subtract it again. A USD account would need the appropriate conversion, not simply a replacement of the JPY symbol with USD. These prices and fees are invented, not Tokyo-session observations.

The Chinese data that moves the night

Use each publisher’s dated calendar for Chinese manufacturing and services surveys, CPI, GDP, retail sales and industrial production. They do not share one universal 02:45 European release time. Preserve the source zone and convert the date before adding a reminder.

A data surprise may affect exchange rates through changing expectations and risk appetite. There is no fixed mapping from a one-point PMI surprise to a given number of pips. Observe the actual response and account for other simultaneous news.

The Bank of Japan and the risk of intervention

BoJ’s outline says that monetary-policy decisions are announced immediately after the meeting, without promising a fixed noon release. Its calendar separately lists Summary of Opinions and MPM Minutes, in principle at 08:50 JST on their own dates. An 08:50 JST publication is ten minutes before our 09:00 window and equals 23:50 UTC on the previous date. In Warsaw it converts to 00:50 on 27 March or 26 October 2026 and 01:50 on 30 March or 23 October. These four examples convert the clock; they do not assert that a document is published on every date. Check which document and publisher your reminder refers to, preserving date and zone.

Japan’s Minister of Finance decides on foreign-exchange intervention; the Bank of Japan executes operations as the government’s agent. Intervention is separate from a monetary-policy meeting and does not guarantee a particular pip change. Consider event exposure without assuming a fixed warning period.

How to trade the Tokyo session from your own time zone

Choose observation hours that fit your normal schedule and your ability to supervise risk. An overnight session is optional; a strategy does not require abandoning a practical daily routine.

If you use pending orders, verify acceptance, trigger rules and execution risk. A stop order does not guarantee its specified price. A scheduled notification can identify a planned release time, but it cannot determine whether the published number is a genuine surprise. Use a tested data alert if that distinction matters.

An active Asian window can be a candidate for measurement if your method and schedule support it. It does not guarantee the best liquidity. Compare dated boundaries and quote records for the Tokyo–London handover separately.

Next steps

  1. Keep broker trading hours separate from the Tokyo window convention. Convert 09:00–18:00 Asia/Tokyo for the observation date, then verify whether your provider and instrument are available throughout it.
  2. Choose an instrument list based on strategy evidence, costs and relevant events. Yen and Australian-dollar pairs are examples to investigate; there is no universal requirement to exclude other majors from this window.
  3. Add each planned Chinese release using its official timestamp and source zone. Set the notification after conversion for the date; do not reuse a universal 02:45 alarm.
  4. Review exposure before a Bank of Japan event using the official schedule and your strategy’s limits. A universal half-hour buffer does not cover an uncertain release time, a later communication or unscheduled intervention.