Forex Trading Sessions: Asia, London, New York
How the three major FX sessions work, when the overlaps produce peak liquidity, which pairs move most in each window, and what the session structure actually implies for a retail trader.
The foreign exchange market never closes during the trading week. That is technically true, and technically misleading. In practice, FX moves in three concentrated windows every 24 hours, driven by which parts of the world are awake and at work. Understanding the sessions is one of the highest-leverage pieces of knowledge for anyone new to the market, because it tells you when to expect activity and when the tape is a graveyard.
This article walks the three sessions, the two overlaps that produce the deepest liquidity, which pairs are most active in which window, and what the session structure actually implies for a retail trader.
Why sessions exist
FX is a decentralized global market. There is no physical exchange; trades happen bank-to-bank, dealer-to-dealer, and increasingly through electronic communication networks (ECNs). What creates the “session” structure is the fact that the biggest FX participants (commercial banks, investment banks, sovereign wealth funds, corporate treasuries, hedge funds) are staffed only during their local business hours.
When Tokyo is at work, Tokyo-based FX desks are quoting prices, providing liquidity, and making markets. When Tokyo goes home, that liquidity walks out the door. London and New York follow the same pattern. The world’s clock, in FX terms, is stacked geographically: Asian markets open first, London hands off from Asia, New York takes over from London, and there is a brief overnight lull before Asia comes back.
The result is three peak activity windows per day (three sessions), two windows of very deep liquidity where two sessions overlap, and one window per day where activity is genuinely thin.
The three sessions in GMT
All FX times are quoted in GMT (or UTC, which is essentially the same). The specifics shift by an hour twice a year due to daylight saving in each region, but the base framework holds year-round.
Asian session (Tokyo): 00:00 to 09:00 GMT. The day begins in Tokyo. The Asian session also covers Sydney, Hong Kong, Singapore, and Shanghai. Peak activity typically runs from 00:00 to 06:00 GMT (Tokyo late-morning through mid-afternoon). Volumes are lower than London and New York overall, but the yen pairs (USD/JPY, EUR/JPY, GBP/JPY, AUD/JPY) trade with genuine depth here because Tokyo is their home market.
London session (Europe): 08:00 to 16:00 GMT. London is the largest single FX center globally. Approximately 40 percent of all FX turnover happens in London. The session officially runs from 08:00 to 16:00 GMT, but activity typically peaks between 09:00 and 13:00. European pairs (EUR/USD, GBP/USD, EUR/GBP, EUR/CHF) are most active here, and any pair with a European currency component gets meaningful attention.
New York session: 13:00 to 22:00 GMT. The New York session picks up from London and runs through the US business day. Approximately 19 percent of FX turnover happens in New York, second globally to London. The session officially runs from 13:00 to 22:00 GMT, but the deepest activity is from 13:00 to 17:00 (when London is still open) and again around key US data releases at 12:30 or 13:30 GMT.
The two overlap windows
Peak liquidity in FX corresponds to session overlaps, when two major centers are staffed simultaneously.
Asia-London overlap: 08:00 to 09:00 GMT. A short one-hour window when Tokyo desks are wrapping up and London desks are just opening. Volumes rise noticeably. This is when many overnight positions established during the Asian session get filled or hedged.
London-New York overlap: 13:00 to 16:00 GMT. This is the golden window. Approximately 60 percent of daily FX turnover happens during these three hours. Both London and New York are fully staffed. US economic data releases (which typically publish at 12:30 or 13:30 GMT) hit in this window, and market-moving flow is at its peak. Spreads are tightest, order execution is best, and volatility (for those who want it) is highest.
Which pairs move in which session
The session structure produces a predictable pattern of which pairs are most active in which window.
Asian session (00:00-09:00 GMT):
- USD/JPY is the most active pair, with Tokyo corporate flow and Bank of Japan announcements typically landing 03:00-04:00 GMT.
- AUD/USD and NZD/USD are active, particularly around Australian and New Zealand data releases and RBA/RBNZ meeting days.
- Yen crosses (EUR/JPY, GBP/JPY, AUD/JPY) get meaningful attention because of Tokyo participation.
- European pairs (EUR/USD, GBP/USD) are relatively quiet with wider spreads.
London session (08:00-16:00 GMT):
- EUR/USD dominates, as the largest pair globally and a European home market.
- GBP/USD and EUR/GBP are highly active.
- USD/CHF is active with Zurich-based Swiss franc participation.
- All European crosses (EUR/JPY, GBP/JPY, EUR/CHF, EUR/GBP) trade actively.
New York session (13:00-22:00 GMT):
- All USD pairs are active as US traders react to European moves and to US data releases.
- USD/CAD picks up meaningfully with Canadian participation joining.
- US equity index correlations on FX pairs (particularly with the yen and Swiss franc as safe havens) are strongest in this window.
The dead zones
Two windows per day where FX genuinely thins out:
Late New York into Sydney: 22:00 to 00:00 GMT. New York closes; Sydney opens but with minimal weight. Spreads widen noticeably. Fill quality deteriorates. Order execution is more prone to slippage. If you are placing an order in this window, expect to pay for it.
The “second dip” during the Asian afternoon: 06:00 to 08:00 GMT. Tokyo starts winding down; London has not yet opened. Volumes ebb between the two peaks. This window is not as thin as the New York-Sydney gap, but it is noticeably quieter than either the Asian morning or the London morning.
A special note on Friday evening and Sunday evening. Friday New York close (22:00 GMT) is the end of the trading week; the market fully reopens at 22:00 GMT Sunday when Sydney comes back. The 48-hour window between Friday 22:00 and Sunday 22:00 is genuinely closed for most retail brokers (some quote indicative prices but do not execute). Weekend geopolitical events or unexpected policy news can produce large gaps at the Sunday open, so any position held over the weekend is exposed to unhedgeable overnight risk.
What the session structure implies for a retail trader
Four operational implications:
1. Trade when the tape trades. The reasonable time to be actively watching a European pair like EUR/USD is during the London and London-New York overlap windows (roughly 08:00 to 17:00 GMT). Outside those windows, the pair is quiet, spreads are wider, and price action is more likely to be noise than signal. For yen pairs, add the Asian session (00:00 to 06:00 GMT) to the reasonable window.
2. Data releases align with sessions. US data releases at 12:30 or 13:30 GMT land in the London-New York overlap, which is why those releases produce the largest single-instant moves of the trading day. European data releases at 09:00 GMT hit in the early London session. UK data at 07:00 GMT hits right at the London open. Japanese data at 23:50 GMT (previous evening in UTC terms) hits in the Sydney-Tokyo transition. Knowing when your pair’s data releases hit tells you when to expect the largest scheduled moves.
3. The weekend gap is real risk. Positions held over the weekend can gap materially on the Sunday reopen. Any position sizing should account for this. A stop-loss that would work fine during the trading week does not protect against a Sunday-open gap that jumps clean through the stop level.
4. Session-driven volatility patterns are stable. The pattern of “London morning is active; Asian afternoon is quiet; Sunday open can gap” has been stable for decades. Strategies that assume this structure (like scalping only in the London-New York overlap, or avoiding the Asian afternoon) have their assumptions built on a durable feature of the market rather than a fragile one.
The daylight saving complication
Twice a year, the session times shift by an hour because of daylight saving in the northern hemisphere. The specific dates when clocks change differ between the US, UK, EU, and Australia, so for a few weeks each spring and autumn there are session-timing anomalies. During those weeks, the London-New York overlap can be one hour longer or shorter than usual depending on which region has changed and which has not.
Practical response: use a broker platform that displays session times in your local timezone, and adjust your mental map of “when things happen” twice a year when the daylight-saving shifts kick in.
The takeaway
The FX market runs on a three-session structure: Asia, London, New York. The two session overlaps produce the deepest liquidity. Approximately 60 percent of daily FX flow happens in the London-New York overlap (13:00-16:00 GMT). Different pairs move most actively in different sessions. Two dead zones exist per day (New York-Sydney transition and Asian afternoon), plus the weekend closure. The session structure is stable, predictable, and one of the highest-signal pieces of context for reading intraday FX.
For the pair-specific session dynamics, read EUR/USD: The Fibre, GBP/USD: Cable, and USD/JPY and the Bank of Japan. For how trading costs vary across sessions (spreads widen in the dead zones), read Trading Costs Explained. For a wider view of what actually moves FX rates during peak session windows, read What Moves Exchange Rates.