Forex Market Hours and Session Overlaps: When Liquidity and Volatility Actually Concentrate

Most coverage of forex market hours is a clock widget. Four coloured bars labelled Sydney, Tokyo, London and New York, a line for the current time, and a thread underneath arguing about which session is best. The widget is not wrong, it is just answering a smaller question than the one that decides outcomes. The session clock matters because of what sits underneath it: where the order flow concentrates, where the spread is tightest relative to the move, and where a position can be held without the conditions working against it. The clock is the surface. The order flow is the structure.

This guide is the structure. The four sessions in UTC and how daylight saving shifts them, why the London open and the London New York overlap are the genuine concentration windows, the character of the Tokyo session, the weekend gap, why a 24 hour market is not a uniformly tradeable one, and the rollover and triple-swap time a desk plans around rather than discovers.

By Ken Chigbo, Founder, KenMacro. 18-plus years in markets, London trading floor and institutional FX. The desk maps its hours to its pairs every day inside the MACRO MASTERY desk.

The four sessions, in UTC, and why UTC is the only honest reference

The market is conventionally divided into four sessions named after the centres that dominate them. Stated in UTC, so they do not drift as the seasons change: Sydney runs from the late evening into the early hours. Tokyo follows through the Asian morning. London opens around 07:00 to 08:00 UTC and runs into the afternoon. New York opens around 12:00 to 13:00 UTC and runs into the evening. There is overlap by design, which is the entire point of the structure, and the overlaps are where the day’s character is set.

The reason to anchor everything to UTC is that the local clock times move through the year. Europe and North America change to and from daylight saving on different calendars, and they do not align with each other to the day. The southern hemisphere shifts in the opposite direction, so when the north springs forward the south is falling back. The consequence is that the gap between, for example, the London open and the New York open is not constant in local time across the year, and the windows above slide by an hour as each changeover happens. Quoting one fixed local clock is how people end up trading the wrong hour for several weeks twice a year. The session structure is stable in UTC. It is not stable on any one wall clock.

The London open: the first real concentration window

When London comes in, the largest single centre for foreign exchange turnover opens its book. Activity steps up sharply against the thinner Asian conditions that preceded it. Spreads on the major pairs tighten as depth arrives, the day’s directional tone often starts to form, and the European data calendar lands into a market that now has the participation to react to it. This is the first of the two windows where the cost of trading is low relative to the size of the move available, because the order flow is there to absorb size without the price slipping as far.

The character of the London open is a transition from a low-information overnight regime into a high-participation one. The move that forms here has the volume behind it that the Asian session did not, which is the difference between a directional move that follows through and one that fades into thin conditions and reverses for no reason other than that nobody was there.

The London New York overlap: where the volume actually sits

This is the window that the clock widget cannot explain and the one that matters most. For a few hours, the two largest pools of currency order flow are open at the same time. London is the biggest centre for foreign exchange turnover. New York is the largest dollar centre. When both desks are live, the depth of resting orders is at its greatest for the day, the major-pair spreads are typically at their tightest, and the largest share of the day’s range tends to be made inside this window.

Why this is structural rather than incidental: the order flow that moves currency is institutional. Corporate hedging, reserve management, real-money rebalancing and the bulk of the speculative book run through the London and New York desks. When both are open, the participation that can absorb and drive size is at its peak, and the US data calendar, which is the densest macro release schedule for the dollar, lands directly into it. More volume has two effects that matter for a position. A trade can be moved without the price slipping as far against it, the cost is lower. And a genuine directional move has the participation behind it to continue rather than stall, the follow-through is more reliable. The overlap is not the best window because activity is good in itself. It is the best window because that is where the depth and the lowest cost relative to the move are, simultaneously.

The Tokyo session character: ranges and the yen bloc

The Asian session has a different personality and trying to trade it like London is a common, expensive mistake. On the European and US major pairs it is generally the quieter part of the day, narrower ranges, less follow-through, more mean-reverting drift inside a band. That is not a flaw to fight, it is the regime to respect. The participation that drives sustained trend in those pairs is largely not at the desk yet.

Where the Asian session does carry its own character is the yen bloc and the antipodean currencies: pairs involving JPY, AUD and NZD. Tokyo is the centre for yen flow, and the Australian and New Zealand data and the China-linked sentiment that moves AUD and NZD land in this window with the relevant desks live to react. So the Asian session is not dead, it is selective. It rewards a focus on the currencies whose order flow actually concentrates there and punishes treating the quiet majors as if London-style continuation were available. Knowing which pair belongs to which session is the difference between trading the regime and fighting it.

Why a 24/5 market is not uniformly tradeable

The market trades around the clock for five days because as one centre closes another opens and there is no single exchange that shuts it. That fact is true and almost useless on its own, because around the clock is not the same as uniformly tradeable. The variable that matters, liquidity, and its two consequences, spread and the size of moves, swing enormously through the day.

There are dead zones. The handover where the US has gone home and Asia is not yet fully in is thin. Public holidays in a key centre hollow out the book even during what is nominally an active session. The hours when no major centre is properly open are wide-spread, low-depth conditions where the cost of being in a position is high relative to whatever small move is available. A spread that is a small fraction of a typical move in the overlap can be a large fraction of the available move in a dead zone, which quietly inverts the economics of the same strategy depending purely on the clock. The market being open is not a reason to be in it. The session structure tells you when being in it is cheap and when it is expensive, and that interaction with cost is the part the widget never shows.

How a desk maps its working hours to the pairs it trades

The practical synthesis is not a search for the single best session. It is matching three things: the pairs being traded, the session where those pairs concentrate liquidity, and the hours the trader can actually be at the desk with full attention. A desk does not trade every pair in every session. It trades the pairs whose order flow is live in the window it is working.

That mapping is concrete. Major dollar and euro and sterling pairs get worked through the London open and the London New York overlap, where their depth and tightest spreads sit. Yen and antipodean pairs get the Asian window where their flow concentrates. A trader who can only be at the desk during the Asian hours is structurally better off focusing on the pairs that actually move then than forcing the European majors into a window that does not pay them. Forcing a pair into the wrong session is the same error as forcing a setup into the wrong regime: the technical work can be clean and the context still does not support it. The hours are part of the edge, not a logistical afterthought.

The weekend gap and carried risk

The market closes late on Friday in UTC and reopens late on Sunday in UTC. The world does not stop over that window. News, geopolitics and policy still happen, but there is no continuous price to absorb them in real time. When the book reopens, the first prints can be away from the Friday close, a gap, and after a significant weekend event that gap can be large. Spreads at the Sunday reopen are also typically wider while depth rebuilds.

The structural point for risk: a stop manages a continuous market. It does not manage a gap. A position carried over the weekend is exposed to a discontinuous move that can open straight through where a stop sat, so the realised loss is not the planned loss. This is a known property of the session calendar, not bad luck. It is sized for deliberately, by treating weekend-carried exposure as a different risk category from intraday exposure, rather than discovered after a bad Sunday open.

Rollover and the triple-swap day

One more fixed point in the session calendar is the rollover. Each day, positions held open are rolled to the next value date and a financing amount, the swap, is charged or credited depending on the interest-rate differential and the direction held. Rollover is conventionally tied to the New York close, around 22:00 UTC, with the exact clock time moving with daylight saving like everything else.

Because spot foreign exchange settles two business days forward, the weekend has to be accounted for on a single weekday. Conventionally that is the Wednesday rollover, where three days of swap are applied at once, the triple-swap day. For a desk that holds positions across rollover this is a scheduled, knowable cost or credit, not a surprise. It feeds the cost calculation on anything carried overnight, and the midweek triple charge is planned into the decision to hold rather than noticed afterwards on the statement. Cost that is scheduled is cost that can be managed. Cost that is discovered is cost that has already happened.

Related reading

• The free macro framework (the five-lens regime read that decides which window to work)
• Why most forex traders lose money (the cost and structure failures session timing interacts with)
• How to plan your trading week (mapping the calendar onto a small number of high-conviction situations)
• How to trade FOMC (the US catalyst that lands directly inside the London New York overlap)

Frequently asked questions

What are the four forex trading sessions?

Sydney, Tokyo, London and New York, named after the centres that dominate them. In UTC: Sydney late evening into the early hours, Tokyo through the Asian morning, London around 07:00 to 08:00 UTC into the afternoon, New York around 12:00 to 13:00 UTC into the evening. The local clock times shift by an hour through the year as Europe and North America change daylight saving on different calendars, so UTC is the only stable reference.

What is the best time to trade forex?

There is no single best time for everyone. Liquidity and volatility concentrate most in the London New York overlap, when both of the largest centres are open at once, and at the London open. These windows carry the largest share of daily volume and the tightest major-pair spreads, so the cost of trading is lowest relative to the move. The right window is the one that overlaps both your pairs and the hours you can be at the desk with full attention.

Why is the London New York overlap so important?

Because the two largest pools of currency order flow are open at the same time. London is the biggest foreign exchange centre, New York the largest dollar centre, so depth of resting orders is at its greatest, major-pair spreads at their tightest, and the largest share of the day’s range tends to be made then. More volume means a position can be moved without the price slipping as far, and a directional move has the participation to follow through.

Is the forex market really open 24 hours?

It trades around the clock from the Sunday open to the Friday close in UTC, roughly 24 hours for five days, because as one centre closes another opens. But around the clock is not uniformly tradeable. Liquidity, spread and move size vary enormously. Handover gaps, holidays in a key centre and hours when no major centre is open are thin, wide-spread conditions where cost is high relative to the move. The market being open is not a reason to be in it.

What is the weekend gap in forex?

The market closes late Friday in UTC and reopens late Sunday in UTC. Over that closed window news still happens but there is no continuous price to absorb it, so the reopen can print away from the Friday close, a gap, sometimes large after a major weekend event, with wider spreads while depth rebuilds. A stop cannot manage a gap, so weekend-carried exposure is a different risk category that is sized for deliberately.

What time is the rollover and triple swap in forex?

Rollover is the daily point where open positions roll to the next value date and the swap is charged or credited, conventionally tied to the New York close around 22:00 UTC, with the exact clock time shifting with daylight saving. Because spot settles two business days forward, the weekend is accounted for on one weekday, conventionally the Wednesday rollover, where three days of swap apply at once, the triple swap. A desk treats this as a scheduled cost, not a surprise.

Educational analysis only, not financial advice. Past performance does not guarantee future results. Always manage risk and never risk more than you can afford to lose. This is macro education and scenario framework, never a signal or a recommendation to trade.