Trading sessions and killzones

What a trading session is, and why it matters

A trading session is the window during which a financial centre is open and its participants are at their desks. The market may run around the clock; the people who trade it do not.

Currency and futures markets are almost continuous, but activity is not. That mismatch is what creates sessions: the market is open everywhere, the volume is concentrated somewhere.

The consequence is direct and checkable on your own chart. The volume of a ten in the morning candle has nothing in common with a three in the morning one. And volume is what decides the cost of an execution, as the lesson on large orders explains.

A level taken out with people on the other side and the same level taken out into a vacuum do not tell the same story, and no indicator will tell you which one you are looking at. The clock will.

⚠️ This is the only element in the whole corpus that is objective data rather than a reading. An hour cannot be reinterpreted after the fact, which is what makes it testable where the rest is not.

The three sessions, in London time

The Asian session, Tokyo's, is counted here like the other two, 8 am to 5 pm local time: 11 pm to 8 am in London in winter, and midnight to 9 am in summer, because Japan keeps the same time all year. Some sources start it an hour later. It is quiet on European pairs, livelier on the yen and the Australian dollar, and it is often where the narrow range forms that the rest of the day will break.

The London session runs from about eight to five. It is the heaviest in foreign exchange, and it is the one that most often gives the day its range.

The New York session is counted two ways. In foreign exchange the convention runs from 8 am to 5 pm New York time, 1 pm to 10 pm in London. The New York Stock Exchange trades from 9:30 am to 4 pm, 2:30 to 9 pm in London. The window that genuinely matters is the overlap with London, 1 pm to 5 pm London time: two centres open at once, so more people on the other side.

None of this is read in the abstract, it is read against the instrument. A US index has no reason to live by the rhythm of an Asian pair, and the internal calendar of a futures contract depends on the exchange that lists it.

Check these windows yourself rather than copying them: daylight saving shifts them by sixty minutes twice a year, and not on the same dates on both sides of the Atlantic.

Killzones, and what they claim to add

The ICT vocabulary cuts these sessions into narrower windows it calls killzones, usually one to three hours around an open. The idea is that most of a day's movement is decided inside those windows.

The claim is a strong one, and it has the merit of being checkable with no argument possible. All it takes is recording, across a hundred sessions, the hour at which the high and the low of the day printed.

That is a spreadsheet, two columns, one evening of work. The result is a histogram of hours, and it is the same for everyone: not an opinion, a count.

What you will probably find is a real concentration, less spectacular than promised, and above all different from one instrument to the next. That is the kind of result that does not sell well on video, and it is the one that serves you.

⚠️ A window that concentrates the extremes does not tell you which way to enter. It tells you when it is worth looking, which is information about time, not about direction.

The asian range, the night bracket London comes for

The asian range is the bracket drawn by the Asian session described above: the high and the low printed between midnight and eight in the morning. It is a range, not a level, and it gets drawn with two lines once the night has closed.

It works as a reference because it is narrow. On the major pairs, the night's range is only a fraction of the full day's, and that narrowness is what makes it readable: a wide bracket frames nothing, a tight one frames something.

What the corpus does with it fits in one sentence. London's first move often comes for one side of the asian range before the day leaves in the other direction. That is the manipulation of the accumulation, manipulation, distribution sequence, and taking the extremes of the range is a liquidity sweep like any other.

The trap sits in the definition, not in the idea. The window is not standardised: the ICT vocabulary draws it from 8 pm to midnight New York time, others start it at 7 pm, others again take the whole Asian session, as above. Two traders comparing 'the asian range' are measuring two different things, and their statistics do not compare. Write your window down alongside your records, or your own history stops being consistent the day you change your mind.

⚠️ A side of the range taken out is not a signal, it is a moment. On the nights when the Bank of Japan or the Australian central bank publish, the range is drawn by the announcement, and reading it as a manipulation stops making sense. Mark those sessions separately in your count rather than throwing them out, their number alone will teach you something.

ICT macros, the CBDR and the opens

Inside the killzones, the corpus cuts further. The ICT macros are windows of twenty to thirty minutes said to concentrate the mechanical moves, the most cited running from 9:50 to 10:10 New York time. The idea extends this lesson's: if activity concentrates, hours can be compared.

Before London, the vocabulary also sets the CBDR, central bank dealers range: the quiet bracket between 2 pm and 8 pm New York time, whose height some use as a yardstick to project the night's targets. And every calendar turn sets its reference prices, the midnight open of New York midnight and the true open of the quarterly cuts, above or below which the corpus reads price as expensive or cheap.

⚠️ None of these windows has any virtue in itself. Their one advantage is being written in advance, hence countable, and the column to keep is the same as above: what happens inside the window, on your market, over a hundred cases.

What the hour genuinely changes

Three things change with the hour, and all three can be measured on your own account, which is rare in this field.

The spread first. It widens when few people are around and tightens when the main session opens. On a small account that spread is a large share of the real cost of a round trip.

Range second. The same breakout strategy does not produce the same number of usable signals depending on the window, and a stop calibrated on London is often too wide for the Asian session.

Follow-through third. A break during a busy session has the volume to go somewhere; the same break in a quiet hour usually runs out of participants.

The best hour to trade looks like one question. It is three, and the useful way to ask it is: at what hour does my cost fall, and at what hour do my signals follow through?

The timing mistake that costs the most

It does not come from the theory, it comes from the time zone. An hour quoted in an American video is New York time, an hour quoted on a British forum is London time, and your chart is probably displaying something else again.

Set your chart once and for all to a zone you understand, and write that zone next to your records. Without it, thirty sessions of history become unusable the moment the clocks change.

The second mistake is reasoning in fixed hours when the market reasons in opens. The gap between Europe and the United States is not constant across the year, and for three to four weeks a year it is one hour smaller.

The third is running the same strategy across every session with the same settings. A stop that suits London is often too tight in New York and too wide in Asia, and the trader concludes the method has decayed when what changed was the clock.

⚠️ Record the hour of every trade in your trading journal. It is the easiest column to fill and one of the most revealing: sometimes a large share of a year's loss turns out to sit inside one or two hours of the day, and only that column shows it.

Practising: a hundred sessions, two columns

Pick one instrument and one time zone, and write them down. Then record, across a hundred sessions, the hour of the high and the hour of the low of the day. Two columns, a hundred rows.

Count how many fall into each hourly window. You get a histogram, and that histogram is the only serious answer to whether killzones exist on YOUR market.

Add a third column if you want to go further: the range of the session. You will then see whether the days when the extreme falls inside the stated window are also the days that move, which is not automatic at all.

Repeat on a second instrument. The comparison between the two will teach you more than the first hundred rows, because it tells you whether you found a property of the market or a property of your sample.

This record is worth keeping and rereading. It is one of the few pieces of work in this course whose result does not go stale from one season to the next, provided you redo it once a year.

Frequently asked questions