ICT daily bias

Every ICT setup assumes a direction, and the daily bias is where that direction is decided, before the session starts.

What a bias is

A daily bias is a forecast with three parts: a direction, a target, and the session it covers. In ICT teaching the direction is framed through liquidity. Above the previous day’s high rest the buy-side orders and below its low the sell-side orders, as the buy-side and sell-side liquidity entry describes; a bullish bias expects the day to reach the orders above, and a bearish bias the orders below. It is a hypothesis about which side the day takes, not a trade. Entries still come from the rest of the vocabulary, inside the day.

The three inputs

Teaching reads three things before the session:

• Daily order flow. Whether recent daily candles have been taking out previous highs or previous lows. A run of days trading above each prior day’s high reads bullish; a run trading below each prior day’s low reads bearish.
• Imbalances to rebalance. The nearest fair value gap or other imbalance on the daily chart. An unfilled gap behind price is a place the day may return to before continuing, and it can point the day against the order flow.
• The next draw on liquidity. The draw on liquidity the higher timeframe points at: an old high or low, or an unfilled gap.

When the three agree, the bias is the direction they share. When they disagree the call is weaker, and a written record of which inputs disagreed is what later lets a reader see whether those days behave differently.

A worked case

Constructed numbers. On GBP/USD the previous day’s high was 1.2765 and its low 1.2702. Each of the last three daily candles traded above the high of the day before, so the daily order flow reads bullish. The nearest daily imbalance is a gap from 1.2680 to 1.2690, below the previous day’s low. The next obvious buy-side pool is last week’s high at 1.2798. Two inputs point up and one points down, toward the gap below price. A trader who weighs the order flow and the weekly high calls the day bullish, with the previous day’s high at 1.2765 as the first target and 1.2798 beyond it, and writes down before the open that a trade below 1.2702 before 1.2765 would score the call as wrong.

How the day is expected to unfold

Teaching pairs the bias with a picture of the session. On a bullish day the Asian session builds a range, London trades below the Asian low in the move the judas swing names, and New York delivers the move toward the target; a bearish day mirrors it. The picture is a template for the order of events, and the kill zones are the clock it runs on. A day can reach its target without following the template, and a bias is scored on where the day went, not on whether it followed the picture.

Daily inside weekly

A daily bias is usually made inside a weekly one. The weekly profiles are templates for when in the week the high and the low form, and a week read as bullish can still hold a bearish day, such as the early-week decline that sets a Tuesday low. The campaign’s frozen definitions compute the daily and weekly bias together and keep a resolution ledger for both, so that each call is scored on its own outcome.

Which previous day

“Previous day’s high” sounds unambiguous and is not. The day’s boundaries depend on the convention the chart uses. Currencies and futures conventionally close the day at 17:00 New York time; many currency platforms draw daily candles from midnight on the server’s clock; and United States equity index futures also have a regular session, from 09:30 to 16:15, whose high and low differ from the full day’s. The same instrument can show three different previous-day highs, and a bias scored against one of them is not the same forecast as a bias scored against another. The convention is part of the bias, and it is stated with it.

The bias and the other levels

Several other tools wait for the bias. The central bank dealers range projects the day’s high and low only once a buy day or a sell day has been called. The weekly and daily opening gaps, the new week opening gap and the new day opening gap, are read as destinations in the direction the bias points. And the IPDA data ranges set the larger frame: a bullish call made from the discount half of the 20-day range, pointing toward its high, is the kind of agreement the teaching looks for. None of these tools supplies the bias; each assumes one.

Scoring a bias

A bias is a forecast only if it can be wrong, and it can be wrong only if the rule for wrong is written first. A usable rule has three outcomes: correct if the day reaches the named side’s level before the other side’s, wrong if it reaches the other side first, and unscored if it reaches neither. The rule is written before the open, with a time stamp. Over many days the hit rate means something only against a baseline, such as the simple rule of calling the direction of the previous day’s candle. A way of forming the bias that cannot beat that rule is not adding information, however well the charts that illustrate it look.

Stating one, step by step

1. Before the session, mark the previous day’s high and low and the week’s high and low.
2. Read the daily order flow: which extremes recent days have taken.
3. Mark the nearest daily imbalance, and whether price is likely to revisit it first.
4. Name the draw on liquidity on the higher timeframe.
5. Write the bias: the direction, the first target, the level that would score it wrong, and the time.
6. Score it at the close, and keep the record, including the days it was wrong.

Common mistakes

Setting the bias after the open, when the day has already shown its hand. Changing it at midday and keeping the second call as the record. Never scoring it, so that every day becomes a lesson rather than a result. Treating the bias as an entry signal. And forming it on a lower timeframe than the day it forecasts.

What a bias does not settle

A stated bias is checkable in the simplest way a forecast can be: it named a side, and the day either reached that side first or did not. Whether a given way of forming the bias beats a simple baseline over many days is a measurement, and the daily bias driver has a row in the claims ledger, where that verdict renders. The two lenses keep the three inputs, which anyone can read from a chart, apart from the claim that they forecast the day.