ICT daily bias: how to call the day's direction before the session opens

The short answer

ICT daily bias is the single directional call, bullish or bearish, that a trader commits to for one trading session, and it is decided from the higher timeframes before the London open rather than read off the live chart as it prints. ICT stands for the Inner Circle Trader, the public name of trader Michael Huddleston.

I treat bias as the first decision of the day and the one everything else hangs off. Once the bias is set, I only look for setups that agree with it, and the method treats trading against your own bias as the most common way traders lose.

The reason a dedicated page exists for this one idea is that most traders can execute every step that comes after the bias but cannot form the bias itself. Forming it is a process, not a feeling, and this page lays that process out in full.

If the wider framework is new, the Smart Money Concepts hub sets up the method this sits inside.

What daily bias means in ICT, and what it is not

Daily bias is a conviction about direction for the session ahead. ICT traders use the term for the bullish or bearish read formed from the daily and weekly charts before price starts moving, and it governs which trades get taken once the session opens.

I am blunt about what bias is not, because the SERP is full of pages that blur it. Bias is not a prediction of where price will close, it is not a trade entry, and it is not a position you hold.

It is a filter, and its only job is to keep you on the right side of the day's likely travel.

The distinction from trend matters too. A trend describes what price has already done, while a bias is a forward call about what price is likely to do next, and the two can point opposite ways when a market is turning.

Why bias comes before everything else

In the ICT method, bias is step one of every trade, and skipping it breaks the rest of the sequence. Without a bias you cannot pick a draw on liquidity, you cannot decide whether a sweep is fuel for your move or a warning against it, and you have no rule for which setups to ignore.

I learned this the hard way by taking clean-looking setups with no bias and watching them fail into the wider direction. The setup was real, the timing was off, because I had no read on which way the day was actually being delivered.

The method treats the session as a delivery from one pool of resting orders to another, which is the buy side and sell side liquidity read. Bias tells you which pool is the destination, and that single call shapes every entry that follows.

The top-down read, from monthly to weekly to daily

Bias is built from the top down, never from the timeframe you trade. I start on the highest timeframe that matters and let each lower frame refine the read rather than override it.

The monthly chart sets the regime, the weekly chart sets the direction institutions are said to be delivering this week, and the daily chart sets the specific target for today. Each frame narrows the lens, and the daily bias is the last, most specific output of that chain.

I resist the urge to form bias from the one-hour or lower. A lower timeframe shows noise and intraday swings, and reading bias there is how traders end up long in a bearish delivery because a five-minute rally looked convincing.

The five-step daily bias process

The framework condenses into a repeatable five-step process, and running it the same way every day is what turns bias from a guess into a habit. Each step feeds the next, and skipping one weakens all the rest.

Step one is weekly structure. I read the weekly candle and the draw on liquidity it implies, asking which major pool of resting orders the week is travelling toward, and that sets the macro direction bias should agree with.

Step two is daily structure. I mark the daily highs and lows and read the most recent market structure shift, because the last shift on the daily often telegraphs the next leg.

Step three is the draw on liquidity. I name the specific target, usually the prior day high or low or an equal high and low, and that target is what the daily bias is betting on price reaching.

Step four is premium and discount. I locate price inside its dealing range, because a bullish bias is far stronger when price is in the discount half and a bearish bias is stronger in the premium half.

Step five is the session scenario. I map how the day could unfold through the killzones and the Power of 3 sequence, which sets the times I expect the move to deliver.

By the end of step five I have one direction written down. That written call is the bias, and the rest of the session is execution against it.

The prior day high and low as the core bias anchor

The prior day high and low, shortened to PDH and PDL, are the single most-used anchors in daily bias work, and for good reason. They are the most obvious levels on the chart, which means they hold the densest clusters of resting orders.

The read is simple. If the daily candle closed strong above the PDH, the next session is biased up toward the next pool above.

If price swept the PDH and closed back below it, the next session is biased down, because the buy stops above were cleared and the move reversed.

ICT traders call this the next day model, and it is the cleanest mechanical version of bias for traders who want a rule rather than a judgement. The model uses the relationship between the close and the prior day extremes to set the call before the open.

I anchor on PDH and PDL because they are unambiguous. Every trader watching the instrument sees the same two levels, which is exactly why they attract the order flow the method expects to be targeted.

The single-timeframe market structure shift shortcut

There is a faster version of bias that skips the full five steps, and ICT traders share it as the daily bias trick. You read the most recent market structure shift on the daily chart, lock your bias in that direction, and then wait for a lower-timeframe shift to confirm the entry.

The shortcut works because a daily structure shift is a strong statement of intent, and trading in the direction of the most recent one keeps you aligned with the force that last moved the market.

I use the trick on slow days when I have no time for the full process, but I treat it as a fallback rather than the main method. It reads only one timeframe, so it misses the weekly context that protects the full five-step version from regime errors.

When the weekly and daily bias disagree

This is the question traders ask more than any other, and almost no page on the SERP answers it with a rule. The weekly says one thing and the daily says another, and the temptation is to abandon the process and guess.

The fix is to treat the conflict as a known case with a written response, not a crisis. The weekly sets the regime, the daily sets the target, and when they clash I size my conviction by which way each frame points.

The rule I keep is that the higher timeframe wins the regime. When the daily disagrees, I do not flip my bias, I lower my conviction and demand a better entry inside the daily's premium or discount before I act.

Bullish versus bearish daily bias

A bullish daily bias means the session is read to deliver upward, toward buy-side liquidity above old highs, and a bearish bias means the session is read to deliver downward, toward sell-side liquidity below old lows. The two are mirrors, and the checklists for each are symmetric.

I score the row. Four checks aligned is a high-conviction bias, three is tradeable with care, and two or fewer is a day to stand aside, because a weak bias is worse than none.

The neutral day: when to sit out

Not every session produces a clean read, and the disciplined response to an unclear bias is to trade nothing. ICT traders treat a no-bias day as a neutral day, and the iron rule is that no bias means no trade.

I mark a day neutral when the weekly is flat, the daily is chopping inside its range, and no clean structure shift has printed. On those days the method offers no edge, and forcing a trade manufactures risk where none needs to exist.

The hardest part of bias work is not the analysis but the patience to sit on your hands. The traders who last are the ones who can take a neutral day without feeling they missed something, because the next high-conviction day always comes.

A worked pre-session walkthrough on NAS100

The framework only clicks when you see it run on a real instrument, so here is a full walkthrough on NAS100, the index most ICT traders actually trade. I run this exact sequence on a Sunday evening before the London open.

Step one, the weekly read. The weekly candle is bullish and travelling toward an old high at 19,800, so the macro direction is up and my bias should lean long unless the daily disagrees.

Step two, the daily structure. Friday closed above the prior day high at 19,640 with a bullish market structure shift, which agrees with the weekly.

The close above PDH leans the next session bullish under the next day model.

Step three, the draw on liquidity. The nearest unswept pool is buy-side liquidity at the 19,800 old high, so the target for a long is explicit and the trade has a destination.

Step four, premium and discount. Price opened the session in the discount half of its dealing range, which is where the method wants to buy, so the location supports the long bias.

Step five, the scenario. I expect a NAS100 Judas swing down at the London open to sweep a session low, then a reversal up through the New York killzone toward 19,800, which gives me my entry time and my invalidation level below the sweep.

The bias is written: bullish, target 19,800, long only off a confirmed sweep. If price instead breaks and holds below Friday's low, the bias is invalidated and I stand aside, because the read was wrong and the plan is to follow the process, not defend the call.

Applying ICT daily bias to round-the-clock crypto markets

The method was built on forex and indices with clean session opens, and crypto trades around the clock with no London or New York bell. The bias framework still applies, but the time anchors have to move.

I swap the killzones for fixed references on crypto. The midnight UTC open replaces the daily candle open, and the eight-hour cycle replaces the session windows, so the same top-down read runs on a clock that fits a 24-hour market.

The structural read carries over cleanly. Weekly and daily structure, the draw on liquidity, and premium and discount work the same way on a Bitcoin perpetual as on EURUSD, which is why the concept maps onto perpetual futures trading without forcing a new method.

When to abandon your read

A bias is a thesis, not a vow, and the method defines exactly when the thesis dies. Hard invalidation is a sweep of the opposite pool followed by a structure shift against your bias, and that combination means the read was wrong and you stop trading the direction.

Soft invalidation is different. Price chopping at equilibrium with no follow-through is not proof the bias is wrong, it is a sign the move has not engaged, and the response is to wait rather than to flip.

I write my invalidation down before the session opens, alongside the bias itself. Deciding the exit point in advance is what stops me from arguing with the market in real time, which is the trap that turns a small loss into a large one.

Common daily bias mistakes

The same errors sink most bias work, and they are all forms of breaking the process. The first is reading bias from a lower timeframe, which captures intraday noise instead of the delivery direction.

The second is ignoring the draw on liquidity. A bias without a named target is a feeling, and the method wants a level price is travelling to, not a vague direction.

The third is forcing a bias when the daily is unclear. A weak read traded with full size is the most expensive mistake in the sequence, and standing aside is the correct call when conviction is thin.

The fourth is switching bias mid-session. The plan is to set the call before the open and trade it, and flipping because a counter-move looks scary is how traders take the worst entry at the worst time.

The fifth is confusing bias with the current move. Price can spend hours moving against the bias to load a pool, and reading that counter-move as a new bias is how traders abandon the correct read at the worst moment.

The honest framing: what is proven and what is doctrine

I want to be straight about the evidence, because the ICT SERP is thick with invented statistics. The specific win rates you will see quoted for daily bias, the 79 percent and the 75 to 85 percent figures, are self-reported by educators with no published method and no peer review, and I will not repeat them as fact.

The external anchor that does hold up is Carol Osler's research at the Federal Reserve Bank of New York, which found that exchange rates tend to stop trending at support and resistance levels and that stop orders cluster just beyond prior highs and lows. That is the real, documented reason the prior day high and low behave like magnets.

What is not proven is the ICT claim that a central algorithm delivers price to those levels on purpose. That is a plausible story that fits the observation, but it is an interpretation rather than a documented mechanism, and no peer-reviewed study validates the five-step framework or its profitability.

I trade the process because the levels are real and the structure is observable, not because the institutional narrative is confirmed. The honest statement is that bias is a disciplined read of real levels, tested on your own data, and anyone quoting a guaranteed win rate is selling something rather than reporting a study.

The same honest split runs through everything in Smart Money Concepts on the site.