RXT

Education

Prop Firms

Learn More

Weekly Insights

How I build a daily bias before the New York open

I build a daily bias the same way every session. Daily first, then 7-hour, then 4-hour. I lock it before the New York open and I do not change it after 10:00 ET. The bias is a direction, not a prediction. It tells me which side of the market I am allowed to take. If the condition for entry never shows up, I take nothing. That is a normal day.

This post is the exact top-down routine I run on NQ and MNQ before New York. I will walk it with the 4 August 2026 session so you can see the levels, not just the theory.

What the daily bias is for

The daily bias answers one question: which way am I allowed to trade today. Long only, short only, or flat. It is not a forecast of the close. It is a filter.

I profile bias top down on three timeframes:

  • Daily

  • 7-hour

  • 4-hour

Each timeframe has a job. The daily sets the directional lean from structure. The 7-hour bridges the session boundary and shows how overnight range is behaving relative to the prior day. The 4-hour is the last alignment check before I lock the plan. I do this work before the New York open. Entries only happen inside the 10:00–14:00 ET window. Bias is locked at 10:00 ET and is not revised mid-session.

The instrument is NQ or MNQ futures. The session boundary I use is 18:00 ET.

Step 1 — Daily timeframe first

I start on the daily chart. I mark the previous day's high, low, and close. I mark the relevant swing points above and below price. I am looking for a clear draw on liquidity.

A draw on liquidity is a level price is likely to run toward because resting orders sit beyond it. Prior day highs and lows, prior swing highs and lows, and session extremes are the references I use. I pick one primary draw in the direction of the structure I am leaning into. I also mark the level that would invalidate that lean.

Invalidation is simple. If price takes out the structural point I am basing the bias on, and holds beyond it, the bias is wrong. I do not invent a new bias at that point during the New York window. I stand down.

On 4 August 2026, the previous day on NQ printed:

  • High: 28964.25

  • Low: 28313.50

  • Close: 28891.75

Bias resolved long off the daily swing low at 27201.50. That swing low was the structural reference. As long as that framework held, I was only interested in long setups. I locked the bias at 10:02 ET.

I do not need the daily to be fancy. I need a direction, a draw, and a level that tells me I am wrong.

Step 2 — 7-hour timeframe

The 7-hour sits between the daily and the 4-hour. Its job is to show how the overnight sessions have built range relative to the prior day and relative to each other. I read Asia and London as ranges, not as stories.

On 4 August 2026:

  • Asia session range: 28831.75 to 29098.75

  • London session range: 29007.50 to 29245.75

London never traded back into Asia's range. That matters. It told me London was accepting higher prices and not returning to the prior overnight block. With a long bias already forming from the daily swing low, that separation supported the lean. It did not create the bias on its own. It confirmed that overnight flow was not fighting the daily direction.

I also mark internal references inside London. A 25% retracement of London's range sat near 29067. The London low at 29007.50 was the deeper reference. Those numbers become candidates for manipulation later. They are not entries. They are levels I write down so I know in advance what has to be swept.

Step 3 — 4-hour alignment

The 4-hour is the final alignment check. I want to see that the 4-hour structure is not in open conflict with the daily lean. If the daily is long and the 4-hour is mid-range with no opposing impulse that has already broken the structure I care about, I treat it as aligned. If the 4-hour has already violated the basis for the daily lean, I do not force the bias.

I am not hunting entries on the 4-hour. Entries come later from an hourly manipulation and a 5-minute confirmation. The 4-hour only answers whether I still have permission to hold the direction I built from the daily and the 7-hour.

When the three timeframes agree, I write the plan in one pass:

  • Bias: long or short

  • Primary draw on liquidity

  • Invalidation level

  • Reference levels that qualify for a sweep (prior day high/low, Asia high/low, London high/low, prior swing points)

  • No trade if manipulation never prints

Then I stop editing the bias.

Why the bias locks at 10:00 ET

Bias is locked at 10:00 ET. Mine locked at 10:02 ET on 4 August. After that I do not flip direction because a candle looks strong or weak.

There is a reason. If I allow mid-session revision, the bias stops being a filter and becomes a feeling. Feelings chase the last hour. The model is built on repetition and defined risk, not on updating a story every time price moves.

The entry window is 10:00–14:00 ET. Outside that window I do not enter. Inside that window I only take setups that agree with the locked bias. If bias is long, I ignore short manipulations. If the long manipulation never happens, I take no trade. That outcome is normal. It is not a failed day. It is the model working as designed.

Risk controls sit on top of this. The day stops after +2R, or after 2 losses. Those rules only make sense if the direction was fixed before the first order.

What a draw on liquidity is in practice

A draw on liquidity is not a vague magnet. It is a specific level I can point to on the chart.

I pick draws from:

  • Previous day high and low

  • Asia high and low

  • London high and low

  • Prior swing highs and lows on the daily and 4-hour

On 4 August, with bias long, I was interested in price continuing to work higher after any deeper reset into overnight references. The London low at 29007.50 was the deeper reference. The 25% retracement near 29067 was the shallower internal level. I write both down. I do not need price to hit every reference. I need to know which levels count if a sweep prints.

The sweep itself is not the topic of this post in full, but you need the definition to understand what the bias is waiting for. A valid setup needs an hourly manipulation: price sweeps a prior reference low or high by at least one tick, then closes back through it. Then a 5-minute confirmation triggers the entry. Stop goes at the 5-minute manipulation extreme, one tick beyond. Target is a static 2R.

No hourly manipulation means no trade. The bias can be correct and still produce no entry. That is expected.

Worked example — 4 August 2026

Here is how the full morning read sat together.

Previous day: high 28964.25, low 28313.50, close 28891.75. Daily lean resolved long off the swing low at 27201.50. Asia ranged 28831.75 to 29098.75. London ranged 29007.50 to 29245.75 and never traded back into Asia. Bias locked long at 10:02 ET.

Reference levels written down before the window:

  • London low 29007.50

  • 25% of London near 29067

  • Prior session and hourly highs/lows as they formed into the open

What happened into the window matters for process, not for hindsight. On MNQ the 08:00 candle low was 29186.50. The 09:00 candle traded to 29185.00 — a sweep of 1.5 points — then closed at 29469. On NQ the same 09:00 candle low was 29191.25 against an 08:00 low of 29188.00. It missed by 3.25 points.

Same moment. Two instruments. Opposite readings on whether a sweep had printed. That is why the reference has to be defined in advance and written down on the contract you actually trade. I do not switch charts mid-setup to force a qualification.

The session high reached 29504.50. The day finished with no entry taken. Bias was long. Price ran. I still took nothing because the model did not give a valid entry under the rules I had written. That is an acceptable outcome. The job of the bias is to constrain direction. The job of the entry rules is to constrain when risk goes on. Both have to agree.

What invalidates the bias

Invalidation is decided before the open, not during the move.

For a long bias, invalidation is a break and hold beyond the structural low I used to justify the lean. On 4 August that framework was the daily swing low at 27201.50. I am not waiting for that level to trade in the New York window on most days. It is the structural anchor. Closer operational invalidation comes from the overnight references I am using for manipulation. If those levels are destroyed in a way that breaks the long thesis and removes any clean reference for a long sweep, I simply have no setup. I do not flip short because I am bored.

For a short bias the mirror applies. I define the structural high, the draw below, and the references that must be swept lower then reclaimed.

I do not revise bias after 10:00 ET. If I am wrong, I pay the stop on a valid entry or I take no trade. I do not rebuild the narrative at 11:30 because the market looks different.

How the three timeframes divide the work

Keep the roles clean.

Daily. Direction and structural draw. Previous day high, low, close. Swing points. Invalidation at the structural level.

7-hour. Overnight range behaviour. Asia and London highs and lows. Whether London re-enters Asia. Internal retracement levels inside London that may become sweep references.

4-hour. Final alignment. Permission to keep the daily lean. Not an entry chart.

When you mix the roles, you start taking entries off higher-timeframe opinions or flipping bias off lower-timeframe noise. The model is Framework on the higher timeframes, then Confirmation, then Entry off candle signatures inside the window. Bias belongs to the framework step.

The week around it, in plain terms

The point of a locked bias and a narrow entry window is consistency of process. In the week of 28 July I took 3 trades and had 3 sessions with no entry. Results on the three trades were +2.0R, +1.67R, and +1.14R. Net +4.81R. Three trades is a small sample and proves nothing on its own. I include it only to show that no-trade days sit next to active days in normal operation. The bias routine is what makes both outcomes acceptable.

What to do tomorrow

Run this checklist before the New York open. Write the answers down. Do not keep them in your head.

  • Mark previous day high, low, and close on NQ or MNQ.

  • Mark the daily swing high and swing low that define current structure. State bias direction in one word: long, short, or flat.

  • Write the primary draw on liquidity and the structural invalidation level.

  • Record Asia range high and low. Record London range high and low. Note whether London has traded back into Asia.

  • Mark the 25% retracement of London's range and the London extreme that matters for your direction.

  • Check the 4-hour for alignment with the daily lean. If it conflicts with the structural basis, stand down.

  • List the exact reference levels that will count for an hourly sweep tomorrow inside 10:00–14:00 ET.

  • Lock the bias at 10:00 ET. Do not change it after.

  • If no hourly manipulation prints — sweep of a written reference by at least one tick, close back through it, then 5-minute confirmation — take no trade.

  • If you enter, stop goes one tick beyond the 5-minute manipulation extreme. Target is 2R. Stop the day at +2R or after 2 losses.

Repetition of this sequence is the work. The bias is a defined filter. It is not a prediction of what the session must do.