Daily, weekly & monthly level map

08:00 ET (20:00 ICT, winter). ES traded overnight between 6,008 and 6,018. Yesterday’s high was 6,020, last week’s heaviest-traded price was 6,022, and this month’s high sits at 6,045. Your chart carries twenty lines from past sessions. Which deserve your attention at 09:30 ET (21:30 ICT), and what will you do if price reaches one? This lesson builds the answer as a short written map: a few places to watch, a reason for each, an if-then plan for both sides, and a check on whether such levels do anything random prices don’t.

Key terms

Level map (bản đồ mức giá): a short list of prices, written before the open, where you will watch how the auction behaves, each with a reason and an expected behavior Definition. A level on the map is a place to look, not a forecast of where price turns.

Confluence (đồng thuận mức giá): two or more levels from different layers sitting within a few ticks of each other, merged into one zone Definition.

Term you need: composite profile. A volume profile built over several sessions instead of one (M2.10). Its POC is the price where most volume traded that week or month; its value area holds about 70% of that volume. The weekly and monthly layers of the map come from it.

Chart first: where the weekly layer comes from

How it works

The three layers

• Daily: prior day high, low and close; overnight high and low; today’s initial balance (after 09:30 ET); VWAP; the developing POC. The day’s own structure.
• Weekly: week high and low, the weekly POC, last week’s value area edges. What the week’s participants traded.
• Monthly: month high and low, the monthly POC, round numbers such as 6,000 on ES. The big-picture boundaries.

Each layer answers the same question at a different scale: where did business happen, and where did it stop? Every line on the map is a description of past trade Definition. Whether price reacts at it today is a separate question, and the answer comes from order flow at the time (M4.09), not from the line.

Label the kind: the label picks the if-then you write

The kind of level tells you which two scenarios to write and what you need to see before acting. It does not tell you which scenario is more likely.

• Prior extreme (a high that was rejected before): write a rejection plan and an acceptance plan. An earlier rejection does not make the next one more likely: our lab found prior-day highs and lows held after a touch no more often than random prices Measured: no edgeNasdaq index.
• Accepted level (yesterday’s high, traded through and built value above): watch whether value keeps building on the new side. “Old resistance becomes support” is a hypothesis to count, not a rule Lore.
• Unfinished business (poor high, single prints, M4.06): a reference to watch. The claim that price gets drawn back to it was tested and showed no edge (myth box below).
• IB extreme: the line where acceptance tests happen during the day (M4.05) Definition.
• VWAP: a context line for the day’s average price, not an entry trigger (M4.04) Context.

Cap the map at 5–8 levels

Keep 5–8 levels House rule. The reason is arithmetic: mark 25 lines within a day’s range and price touches several of them every session by chance alone, so every move looks “explained” and nothing can be scored. Six lines you can plan, watch and grade. If your chart looks like sheet music, delete until the map fits on the morning sheet (M4.08).

Confluence: fewer lines, not a stronger level

When the overnight high, the prior high and the weekly POC sit within four points of each other, merge them into one zone: one place to watch and one pair of if-then plans House rule. Confluence is not a multiplier. Our lab compared zones where two or more level families stack with single levels and found no difference in how often they held Measured: no edgeNasdaq index. A stacked zone gets the same size, stop logic and need for order-flow confirmation as a lone level.

Update intraday, and check the higher layers

As the day prints new extremes and accepts new prices, add them to the map rather than redrawing it; after the close it becomes tomorrow’s history layer. Check the higher layers even on scalps: a daily long just under the weekly high is a long into a place you chose to watch, so write down what flow you need to see there first.

Myth / hypothesis: “Stacked levels are stronger, round numbers attract price, and unfinished business pulls price back” Lore. Our lab’s stacked zones held no better than single levels. One practitioner’s NQ study found round 00/50 handles did not reverse on the first touch (−3.6 pp [−6.6, −0.6], n = 2,031, claim RL-03), and in our lab the top round-number tier was reached and held slightly less often than random prices. Poor highs and lows came back about as often as extremes with tails (AP-05, inconclusive); single prints did not fill or hold more than chance (AP-07, not supported). The Halfback, the midpoint of the prior day’s range, is popular and not yet measured by our lab Lore. The practitioner study is one trader’s pre-registered test; its intervals are not clustered by session (too narrow), so on its own it caps at Plausible (Order Flow, Measured).

Putting it together: the value map behind the map

Every level on this map comes from somewhere, and that somewhere is the value map from M2. It has five layers: (1) composite structure (where longer-term value sits), (2) prior session (POC, VAH, VAL, profile shape, poor or excess extremes), (3) overnight (inventory, midpoint, unfinished business), (4) developing (today’s POC migration, IB, shelves), and (5) the places where the layers agree. Where layers agree, you get one zone instead of several lines.

Turn it into the daily value narrative, one paragraph every morning: “Value migrated higher for three sessions; yesterday printed a P-shape with an excess low at 5,960; overnight was light, midpoint 5,988; we open at 5,992 inside yesterday’s value, upper third. First question: does price accept above the overnight high at 5,998, or rotate back toward the POC at 5,985?” That paragraph is the skeleton the map hangs on.

One rule keeps the map honest: every level needs a reason (which layer put it there) and an expected behavior (hold or break, accept or reject) with the evidence that would confirm each. A level without an expected behavior is decoration. Score the map each evening against random prices at the same distance (the drill below).

What the data says

So keep the map, but read it differently. It tells you where to watch; it does not do the work. Any edge has to come from what you see at the line and from your execution, and you measure that in your journal.

Worked example: the 6,018–6,022 zone, long side

Illustrative ES day (synthetic numbers; ES ticks are 0.25, $12.50 a tick). Morning map, 8 levels: prior high 6,020 / prior low 5,995 / overnight high 6,018 / overnight low 6,008 / IB (forms 09:30–10:30 ET) / VWAP / weekly POC 6,022 / month high 6,045.

Confluence scan: 6,018–6,022 holds the overnight high, the prior high and the weekly POC. Merged into one zone, it is the place you watch most closely because the day’s two main scenarios sit on either side of it. The lone 5,995 prior low gets the same treatment.

Plan A, written at 08:00 ET: “If price trades above 6,022 and is accepted there (M4.09: at least 15 minutes above, volume building above the zone, a 5-minute close above 6,023), then long, first target 6,035 (M5.07). Invalid back below 6,018.” In numbers: entry 6,023.75 after the confirming close, stop 6,017.75 (one tick under the zone), risk 6.00 points = 24 ticks = $300 per contract. Target 6,035 is 11.25 points, $562.50. You size from the $300 stop, not from how many layers stack at the zone.

On this day plan A never triggered. At 10:40 ET (22:40 ICT) price spiked to 6,021, printed a long upper wick and fell back under 6,018 within minutes. Acceptance above 6,022 never happened, so there was no long. The map didn’t predict the wick. It told you in advance which evidence would make you act and which would keep you out.

The mirror and the failure

Mirror (short side). Plan B was written next to plan A: “If a probe into 6,018–6,022 fails and price closes back below 6,018 with sellers in control, then short toward VWAP and 6,008; invalid above 6,023.25.” After the 10:40 wick, a 5-minute close at 6,017.25 gives entry 6,017.25, stop 6,023.25 (6.00 points = $300), target 6,008 (9.25 points). The fade was pre-authorized, so you didn’t have to improvise.

Failure. Another morning, same map. At 09:50 ET price runs through 6,018–6,022 without a pause, never trades back into the zone, and the day closes near 6,040. The “most important zone” gave nothing to trade on either side, and the lone 5,995 low never got a visit. The map didn’t fail: on many days nothing tradable happens at the lines, and a random price at the same distance would have been crossed the same way, as the evidence box says.

Score your map

Prop-firm note. The stop you pre-compute on the map ($300 a contract in the example) also has to fit the firm’s daily loss limit and position caps; the PROP course covers those rules.

Common mistakes

Common mistake: The 25-level chart: every swing, every Fibonacci retracement, every indicator line, until the map explains everything and predicts nothing. A reaction at one of 25 lines is coincidence. Choosing which 5–8 places you will watch is the preparation.

Common mistake: Sizing up at a stacked zone because “three timeframes agree”. That treats a description as an edge. Our lab found stacked zones held no better than single levels. Size from your stop and your risk budget, the same at every level.

Common mistake: “It reacted, so the level works.” One reaction is one touch. Random prices at the same distance also react about as often. Only a count against a placebo, over many sessions, tells you anything.

Drill: score your map against random prices

For 20 sessions, write your 5–8 level map before 09:30 ET and score it after the close.

1. For every level, write its distance from the 09:30 ET open in points. Put a placebo at the same distance on the other side of the open (open 6,012, level 6,020 → placebo 6,004).
2. After the close, mark each level and each placebo: touched (yes/no). For each touch, mark held: price moved at least 4 ES points (16 NQ points) away on the side it came from before trading 2 ES points (8 NQ points) through it. Write the rule down before session 1 and don’t change it.
3. Count: held ÷ touched for your levels, and held ÷ touched for the placebos.
4. Write one sentence comparing the two rates, then one sentence on how wide the uncertainty is.

What we don’t know yet

• Whether order-flow confirmation at a zone beats acting at the zone without it: not yet measured (a forward test is on the lab’s list; M4.09 writes the rule so it can be scored).
• Monthly highs, lows and POCs as levels: not yet measured by our lab.
• Prior-week extremes and weekly or monthly opens: inconclusive so far (our lab +2.7 pp [−1.6, +7.3]; the practitioner study +3.4 pp [−2.2, +8.9], n = 649, claim RL-05). More sessions are needed before either side can be claimed.

Recap

• A level map is a short written list of places to watch, built from three layers (daily, weekly, monthly), each level with a reason and an expected behavior.
• Cap it at 5–8 levels and label each by kind; the label tells you which two if-then plans to write, not which one wins.
• Merge levels that sit together into one zone; our lab found stacked zones held no better than single levels, so size and confirmation stay the same.
• Our lab measured map levels holding at the random-price rate; act on the order flow you see at the zone (M4.09), and score your map against random prices.

Further reading

• How To Trade A Market Profile Ledge: context-checked levels; what makes a level worth watching.
• Auction Market Theory: Understanding of market and orderflow dynamics: acceptance and rejection at levels in AMT terms.
• Trading Key Market Auction Reversals: multi-timeframe levels and initiative flips, read as hypotheses to score.
• Order Flow, Measured: one practitioner’s public, pre-registered NQ study; the round-number and weekly-open rows (RL-03, RL-05).
• Our lab: pre-market level families vs a distance-matched placebo, 721 sessions and 226 tests. No family, and no stacked zone, held more often than random prices at the same distance.

This lesson connects to: M2 (the value map behind the levels) · M4.01 (acceptance: how levels get tested) · M4.08 (the morning sheet, where the map gets used) · M4.09 (order-flow confirmation at a zone: the evidence you act on) · M5 (every setup is located on this map).

Educational content only — not financial advice. Futures trading involves substantial risk of loss.