Multi-session structure

Friday 2 October 2026, 09:30 ET (20:30 ICT). NQ opens at 31,147.75, more than 300 points above the value it had built over the previous seven sessions. One day at a time, each of those sessions looked like nothing much: a rotation, a dull close. Side by side they were one auction that had agreed on a price range for a week and a half. The question this lesson answers: when value leaves a range like that, how do you read what happens next without pretending you know the ending?

A multi-day balance (cân bằng đa phiên) is a run of sessions whose value areas overlap, so the market keeps agreeing on roughly the same prices Definition. A value migration (dịch chuyển giá trị) is the opposite: each session’s value sits at new prices, away from the last. The tools are the ones you already have (M2.02 value area, M2.04 composite); only the timeframe changes.

Chart first: one auction across sixteen sessions

How it works

Overlap or migration: compare the value areas

Put the last five to eight daily value areas side by side and ask one question: do they overlap? If most of them share prices, the auction is in multi-day balance. If each one sits beyond the last, value is migrating. In the figure, 23 September to 1 October overlaps heavily (every daily value area falls inside 30,500–30,935) Definition. A composite of those sessions (M2.04) summarises the balance in one profile: value 30,550–30,825, POC 30,632.5.

Inside a multi-day balance the vocabulary of M1.02 applies at a larger scale: activity near the edges is responsive, a push beyond an edge is initiative Definition. That vocabulary describes the auction. It is not a tested rule that fading the edges or buying the break pays: in our lab’s test, the prior session’s value-area edges held no more often than random prices at the same distance Measured: no edgeNasdaq index (M2.02).

A break has three stages

A break out of a multi-day balance is not one event. Read it in three stages:

1. The claim. A session’s value prints entirely outside the balance. On 2 October value was 31,005–31,170, above the highest daily value-area high of the balance (30,935). One side acted with initiative.
2. The test. Price comes back toward the old edge. This is where the auction checks whether the new prices are accepted. On 5 October the opening dip stopped at 31,044.25, above the old range, and value built higher (daily POCs 31,227.5, then 31,532.5).
3. The verdict, for now. Value that keeps building outside means the break is accepted; value that rebuilds inside means it failed. Both are descriptions after the fact, and both can change. NQ held outside for three sessions, then traded back to 30,792 on 8 October, built value inside the old range on 12–13 October and printed value below it on 14 October.

The third stage is the part most traders skip. “Accepted” is a description of the sessions so far, not a promise about the next one.

The POC trail

Plot each session’s POC as one point and join them. A flat, tangled trail is balance; a trail that steps one way is migration Definition. The October trail reads 30,632.5 (1 Oct) → 31,067.5 → 31,227.5 → 31,532.5 → 31,352.5 → 31,237.5 → 31,092.5 → 30,892.5 → 30,772.5 → 30,172.5 (14 Oct): up for three sessions, then down through the old balance. It is the fastest way to answer “balancing or migrating?” before the open. It answers it about the past.

Myth / hypothesis: “Value migrating higher today means the next session goes up.” A public practitioner study of NQ tested it: sessions after higher, overlapping or lower value went in that direction no more often than other sessions (−0.4 pp [−3.0, +2.3], n = 1,376) Lore. Read the trail as a description, not as tomorrow’s direction.

Gaps and ragged extremes across days

An overnight gap that opens beyond prior value and stays there is initiative across sessions; a gap that trades back into prior value is responsive Definition. Whether either one predicts the rest of the day is a separate question (M1.06).

Myth / hypothesis: “A ragged multi-day high, with no excess, has to be revisited before the larger move can reverse.” The single-session version (poor highs and lows get revisited) was inconclusive in the same NQ study (+1.0 pp [−2.8, +4.8], n = 3,766) Lore, and the multi-day version is not yet measured. Keep it on the watch list, not in the plan as a target (M2.09).

What the data says

Why it is still worth knowing

Multi-session structure is shared vocabulary: “we are inside last week’s balance” tells another trader a lot in one sentence. It frames risk, because the old balance edge is the natural place to say “if value comes back inside here, my reason for being long is gone”. And it turns the breakout claim into a hypothesis you can score yourself: write the balance, the break and your acceptance rule down before the session, then record what happened.

Worked example: the long side, 2 to 7 October 2026

Historical NQ prices; the framing is an illustration, not a call.

1. Evening of 1 October: mark the balance. Seven sessions, every value area inside 30,500–30,935; composite value 30,550–30,825, POC 30,632.5. Old top of value: 30,935.
2. 2 October: the claim. Value prints 31,005–31,170, entirely above. You write the conditional plan for the next session: “If pullbacks stay above about 30,935 and value builds above 31,005, the break is being accepted. If value builds back inside 30,550–30,935, the break failed.”
3. 5 October, 09:30 ET (20:30 ICT): the test. The open dips to 31,044.25 and holds above the old range; by the close value has built higher, POC 31,227.5. A long taken on that test has a written invalidation (value back inside the old range) and a stop beyond the session low, not a target from the old range.
4. 6–7 October: still outside. POCs 31,532.5 and 31,352.5. The break is accepted so far, and the POC trail is rising.

The mirror and the failure

The failure, 8 to 14 October. On 8 October NQ traded down to 30,792 at 13:27 ET (00:27 ICT), inside the old balance, and closed at 30,976.25. On 12–13 October value sat back in the old range (composite 30,755–31,035). On 14 October value printed below it, at 30,090–30,485. The long from step 3 met its written invalidation days before that. The trader who held “because the break was accepted” was holding a description that had stopped being true. One example cannot tell you how often this happens; the study above says the breakout rule has not shown an edge.

The short mirror, 30 October to 6 November. Four sessions balanced with composite value 28,245–28,710 (daily value areas inside 28,290–28,890). On 5 November value printed at 26,830–27,365, entirely below, on a session that closed at 26,745.75. On 6 November the high was 27,425, nowhere near the old bottom, and value stayed below (27,105–27,310). As of the last recorded session the break was accepted; the same “for now” applies. The short-side invalidation would have been value rebuilding inside 28,290–28,890.

Common mistake: Treating the breakout day as the verdict: “value migrated, new trend, I’m all in.” A break is a claim, the test is where the auction answers, and even an accepted break is only accepted so far. Write the invalidation (value back inside the old range) before you enter, and act on it when it prints.

Common mistake: Turning the open lead into a rule. “+30 points in the holdout” is one year of one practitioner’s study, after a development result that showed nothing clear. Treating it as an edge is treating a hypothesis as a base rate. It belongs in your forward-test log (M5.07), not in your size.

Common mistake: Expecting a failed break to run to the composite POC or the opposite edge because “the trapped buyers have to get out”. Trapped inventory explains why a failed break can move fast (M1.07); the same study found failed breaks of prior-day extremes reached the other side no more often than baseline Lore. Price did fall through the old balance in October; in another month it may stall in the middle.

Drill: build a POC trail

Use the last ten RTH sessions of NQ on your platform.

1. Write each session’s POC, VAH and VAL in a table.
2. Mark each session as “overlaps the previous value area” or “outside it”.
3. Draw the POC trail and label each stretch: balance or migration.
4. If a session’s value printed entirely outside a balance of three or more sessions, write the next session’s acceptance and invalidation rule before you look at it, then record what happened.

Do it on the platform

Recap:

• Several daily profiles are one auction: overlapping value areas describe a multi-day balance, value stepping away describes migration.
• A break out of a multi-day balance has three stages: value prints outside (the claim), the old edge is tested, and value is accepted outside or comes back inside. “Accepted” means so far.
• The POC trail shows migration directly and describes the past; it did not forecast the next session in the NQ study.
• The breakout-continuation rule is Lore: inconclusive in development, with an open lead from one holdout year that is still being forward-tested.

Further reading:

• Mind Over Markets — James Dalton — auction-market theory and balance across timeframes, from the people who named it; read its rules as hypotheses.
• Linnsoft FT71 resources — profile tooling and auction-market education for multi-day analysis.
• Order Flow, Measured: All claims. A public practitioner study on NQ; rows AP-03 (multi-day balance breakout), AP-02 (value migration and the next day), AP-04 (failed breaks) and AP-05 (poor highs and lows).
• Our lab: daily and composite RTH value areas for NQ, 23 September to 14 October and 30 October to 6 November 2026, from our own tick recording.

Prerequisite: M2.04 Composite profiles, M2.09 Poor highs & lows · Next: M2.11 Overnight inventory, the same auction read across one night instead of many days · Connects to: M1.02 (balance and imbalance), M1.07 (trapped inventory), M5.07 (the breakout–retest setup and its forward test)

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