Value Area High & Low (VAH/VAL)

You know the fairest single price (M2.01). But no serious auction does all its business at one price — it does most of it in a zone. Where does that zone start and end, and what changes when price trades on the far side of its edges?

The value area (vùng giá trị) is the range of prices that holds roughly 70% of the session’s volume, built outward from the POC Definition. Its top edge is the Value Area High (biên trên vùng giá trị, VAH); its bottom edge is the Value Area Low (biên dưới vùng giá trị, VAL). Why 70%? It is a convention from Peter Steidlmayer’s Market Profile work at the Chicago Board of Trade in the 1980s. It is often explained as “about one standard deviation” (≈68% of a normal distribution), but that analogy is loose: real profiles are frequently skewed or double-humped, nothing like a bell curve. The percentage is a standard everyone agrees to use; the concept — where most business happened — is the substance.

Chart first: the zone on one profile

What to see: most of the session’s business sits in a band a few points wide around the POC, and the band’s two edges are the VAH and VAL. Everything outside the band traded thinly.

Two names for the two sides of an edge

Auction theory gives trade on each side of an edge a different name Definition. Inside the value area, trade is called responsive and two-sided: buyers and sellers are both active at prices the session has already accepted, and price rotates rather than travels. Outside the value area, price is probing for new value. The question is then whether the probe is accepted (volume starts building beyond the edge — value migrating) or rejected (a quick return inside — the probe failed).

These words describe what the auction has done so far; they do not forecast what it will do next. That is why VAH and VAL are worth marking: they are where the question “business as usual, or something changing?” gets asked. Whether price actually turns at them more often than at any other price is a separate, testable claim, and our lab tested it (below).

Value-area placement tells the day’s story at a glance. Today’s value overlapping yesterday’s = balance continuing. Today’s value sitting entirely above or below yesterday’s = value migrating — initiative in slow motion. In day-type language (M1.04): a Normal Day shows value central and overlapping; a Trend Day shows value migrating with the POC dragged toward the extreme. Both labels are given at the close.

Myth / hypothesis: “Value migrated higher today, so tomorrow goes up.” One practitioner’s public, pre-registered NQ study tested it: days after migration went in its direction no more often than other days (−0.4 pp [−3.0, +2.3], n = 1,376) Lore. Read that study with three caveats: it is one trader’s work, its intervals are not clustered by session, and it can support a claim at Plausible at most. Placement describes the auction so far; it does not forecast tomorrow.

The value area is descriptive context, never a signal. “Price is at VAH” is a location; what you need next is the reaction — acceptance beyond or rejection back inside. The edge is where the auction asks its question; the next 30–60 minutes give the answer House rule.

How the value area is actually computed

Your platform does this for you, but doing it once by hand removes the mystery. The classic CBOT method works in pairs of rows Definition:

1. Start at the POC. Its volume is your running total.
2. Look at the two rows directly above the current value area and add their volumes. Do the same for the two rows directly below.
3. Add whichever pair is larger to the value area.
4. Repeat from step 2 until the running total reaches 70% of the session’s volume. If one side runs out of rows, keep adding from the other.

Here is the profile from the chart, 1,000 contracts in total, so the target is 700:

Price   Volume
5,996     20
5,995     30
5,994     50
5,993     80
5,992    110
5,991    140
5,990    180   <- POC
5,989    130
5,988    100
5,987     70
5,986     50
5,985     40

• Start: POC 5,990 = 180.
• Round 1: above pair 5,991 + 5,992 = 250; below pair 5,989 + 5,988 = 230. Add above → 430.
• Round 2: above pair 5,993 + 5,994 = 130; below pair still 230. Add below → 660.
• Round 3: above pair 130; below pair 5,987 + 5,986 = 120. Add above → 790. That passes 700, so stop.

Value area = 5,988 to 5,994: VAL 5,988, VAH 5,994, holding 79% of volume. Notice the overshoot: adding two rows at a time can jump well past 70%. Many platforms therefore add one row at a time (compare the single next row above with the single next row below). Run that version on the same numbers and you get 5,988–5,993 with 74% — a VAH one point lower. Neither is wrong. They are different conventions, and that is why two platforms can print value areas a tick or a point apart.

Two more sources of disagreement. A TPO value area (M2.06) runs the same procedure on TPO counts — time — instead of contracts, so it measures where price spent time, not where volume traded; on most days the two are close, on fast news days they can differ a lot. And the session definition (ETH vs RTH) and row size change the inputs, exactly as they change the POC (M2.01). Pick one setup, write it down, and use it every day.

What the data says

The popular use of the value area is as a pair of levels: “the prior VAH is resistance, the prior VAL is support.” Our lab tested exactly that. It took the prior session’s VAH and VAL every day, waited for price to touch them, and asked how often price then turned away before pushing through. It asked the same of random prices at the same distance from the open.

Why it is still worth knowing

• Shared vocabulary. “Opened inside value”, “accepted above VAH” and “value migrated” are how profile traders, and later lessons, describe a session in one line.
• It describes the auction. Where value sits, and whether today’s value overlaps yesterday’s, tells you whether the market is still trading the same prices or has moved on.
• It frames risk. If your idea is “the probe above VAH will be accepted”, a fast return deep into value is a clear sign the idea was wrong — a natural place to define your stop.
• It is a hypothesis you can test. If you believe a particular kind of edge reaction works for you, log it with a written rule and compare it with random prices (EV1.04).

Worked example: the VAH that flipped sides

Illustrative ES numbers. Yesterday’s profile: POC 5,985, VAH 5,992, VAL 5,978 — 70% of volume inside. RTH opens at 5,990: inside value, upper third. The first hour rotates 5,986–5,993 — two-sided trade inside value, as the vocabulary describes. Nothing in the profile says which edge will be tested first, or whether either will turn price. Nothing to do but wait for an edge to be tested.

At 12:00 ET (00:00 ICT, winter) price pushes to 5,997 and stays there, with volume building between 5,995 and 6,000 over the next hour. Read: an initiative probe beyond VAH is being accepted — value is migrating higher. Yesterday’s VAH at 5,992 becomes the first level to watch on any pullback: a reference, not a floor. The trader who shorted “the top of value” at 5,992 without waiting for rejection just fought the migration — they faded a level instead of reading the auction. The sequence that respects the evidence: probe beyond VAH → watch for acceptance (volume building outside) vs rejection (fast return inside) → then decide. The auction gave a full hour of evidence; the impatient trader used none of it.

Worked example 2: when the value area hides the story

Illustrative NQ numbers. NQ prints a double-distribution day: a morning hump around 17,950 and an afternoon hump around 18,010, with thin trade between them. The platform’s value area runs 17,945–18,015 — wide, and it includes the thin middle at 17,975–17,990 where almost nothing traded.

Read: on a double distribution the single value area is a poor summary. Its VAH and VAL are real, but “inside value” is no longer one zone of agreement — it is two zones with an empty corridor between them. Price returning to 17,980 tomorrow is technically “inside value,” yet it is sitting in the thin part. In the same practitioner study, price crossed thin low-volume areas somewhat faster than other areas: a weak tendency PlausibleOne trader's study, covered in M2.03. The fix is to read the shape (M2.05) before you trust the value-area lines, and mark each hump’s own peak.

What the value area does not tell you

• That edges hold. In our lab’s test, prior VAH and VAL held about as often as random prices. Fading an edge looks right on balance days and wrong on migration days, and you only learn which day it was afterwards.
• Why 70% is special. It isn’t. It is a shared convention, which is useful precisely because many traders watch the same lines.
• Exact prices across platforms. Pair vs single-row expansion, ETH vs RTH, volume vs TPO — each moves the edges a little.

Common mistake: Fading VAH/VAL automatically — “it’s at the top of value, so it must come back.” That treats a line everyone watches as a line that works. In our lab’s test, prior VAH and VAL held no more often than random prices at the same distance, so a blind fade has no measured edge, and it gets run over on every migration day. If you trade at an edge, wait for the auction’s answer (a fast return inside, or volume building outside), write your rule down, and score it as your own hypothesis.

Recap:

• The value area holds ~70% of the session’s volume — a Steidlmayer/CBOT convention, only loosely linked to one standard deviation; VAH and VAL are its top and bottom edges.
• It is computed outward from the POC, adding the larger neighbouring pair (or single row, depending on platform) until ~70% is reached; TPO value areas use time instead of volume.
• Inside value, trade is described as responsive and two-sided; outside value, price is probing, and the probe is either accepted (volume builds beyond) or rejected (fast return inside).
• Value-area placement across days describes the market: overlapping value = balance continuing; non-overlapping value = value migrating. It does not forecast the next day.
• Tested as reaction levels, the prior VAH and VAL held about as often as random prices (measured, no edge, Nasdaq-index proxy): mark them as places to ask the question, not as support and resistance.

Further reading:

• Volume Profile Explained: How to Find Where the Market Actually Traded (Bars, FRVP & Anchored) — how VAH and VAL are drawn on live charts; treat any “support and resistance” claims in it as hypotheses.
• Volume Profile (Trading Strategy) Webinar — read order: POC, value area, then nodes, as one repeatable process.
• Our lab: the prior session’s VAH, VAL and POC compared with random prices at the same distance from the open, on 721 sessions of QQQ and SPY with a small NQ/ES futures check — no edge as reaction levels.

Prerequisite: M2.01 Point of Control · Next: M2.03 High & Low Volume Nodes · Connects to: M2.05 (shapes, including double distributions), M2.07 (IB extension beyond value), M2.10 (value migration across days), M1.04 (day types in profile language)

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