VWAP as a reference

VWAP (giá bình quân theo khối lượng) = Σ(price × volume) ÷ Σ(volume) from the session start Definition. It is the average price paid by every buyer, and received by every seller, since that start. You learned the mechanics and the bands in M3.07; this lesson is the applied version: how VWAP earns its place on the morning sheet and in a written trade plan.

Chart first: losing VWAP, then taking it back

Why the line is on everyone’s chart

VWAP is the standard execution benchmark: a fund buying through the day is often judged by whether its average fill beat the day’s VWAP Definition. A VWAP algorithm slices a large order and trades it roughly in proportion to the market’s usual volume, so its average fill lands near VWAP. It tracks volume over time; it does not stand at the VWAP price buying every dip to it.

Myth / hypothesis: “Institutions buy below VWAP and sell above it to defend the line, so it becomes self-reinforcing: price reacts there because everyone expects it to” Lore. A tidy story, but our lab has no test of a plain VWAP touch: not yet measured. The two related claims tested on NQ, band reversion and anchored-VWAP support/resistance, were not supported (M3.07).

What VWAP does give you for certain is a shared reference: the number other traders, algorithms and your own journal talk about, and a precise line against which a read is right or wrong.

Above/below as a description of control. Price holding above VWAP means the average buyer since the open is ahead and the auction has been accepting higher prices Context; below, the reverse. That is a fact about the session so far, not a forecast. In the public NQ data, trend days do hold one side of VWAP after 10:30 ET PlausibleOne trader's study, but “trend day” is a label given at the close, so that result describes trend days; it does not tell you this morning is one (evidence box below). Use the side as a filter for which plans you write, not as an entry. In balance, VWAP is the middle of the rotation, and trading every cross mostly pays commissions.

VWAP bands (±1σ, ±2σ) measure how stretched price is relative to the session so far, not where it turns. Band reversion was tested on NQ and touches reverted slightly less often than the baseline Lore (numbers in M3.07); ±3σ is not yet measured. On a one-sided day (M4.03) price can ride the +2σ band for hours. Treat a band touch as context for a planned trade, not a fade signal.

The reclaim: a control change worth planning for

Price loses VWAP, then reclaims it with acceptance: time, closes and volume back on the other side (M4.01). The reclaim is an observable event, not a prediction. After it, the average buyer since the open is back near breakeven or ahead, and the session’s description has changed Definition. That is reason enough to have a plan ready for it.

Myth / hypothesis: “The reclaim is one of the most reliable intraday events: traders positioned against VWAP are trapped, and their exits fuel the move” Lore. The trapped-trader story is plausible as a mechanism, but no one we know of has tested VWAP reclaims against a baseline: not yet measured. A reclaim earns a written plan with a cheap stop, not a size-up.

Anchored VWAP (AVWAP) (VWAP neo) measures from an event instead of the open: the overnight high, a news release, the week’s open Definition. On a news day, “the average price since the repricing” is often the more relevant question. Swing anchors as support/resistance were tested on NQ and not supported Lore; release anchors are not yet measured. Know your plot: RTH VWAP starts at 09:30 ET (21:30 ICT, winter); a 24-hour VWAP starts at the 18:00 ET Globex open and averages in the thin overnight trade. They answer different questions and can sit many points apart.

What the data says

Worked example: the reclaim at 6,012

Synthetic session, not a real date (the figure above). ES, 5-minute chart, tick 0.25 points, $50 per point per contract.

Price leaves VWAP in the first 15 minutes and stays below it all morning: the session so far is the sellers’, so the plans you wrote were short-side plans. At 11:30 ET (23:30 ICT, winter) price rallies from 6,005 through VWAP at about 6,012 and prints three 5-minute closes holding above it with rising delta. By the M4.01 checks that is acceptance back above: the description of the day has changed, so the short plans are off.

The long plan you wrote in advance for this case (the M5.08 pullback):

• If price retests 6,012 from above and the retest holds (a 5-minute close back above the line), long 1 ES at 6,012.00.
• Stop 6,009.00: acceptance back below VWAP means the premise is dead. 3 points, $150 per contract before costs.
• Target 6,020.00, just under the +1σ band (about 6,021): 8 points, $400. Reward to risk about 2.7 to 1.

Without the reclaim, 6,012 was the line you were shorting under. The level did not move; the auction’s description did. The reclaim did not predict anything: it changed which written plan was live and gave you a cheap place to be wrong.

The failure. Same morning, another path: the 12:00 retest slices through 6,012 and two 5-minute bars close below 6,009. The stop is hit, $150 plus costs. Until reclaims are scored against a baseline, assume this ending is common, and size for it.

Prop-firm note. The stop you place is the number an evaluation’s loss limits see. Check your firm’s current rules; the Prop Firm course covers loss limits and order handling.

Common mistake: Fading the ±2σ band mechanically: “extended, must revert.” On one-sided days the bands widen behind price and it can press +2σ for hours while mean-reversion shorts are stopped one after another. Band reversion is graded Lore on NQ in either regime, so a band touch alone is never the entry; a fade needs its own written rejection rule (M4.09) and a stop.

Common mistake: Treating “trend days hold one side of VWAP” as a forecast. It is a description of days labelled at the close. Holding above VWAP at 10:30 ET tells you who has been ahead, not what the afternoon will do.

Drill: score your own reclaims

For ten sessions, mark every time NQ or ES loses RTH VWAP for at least 30 minutes and then retakes it by your written M4.01 acceptance checks.

1. Before the session, write the rule: how many closes, which bar size, what volume or delta check.
2. For each reclaim, log time, price, VWAP, and whether price reached the +1σ band before it closed back beyond VWAP on the other side.
3. Do the same for every plain VWAP cross that did not meet your acceptance rule.

Recap:

• VWAP is the average price traded since the session start; the “institutions defend it” story is a hypothesis, not yet measured.
• Above or below VWAP describes who has been ahead so far; it filters which plans you write, it does not time entries.
• Bands measure stretch, not turns: band reversion was tested on NQ and not supported.
• A reclaim with acceptance is a control change worth a written plan with a cheap stop; its edge is not yet measured.
• Anchor to events on news days, and know whether your line starts at 09:30 ET or 18:00 ET.

Further reading:

• Volume-Weighted Average Price: benchmark mechanics, slippage, participation algorithms.
• Optimal Execution of Portfolio Transactions (Almgren and Chriss): the standard model for splitting a large order over time, the problem VWAP algorithms solve.
• Order Flow, Measured: the public practitioner study behind the VWAP band, anchored VWAP and trend-day results (claims RL-01, RL-02, RL-06).
• Our lab: VWAP band reversion graded Lore on the NQ study; the reclaim is not yet measured.

This lesson connects to: M3.07 (VWAP mechanics and the band evidence) · M4.01 (acceptance, the reclaim’s test) · M4.03 (day types as management labels) · M4.09 (writing a confirmation rule you can score) · M5.08 (pullback to VWAP or value).

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