NQ · LIVE ENGINE Algomyth
EN
Order flow

VWAP for Futures: Deviations, Anchors and Real Setups

Where the calculation starts, what the 2 sigma band is worth, the four anchors that matter, and the days VWAP is worth nothing.

Published 7 min read By the Algomyth desk

Wall Street street sign in New York, home of the institutional desks whose executions are benchmarked against VWAP

VWAP is the most watched line in intraday futures and the most misused. Most traders drop it on the chart, watch price respect it a few times, and start treating it like magic support. It is not. VWAP is a volume weighted average price, and its usefulness depends entirely on where you start the calculation and where you are in the session.

Here is what actually works: how it is calculated, why institutional desks care, what the deviation bands do, where to anchor it, and the days it is worth nothing.

What VWAP is and why so many people watch it

VWAP stands for Volume Weighted Average Price. The formula is simple: multiply each traded price by the contracts that crossed at that price, add it all up, divide by total volume for the period.

Translated: it is the real average price the instrument has traded at since the calculation began, weighted toward the prices where size actually changed hands. A 20-period moving average treats a 300-contract bar exactly like a 12,000-contract bar. VWAP does not.

That is why it matters. Institutional desks executing large orders get benchmarked against VWAP. If a portfolio manager has to buy 4,000 contracts across the day, the execution gets graded on whether it filled above or below session VWAP. That produces real, repeatable behavior: below VWAP, buyers who need to execute show up, above it sellers do. Not every time, but often enough for the line to mean something.

That does not happen with a 21 EMA. Plenty of people watch a 21 EMA. Nobody has a financial incentive to execute against it.

New York Stock Exchange facade draped with American flags, home of the institutional desks whose executions are benchmarked against VWAP
VWAP matters because real money gets graded against it.

Where the calculation starts: the detail that changes everything

A VWAP means nothing until you know when it started accumulating. This is where half of all traders have it misconfigured without realizing.

Trading screen illustrating vwap for futures: deviations, anchors and real setups
Where the decision actually gets made: the screen, not the theory.

Session VWAP resets when the session starts. The catch is what your platform thinks the session start is. On CME futures the electronic session opens at 6:00 p.m. ET the previous day. If your VWAP starts there, by 9:30 a.m. it already carries 15 hours of thin overnight volume, and the line ends up glued to a price nobody cares about.

Traders working the US open almost always use one of these two:

  • VWAP anchored to the 9:30 a.m. ET open. This one reflects what today real volume is doing.
  • Full 24-hour session VWAP, useful for overnight context, especially on days with a European catalyst.

If you only get one for intraday index futures, take the one anchored to the New York open. Thirty minutes of setup completely changes the quality of the line. Session-by-session timing is broken down in futures market hours.

Deviation bands: 1, 2 and 3 sigma

The bands are standard deviations of price from the VWAP itself over the same period. They plot symmetrically above and below, and behave a bit like Bollinger Bands except anchored to an average price with an economic reason behind it.

What each one is worth:

BandWhat it tells youPractical use
VWAP (center line)Real average price of the sessionNatural target on rotational days
1 sigmaNormal working range, around 68 % of the timeWeak alone, useful as a trend filter
2 sigmaMeaningful extension, price is rich or cheapScale-out zone and fade zone with confirmation
3 sigmaExtreme, usually news drivenDo not fade blind, wait for flow to stall

The 2 sigma band holds about 80 % of the value here. Price at 2 sigma on a day with no scheduled data and declining volume is one of the few spots where mean reversion has real numbers behind it. On MNQ that sits roughly 40 to 90 points off VWAP depending on the day.

One warning: on strong trend days price rides the 2 sigma band for hours. Fading it there is the fastest way to lose an account before lunch. Before trading against the band, check whether price has touched VWAP at least once that session. If it has not, it is not a rotational day.

Anchored VWAP: the four anchors worth using

Anchored VWAP (AVWAP) is the same math starting from a point you choose. This is where it gets genuinely useful, because you can measure the average price of one specific move instead of an arbitrary calendar day.

Futures trading desk related to vwap for futures: deviations, anchors and real setups
The number you did not calculate is the one that costs you the account.

1. The New York open

The basic one. Average price of everything since real volume arrived. Used as an intraday bias line: above it, buyers have the edge, below it, sellers.

2. Yesterday high or low

Anchor at yesterday extreme and you get the average price paid by everyone who has traded since that peak. Tells you whether the people who bought the high are underwater.

3. A news release

Anchor on the CPI, payrolls or FOMC candle. That VWAP is the average price of all the paper that has traded with the new information already public. It usually holds as a level for two or three days.

4. The start of the move in progress

If the market has been climbing for three days, anchor where the climb began. While price stays above it, the move is healthy. When it loses that AVWAP and cannot reclaim it, the impulse is done.

Four anchors is plenty. Twelve lines on screen do not give you context, they give you paralysis.

Trading screen crowded with green, orange and purple market data, the result of stacking too many anchored VWAP lines on one chart
Four anchors help. Twelve leave you unable to decide anything.

How VWAP behaves in each block of the session

Same line, different behavior depending on the clock. All times Eastern:

9:30 to 10:30 a.m. The open-anchored VWAP is unstable for the first 15 minutes because it has almost no accumulated volume and swings hard. Do not treat it as a level before 9:45. After that, it is the main reference of the day.

10:30 a.m. to 12:00 p.m. The best block. VWAP has settled, volume is real, reactions are clean. This is where the 2 sigma band gives its best entries.

12:00 to 1:30 p.m. Lunch. Volume drops, price drifts and VWAP flattens. Levels still hold but moves are short. If you trade here, cut the target.

1:30 to 4:00 p.m. Volume returns. VWAP regains relevance and a final test before the close is common. The opening block is covered in detail in the NY open strategy guide.

2 sigma reversion

Conditions: no major data on the calendar, price has touched VWAP at least once, and it reaches the 2 sigma band on declining volume. Enter after the first candle that closes back inside the band. Stop a few points beyond the extreme. First target 1 sigma, second target VWAP. On MNQ, typical risk of 10 to 14 points against 25 to 50 points of target.

VWAP retest in trend

The most boring and most reliable one. Price holds above VWAP all morning, pulls back into it, and bounces. Enter on the bounce, stop below VWAP plus a cushion, target the prior high. It works because you are trading with the bias and against the noise, not the other way around.

Losing and accepting the other side

Price crosses VWAP and stays on the far side for more than 15 minutes without reclaiming it. That is an intraday bias flip. Entry comes on the first pullback that fails to cross back. Best setup on reversal days, and the one that stops you from buying a morning that already turned.

When VWAP is worthless

The honest list:

  • The first 10 minutes. Barely any accumulated volume, the line whips. It is not a level yet.
  • Scheduled data days. At 8:30 a.m. on CPI, price cuts through VWAP like it is not there. Wait 20 minutes.
  • Instruments without real volume. A CFD VWAP is computed off volume the broker made up.
  • Higher timeframes. A weekly VWAP exists, but as an intraday level it does nothing.
  • When price has been 40 points away for hours. VWAP is not failing. The day is not rotational. Change tools, not your opinion of the tool.

Setting it up in NinjaTrader 8

NinjaTrader 8 ships a session VWAP in the indicator list and supports anchored versions. Three settings worth checking:

  1. Session template. This decides whether it accumulates from the electronic open or the regular session. It changes results more than anything else here.
  2. Standard deviations. Turn on 1 and 2. Add 3 only if you trade high volatility days.
  3. Price type. The default is typical price ((high + low + close) / 3). Leave it unless you know why you are changing it.

If you plan to anchor VWAP to a specific bar, you need tick data downloaded back to that date. Without history the line is miscalculated or simply does not draw. Volume profile has the same requirement, covered in the volume profile guide.

Hand pointing at a candlestick chart on a touchscreen surrounded by market data devices, the moment of judging whether price has accepted the far side of VWAP
The decision is not the touch. It is whether price stays on the other side.

VWAP is not a system. It is a fair-value reference telling you whether you are paying rich or cheap versus what the market has paid today. Use it to choose where to try and how big, and let something else give you the entry.

Sources and further reading

vwap futures trading

Educational content, not investment advice. Trading futures involves substantial risk of loss and is not suitable for every investor. Any reference to Algomyth results is hypothetical, simulated over historical data, and shown in full on the performance page; past performance does not guarantee future results.

Algomyth Live

Take the seat: watch the engines live

Three specialist engines on Nasdaq, S&P and Gold, every signal streamed the moment it happens, with the stop and the target written before the trade. From $149 a month per market.