What Is VWAP in Futures Trading?

VWAP is the volume-weighted average price: the total value traded over a chosen period divided by the total volume traded in it, so a price where a lot of contracts changed hands counts for more than a price where two did. On a futures chart it is drawn as one line that begins at a start point you configure and accumulates from there, which means the number it shows depends as much on where you told it to begin as on what price has done since.

That configurability is the part that matters on futures, where the session is long and the platform has to be told which part of it counts.

How is VWAP calculated?

VWAP multiplies each price by the volume that traded at it, adds those products across the period, and divides the total by the period's volume. Sierra Chart publishes the calculation for its study in exactly that form: the sum of each bar's input price multiplied by that bar's volume, divided by the summed volume of the bars in the period.

The platforms differ in what "the price of a bar" means. Quantower's documentation writes the algorithm using the average of a bar's high, low and close, and offers a choice of seven price inputs when the calculation runs off bars rather than off ticks, ranging from a single price such as the close to four-component averages. Sierra Chart's equivalent input defaults to using the bar's last trade price and the bar's total volume, with a separate setting that pushes the calculation down onto the underlying data instead.

The divisor is what gives the line its character. Because it is cumulative volume, each additional trade has a smaller effect than the one before it. Ten thousand contracts into a session, a single contract printing ten points away from VWAP moves the line by about a thousandth of a point; a hundred thousand contracts in, by a ten-thousandth. That is arithmetic rather than a market property: the same quantity traded at the same distance from the line moves it less late in a period than early in one, purely because the divisor has grown.

What is the difference between VWAP and a moving average?

VWAP weights by volume and a simple moving average does not. Both produce an average price line, but the formula Sierra Chart publishes divides by the total volume of the period, while an unweighted average of the same bars would divide by the number of bars — so one bar that traded fifty times the volume of its neighbour counts fifty times as much in the first and once in the second.

The second difference is the window. A simple moving average of a set number of bars is by definition a rolling window: each new bar enters the average and the oldest leaves it, so the length never changes. VWAP as documented by both platforms resets at a period boundary and accumulates until the next one, so it has a beginning and an end rather than a rolling length.

Neither difference makes one of them better. They answer different questions, and only VWAP needs volume at all. Note which volume: VWAP is computed from contracts that actually traded, which is a different quantity from the resting order sizes a depth of market ladder puts beside each price.

When does VWAP reset on a futures chart?

VWAP resets whenever the period you configured ends, and on a futures chart that boundary is a setting rather than a fact about the market. Sierra Chart's study restarts at the beginning of each new period, with the period set by two inputs, and when the period is a day it reads the start of the trading day from the session times configured for the chart.

This is where futures make the setting load-bearing. A CME futures session opens the previous evening and runs through the following afternoon, as our trading hours article sets out, so "the start of the trading day" is itself a choice: both platforms take it from session settings you control, and a line started at the evening open has absorbed the overnight trade while a line started at a session boundary you defined later in the day has not. They are different numbers on the same chart at the same moment.

Two documented traps follow from that. Sierra Chart notes that when the period type is days, the count refers to calendar days rather than trading days. And Quantower handles the same problem from the other end, with a session template setting that selects which trading session the calculation runs inside, drawing on predefined sessions or ones you define yourself. Before reading anything into a VWAP line, it is worth confirming which of these it is.

What is anchored VWAP?

Anchored VWAP is the same calculation with its start point chosen by hand instead of by a session boundary. Quantower's anchored tool fixes the left edge of the range at a point you specify and lets the right edge advance with each new bar, while a second mode fixes both edges so the window stops where you put it.

Sierra Chart exposes the same idea twice. One input tells the study to ignore the period settings entirely and calculate from the first bar on the chart, or from a start date and time you enter — and the platform requires both a date and a time, not a time alone. Separately, a drawing tool lets you place the start and end interactively, adding a study for each line drawn.

The arithmetic consequence is worth stating plainly. A line anchored to a recent point has little volume in its divisor, so it moves quickly; a line anchored weeks back has an enormous divisor and barely moves at all. The choice of anchor is therefore a choice about how responsive the line is, and two traders anchoring to different events are not disagreeing about the market so much as computing different averages.

What are VWAP standard deviation bands?

VWAP bands are lines placed a set distance above and below the VWAP line, with that distance taken from a dispersion measure computed around it. Sierra Chart documents a variance built from each bar's squared distance from VWAP weighted by that bar's volume and divided by the period volume, with the standard deviation as its square root, and it places up to four lines above and four below.

The spacing is a multiplier applied in steps: the first band sits one multiplier away, the second two, the third three, the fourth four. The platform also offers alternatives to the standard deviation for that offset, including a fixed price offset and a percentage of VWAP, which means a band drawn on one chart and a band drawn on another can be built from different quantities while carrying the same label.

Quantower adds a different measure again. Its documentation calls the maximum permissible deviation similar to a standard deviation, but defines it as the period's high minus the period's low, divided by two — a half-range, computed without reference to volume at all, so it will not generally land in the same place as a volume-weighted standard deviation of the same nominal size.

The practical point for anyone reading someone else's chart: "the two-sigma band" is not a single well-defined level in futures. It depends on the method chosen, the multiplier entered, the anchor, and the bar size — Sierra Chart says outright that its band lines can come out different on different bar timeframes, because the calculation uses the chart bar values and a five-minute chart holds only every fifth value a one-minute chart does.

What was VWAP actually built for?

VWAP was built to measure execution quality, not to forecast price. McCulloch and Kazakov, in a 2007 University of Technology Sydney research paper, credit its development as a quality-of-execution measurement to Berkowitz, Logue and Noser, and quote their argument that measuring market impact requires a benchmark that is an unbiased estimate of the prices any randomly selected trader could have achieved over the period.

The use case behind that is institutional. A large order broken into pieces across a day can be scored against the day's VWAP to quantify what the order's own size cost in liquidity impact. Busseti and Boyd, in a 2015 paper on optimal execution, give three examples of a benchmark price a client and a broker might agree on — the price at the start of the schedule, the price at the day's close, and VWAP — describe VWAP as the most common of them, and note that it assigns most of the risk from market price moves to the client, leaving the broker to optimise execution alone. Their model also refuses the convenient assumption that the day's total volume is known in advance, treating it instead as a random variable — which is precisely the uncertainty a trader watching a live VWAP line is sitting inside.

What does the research actually show about VWAP?

The research cited here is about executing large orders in equity markets, and it does not test VWAP as a directional signal. McCulloch and Kazakov built their empirical work on NYSE trade records pulled from the TAQ database, reporting 203,158 sample paths of intraday relative volume. Busseti and Boyd simulated executions on the thirty Dow Jones Industrial Average stocks over sixty market days of NYSE data from late 2012, splitting each day into 390 one-minute intervals. Neither paper uses futures data.

One finding from the 2007 paper travels usefully as an idea rather than as a number: the residual risk of failing to match VWAP scales with price variance and inversely with a power of the stock's final trade count, so the benchmark is easier to track in heavily traded names. That was measured on stocks. Whether the same relationship holds between a heavily traded index future and a thinly traded contract is a question that paper does not answer, and it is not the question behind the choice between an E-mini and a Micro E-mini, which turns on position size.

A 2005 preliminary working paper by Bialkowski, Darolles and Le Fol, working on the forty CAC 40 constituents as of the beginning of September 2004 over a year of Euronext tick data ending that August, takes it as established that intraday volume moves around a U-shaped seasonal pattern. McCulloch and Kazakov put the same shape in words: markets are on average busy at the open and the close and quieter through the middle of the day. If volume clusters that way, a VWAP accumulating through a session is weighted toward those clusters. Both are equity studies — one French, one American — and the 2005 paper is explicitly a preliminary version rather than a published result.

What does VWAP mean on a funded account?

VWAP changes nothing about the rules a funded account runs under. Accounts here are simulated, with a real payout on the profits, and a VWAP line is only a reference: the drawdown line is a number, and no indicator moves it. The arithmetic that sizes a trade is still stop distance in ticks multiplied by tick value, divided into the distance to that line, which is worked through in our post on futures position sizing.

Where VWAP touches the account is in the trades it is used to argue for. Quantower's own documentation offers one: it describes trading the return of price to an hourly VWAP on a five-minute ES chart, and adds that the tactic requires a meaningful gap to have opened between the line and the close first. That is a platform's worked example, not a tested result, and the account rules do not grade it either way — the flat-by time arrives whether or not price has come back, and the drawdown line does not move because a line said it should.

Two practical checks are worth doing before a VWAP means anything to you. First, confirm which session your platform anchored it to: on both of the platforms documented above that is a setting you control rather than a default you can assume, and the supported platforms article lists what runs here. Second, confirm the instrument actually trades enough volume in the window you are looking at for a volume-weighted average to mean much; the tradable list is on the instruments page.

FAQ

Is VWAP the same number on every platform?

Not necessarily. Sierra Chart's documentation says its study can run off a bar's last trade price and total bar volume or off the underlying tick data, and that values will differ between those settings and between bar types without a consistent start time, such as volume or range bars. Quantower offers a choice between tick data and current-timeframe bar data, and seven different price inputs for the bar-based version. Same formula, different inputs, different line.

Does VWAP reset every day on futures?

Only if you configure it to. Sierra Chart resets at the start of each new period and takes the start of a trading day from the session times set on the chart, and its day-count input refers to calendar days rather than trading days. Quantower calculates within a session selected from a session template. Because a CME futures session spans an evening open and the following afternoon, "daily" VWAP has more than one defensible meaning on a futures chart.

What is MPD in Quantower's VWAP settings?

MPD stands for maximum permissible deviation, and Quantower's documentation defines it as the period's high minus the period's low, divided by two. It is offered alongside standard deviation bands as an alternative way to place lines around the VWAP. Because it is built from the range rather than from volume-weighted dispersion, it will not generally produce the same distance as a standard deviation band of the same nominal size.

Does VWAP predict where price is going?

None of the sources here tests that. Quantower's VWAP page does suggest directional readings — it calls price below VWAP undervalued and price above it overvalued, reads a crossing as a momentum shift or trend change, and describes the line as dynamic support and resistance in a sideways market — but it offers no study behind any of them. The academic papers used above ask a different question entirely: how to schedule a large order so its fills track the market's VWAP, measured on US equities rather than futures. A win rate quoted for a VWAP strategy with no study behind it is not evidence.