What is VWAP and what was it built for?
VWAP is the volume-weighted average price: the total value traded divided by the total volume traded over a period. Unlike a moving average, which weights every bar equally, it weights each price by how much actually changed hands there.
Its original purpose is execution benchmarking rather than signal generation. An institution filling a large order over a day is measured against the day's VWAP, because beating it means the execution was better than the market's own average. That is a well-defined problem and VWAP solves it exactly.
The retail usage is different and much looser: treating VWAP as a dynamic level that price reverts to or respects, and taking positions on the basis of price sitting above or below it. That is a separate claim from the benchmarking one and requires its own evidence.
Anchored VWAP extends the idea by starting the calculation at a chosen event rather than at the session open. That makes the anchor point a parameter, and a parameter chosen by looking at the chart is a familiar problem in a new place.
Why is VWAP different in forex?
Forex has no central exchange, so there is no consolidated volume figure, and what charting platforms display as volume is tick volume — a count of how many times the price updated, not how much was traded.
Tick volume correlates with activity and is not the same quantity. Two brokers will show different tick volume for the same period because they aggregate different liquidity providers and update at different rates. A VWAP computed from it is therefore broker-specific, and the same chart on a different feed gives a different line.
That has a direct consequence for testing. A strategy keyed to VWAP levels graded on one data source may not reproduce on another, and the difference is not noise that averages out — it is a systematically different input series.
The benchmarking purpose does not survive this either. Beating a VWAP built from your own broker's tick count is not a meaningful execution standard, because the denominator is not the market's volume.
None of this applies to exchange-traded instruments. Index futures and equities have genuine consolidated volume, and VWAP on those means what it is supposed to mean.
Does price actually revert to VWAP?
Whether price reverts to VWAP is a testable question, and testing it requires a fixed horizon and a count of the excursions that never returned — the same discipline any level-based claim needs.
The trap is identical to the one in gap-fill statistics. Looking back at a chart and checking whether price returned to VWAP will find that it usually did, because VWAP is an average of recent prices and recent prices are near where price has been. The statistic is close to being true by construction, and it is useless without a horizon.
A fair test enumerates every excursion beyond a normalised distance from VWAP, and grades each at a fixed horizon: returned, partially returned, or still away at expiry. Reporting all three is what makes the number interpretable.
The comparison that matters most is against a plain moving average of the same period. If VWAP performs no better than an unweighted average, the volume weighting is contributing nothing, and the extra complexity and the data dependency are being paid for no benefit. In forex specifically this comparison is worth running before anything else.
How should you test a VWAP strategy?
Fix the anchor and the session boundary in advance, normalise distance by volatility, test against an unweighted average as a control, and state which data source the volume came from.
The anchor is the first commitment. Session VWAP resets at a stated time, which must be anchored to a market clock and held consistent across daylight-saving transitions. Anchored VWAP starting from a chosen swing or event needs a mechanical rule for choosing that event, otherwise the anchor is selected with hindsight and every result is contaminated.
Distance from VWAP has to be normalised. Two units away means something different on a quiet day and a volatile one, and a fixed threshold turns the rule into a volatility filter without anyone intending it.
The control test is the one that decides whether VWAP is earning its place. Run the identical rule with a simple moving average substituted, on the same data and the same parameters, and compare. This is cheap, and it is the only way to attribute a result to the volume weighting rather than to the fact that any average of recent price sits near price.
Where does QuantParadox stand on VWAP and volume?
QuantParadox does not offer volume as a setup filter or as a sizing input, and that is a deliberate position based on our own testing rather than a gap in the feature list.
The finding is published separately: in decade-scale testing across forex markets, requiring above-average volume did not measurably improve setup performance. A volume filter is an easy feature to build and an easy one to market, and the result went the other way. Building VWAP-based filters on top of forex tick volume would be building on the same input that did not carry information in that test.
What the platform does grade is the price component of level-based rules. A strategy that reacts to a dynamic average, a session reference price, or a distance-from-mean condition can be expressed and tested on price alone, and where a rule genuinely needs consolidated volume the honest answer is that forex does not have it to give.
For exchange-traded instruments the argument is different, and this is a real limitation to state plainly: if your strategy depends on genuine exchange volume on individual equities, a platform built around that data is the right tool and this is not it.