VWAP is the average price of a market weighted by volume, usually calculated through the current trading session. Traders use it as a reference line, execution benchmark, and sometimes as a rule input - but VWAP itself is not proof of an edge.
That distinction matters. A line that many desks watch can be useful information. It can also become a clean-looking excuse to trade without a test. The honest version is simple: define the VWAP rule, include real trading costs, and see whether it survives out-of-sample data like any other hypothesis.
What VWAP actually is
VWAP means volume-weighted average price: total price times volume, divided by total volume. Charting platforms often approximate it from intraday bars by multiplying each bar's typical price by volume, cumulating that value, then dividing by cumulative volume (StockCharts ChartSchool: VWAP, Charles Schwab: volume-weighted indicators).
VWAP = cumulative(price x volume) / cumulative(volume)
Common bar approximation:
typical price = (high + low + close) / 3
VWAP = cumulative(typical price x volume) / cumulative(volume)Traditional VWAP is an intraday tool. StockCharts describes it as a calculation that starts when trading opens and ends when it closes, and Schwab describes the denominator as the total volume traded that day (StockCharts ChartSchool: VWAP, Charles Schwab: volume-weighted indicators).
How traders use VWAP
VWAP has two different jobs, and mixing them up is where bad analysis starts.
| Use | What VWAP tells you | What it does not tell you |
|---|---|---|
| Execution benchmark | Whether an order filled better or worse than the market's volume-weighted average during the execution window | Whether the trade idea was good |
| Intraday reference | Whether price is trading above or below the day's volume-weighted average | Whether price must reverse or continue |
| Rule filter | A mechanical condition, such as only taking long setups above VWAP | Whether that filter improves expectancy after costs |
The execution use is the more institutional one. The CFA Institute Research Foundation describes market VWAP as a common benchmark for computing transaction costs, and Interactive Brokers offers a best-efforts VWAP algo that seeks the volume-weighted average price over the order window (CFA Institute Research Foundation: Trading and Electronic Markets, IBKR Campus: VWAP Algo). Nasdaq also lists VWAP among execution algorithms, with a schedule based on historical volumes (Nasdaq: execution algorithms).
For chart traders, VWAP becomes a reference point. Schwab notes that price below VWAP can suggest sellers are dominating, while price above VWAP can suggest buyers are in control; StockCharts describes VWAP as a way to compare current price with the volume-weighted average for the current day (Charles Schwab: volume-weighted indicators, StockCharts ChartSchool: VWAP).
That is useful context, not a signal. "Above VWAP" can mean strength in a trend day. It can also mean an extended move that is late for a new entry. The line does not decide that for you.
Anchored VWAP
Anchored VWAP applies the same volume-weighted logic from a chosen starting point instead of the session open. StockCharts defines Anchored VWAP as VWAP from a specific anchor, and TrendSpider describes the anchor as a selected event such as a major low, breakout, news event, or other meaningful point (StockCharts ChartSchool: Anchored VWAP, TrendSpider: Anchored VWAP).
That makes Anchored VWAP more flexible and more subjective. A session VWAP has a default start. Anchored VWAP asks the trader to choose the event that matters.
Session VWAP:
open ---------------------------- close
^ calculation starts here
Anchored VWAP:
news / high / low / breakout ---- latest bar
^ trader chooses the anchorThis is where discretion enters. Anchoring to a swing low, a gap, or a failed breakout can all be defensible. They are not the same test.
What the critics get right
VWAP can be precise and still be incomplete. The calculation is objective once the data window is fixed, but its interpretation is not.
The transaction-cost literature is clear about one limitation: beating VWAP is not the same thing as measuring the full economic cost of a trade. The CFA Institute Research Foundation notes that VWAP transaction-cost estimates can miss price impact when a large trader participates in the market, and transaction-cost analysis generally separates explicit costs from hidden costs such as slippage and market impact (CFA Institute Research Foundation: Trading and Electronic Markets, MillTech: transaction cost analysis).
For directional traders, the critique is simpler: VWAP is a benchmark, not causality. Price crossing VWAP may describe a change in intraday control, but it does not prove that the next trade has positive expectancy. If a VWAP setup only looks obvious after the move, you are reading a chart, not testing a method.
How you'd actually test it
To test VWAP trading properly, first turn the idea into a rule a computer can repeat without opinion.
For example:
- Define the market and session. VWAP in US equities, index futures, FX CFDs, and crypto will not behave identically because the trading day and volume feed differ.
- Define the VWAP source. Tick data, one-minute bars, and five-minute bars can produce slightly different lines.
- Define the setup. Example: "price opens below VWAP, reclaims VWAP, then holds above it for N bars."
- Define the entry, stop, exit, and time stop. No discretionary "it looked clean" filter.
- Include spread, slippage, commission, and swap where relevant.
- Split in-sample and out-of-sample. Optimise on one period, judge on data the rule never saw.
Then compare variants:
| Variant | Question |
|---|---|
| VWAP trend filter | Does trading only above VWAP improve expectancy versus the same setup without VWAP? |
| VWAP mean reversion | Does fading distance from VWAP work only in sideways sessions, or does it get damaged by trend days? |
| Anchored VWAP | Does a fixed anchor rule beat random or obvious hindsight anchors? |
| Session filter | Does the rule depend on the open, lunch, close, or a specific market session? |
This is the realbacktesting standard applied to an indicator concept: not "does VWAP feel respected?" but "does a defined VWAP rule survive real costs and out-of-sample data?" The same discipline sits behind how to verify a cTrader backtest and why a cost-free backtest lies.
Frequently asked
Is VWAP a trading strategy?
VWAP is not a strategy by itself. It becomes part of a strategy only when you define a setup, entry, stop, exit, cost model, and test window.
Is VWAP better than a moving average?
VWAP answers a different question. A moving average weights bars by time, while VWAP weights price by traded volume inside a defined window.
Does VWAP work in forex?
VWAP can be calculated on forex data if the platform has a usable volume proxy or trade-volume feed, but spot FX is fragmented and largely over-the-counter rather than a single central exchange tape (BIS: The foreign exchange market, Investopedia: forex market). That makes the data source part of the test, not a detail to ignore.
What is the difference between VWAP and Anchored VWAP?
Traditional VWAP usually resets at the session open. Anchored VWAP starts from a trader-chosen event or bar, so the anchor choice becomes part of the method.
The stubborn takeaway
VWAP is a good reference price and a bad shortcut. If the rule is not precise enough to backtest, the line is only giving your opinion a cleaner shape.