Volume-Weighted Average Price (VWAP)
VWAP is the average price an asset traded at over a period, weighted by volume, serving as a benchmark for execution quality and intraday fair value.
Volume-Weighted Average Price (VWAP) is the average price at which an asset has traded over a specified period, with each trade weighted by its volume. It is computed by summing the price times volume for every trade (often using the typical price, the average of high, low, and close for each interval) and dividing by the total volume over the period. Because it weights by volume, VWAP reflects where the bulk of trading actually occurred, not just where prices touched, making it a more representative measure of fair value than a simple average.
VWAP is most commonly anchored to a single trading session and reset each day, giving an intraday benchmark. Institutional traders use it heavily for execution: buying below VWAP or selling above it is considered favorable execution, and many algorithmic order types are designed to track VWAP to minimize market impact. Discretionary traders also use VWAP as a dynamic support/resistance line and a gauge of intraday sentiment, with price above VWAP read as bullish control and below it as bearish. Anchored VWAP, started from a meaningful event such as an earnings release or a swing low, extends the concept to multi-day analysis.
The main limitations are that standard VWAP is intraday and resets daily, so it is less useful for long-horizon investing, and like all averages it is backward-looking. It is a benchmark and context tool, not a predictive signal by itself.
On hedgewing.ai, VWAP and price-relative-to-VWAP are natural members of the engineered feature set because they combine price and volume into a single, economically meaningful reference. Feeding the ensemble where price sits relative to volume-weighted fair value gives the LSTM, GRU, TCN, and Transformer models information about institutional positioning that pure price features miss, and as with every feature its predictive contribution is measured through walk-forward backtesting.
Related terms
Support and Resistance · Moving Average · Algorithmic Trading · Feature Engineering
Back to the hedgewing.ai glossary · See AI stock forecasts