Anchored VWAP has one setting, and it is not a number. Every other indicator asks you for a lookback: 14 periods, 20 days, 2 standard deviations. Anchored VWAP asks you for a date, and the whole method lives in that choice. Get the anchor right and the line marks where a specific group of buyers is collectively break even. Get it wrong and it is a squiggle.
Brian Shannon, CMT, is the reason most traders have heard of it. He has taught it at alphatrends.net for years, presented it to the CMT Association, and wrote Maximum Trading Gains With Anchored VWAP (2023) about it. This guide covers what the indicator computes, where he says to put the anchor, how to read it, and the honest limits of testing it.
A note on integrity before we go further: InvestingPaths is an independent research site and is not affiliated with, endorsed by, or connected to Brian Shannon, Alphatrends, or the CMT Association. Everything below is our summary of publicly published material, attributed as such. Trading involves substantial risk of loss, and nothing here is financial advice or a promise of any result.
What it actually computes
VWAP is the volume weighted average price: every trade is weighted by its size, so a million share print counts a million times more than a single share print. It answers a question a simple moving average cannot, which is what did people actually pay, rather than what the price happened to be at a series of moments.
Standard VWAP resets every session, which makes it a day trader's tool and nothing else. Anchored VWAP removes the reset. You pick a starting bar and the average runs from there to now, accumulating.
That single change turns it from an intraday reference into something that can span years, and it changes what the line means. In Shannon's framing, VWAP is an objective measure of what the average participant has paid over a chosen period, so price above the anchored line means buyers from that anchor are collectively in profit and in control; price below it means the opposite. The line is not support because of geometry. It is support because of the people sitting on it.
Where to put the anchor
The anchor should be a bar where a large number of people made a decision. Those are the bars where a lot of stock changed hands at once, and where a new consensus about the company was formed.
| Anchor | What the line then measures |
|---|---|
| Earnings day | What every buyer since the last set of numbers has paid |
| A major news or FDA or merger day | The cost basis of everyone who acted on the news |
| The highest volume day on the chart | Where the largest single transfer of ownership happened |
| A significant swing high | The average price paid by everyone who bought the top, the classic overhead supply level |
| A significant swing low | The average price of everyone who bought the bottom, and the level they defend |
| The IPO or first trading day | Every holder in the stock's history |
| The start of the year or quarter | Where institutional performance is being measured from |
The swing high anchor is the one worth understanding first, because it explains the behaviour people find spooky. Anchor to an obvious top, watch the stock fall for months and then rally back, and price frequently stalls exactly at that line. Nothing mystical is happening: that is the average price paid by the trapped buyers, and it is where they get out at break even.
How to read it
The rules Shannon teaches are about control, not crossings.
- Above the anchored line, buyers from that event are in profit. Pullbacks toward the line are where buyers who missed the move get a second entry at the average price, so the line acts as support.
- Below it, the same population is underwater, and rallies into the line meet people who want out at break even, so it acts as resistance.
- The slope matters. A rising anchored VWAP with price above it is a healthy trend. A flat one is a market with no directional consensus since the anchor.
- Multiple anchors are the point. Several lines from several events, converging in the same area, mark a level that many different groups of buyers care about. That confluence is the setup, not any single line.
There is no overbought or oversold reading, no threshold, and no signal. This is a positioning map rather than an oscillator, and treating it like one is the most common mistake made with it.
Setting it up in TC2000 and elsewhere
Anchored VWAP is now standard across charting platforms. In TC2000 it is available as a chart study you place on a selected bar rather than something you configure with parameters, which is the point: the placement is the configuration. Most competitors (TradingView, Deepvue, TrendSpider, Thinkorswim) ship a version, and they agree on the maths.
The differences worth knowing before you pick a platform:
- How many anchors can you keep on one chart at once. Confluence is the technique, so a tool that allows one anchored line at a time makes the method awkward.
- Whether the anchor survives a timeframe change. Some implementations recompute from the visible bars rather than the anchored date.
- Whether you can scan on it. Almost nowhere can you. See below.
Our TC2000 pricing breakdown covers which tier you need, and the TC2000 versus TradingView comparison covers the charting differences in general.
The honest part: you cannot scan or backtest this properly
This is the section most articles about anchored VWAP leave out, so here it is plainly.
No mainstream screener lets you scan for it, because a scan needs a rule that applies uniformly to thousands of symbols, and the anchor is chosen per stock by a human looking at that stock's chart. "Anchor to the last earnings gap" can be automated. "Anchor to the bar that mattered" cannot, and the second one is the method.
Our backtester has no VWAP rule at all, anchored or otherwise, and we would rather say that than sell you an approximation. What the 70 rule library does contain is the family of ideas that overlap with it: Price vs moving average, Distance from a moving average, Volume vs its own average, and Price vs Keltner channel. Those measure related things, and none of them is a cost basis.
What you can test, and what is worth testing, is the claim underneath the technique: that buying pullbacks within an established uptrend paid better than buying strength outright. That is a mechanical question, and the trend pullback template is the closest expressible version of it, free and with no account. If the general idea does not survive contact with daily bars, a better placed anchor is unlikely to rescue it.
Beware backtests you find elsewhere. Any published anchored VWAP backtest has had to automate the anchor choice, usually to a fixed lookback or a mechanical event. That is a legitimate study of that rule, and it is not a study of what a discretionary trader does with the tool.
Frequently asked questions
What are the best anchored VWAP settings? There are no numeric settings. The only input is the anchor bar, and the useful anchors are high volume events: earnings, major news, an obvious swing high or low, an IPO, or the highest volume bar visible on the chart.
Where do you anchor VWAP? To a bar where ownership changed hands in size and a new consensus formed. Shannon's own guidance is to anchor to the significant areas on a chart: highs, lows, big volume days and events.
What is the difference between VWAP and anchored VWAP? Standard VWAP resets at the start of every session, so it only describes today. Anchored VWAP starts at a bar you choose and keeps accumulating, so it can describe months or years.
Does anchored VWAP work on daily charts? Yes, and that is where the long horizon anchors live. Standard session VWAP is intraday only; the anchored version is timeframe agnostic because the anchor is a date rather than a number of bars.
Can I screen for stocks near their anchored VWAP? Not in any mainstream screener, because the anchor is chosen per chart. The nearest practical workflow is to screen for candidates on ordinary conditions, then anchor by hand on the ones that pass.
Test the idea underneath it
The anchoring is discretionary, but the assumption it rests on is not. The trend pullback template tests whether buying dips inside an established uptrend was paid on real daily bars, free and with no account, and the backtester lets you vary how deep a pullback has to be before it counts.