VWAP deep dive: anchored VWAP and standard-deviation bands
The short answer
Standard VWAP resets every session, so it only describes the current day. Anchored VWAP uses the same volume-weighted formula but starts its cumulation from a chosen event, a swing high or low, an earnings gap, a breakout, so it runs across days and marks the average price paid by everyone who traded since that pivot. That is why it often behaves as support or resistance: those participants sit near break-even. Standard-deviation bands then wrap VWAP in a volume-weighted volatility envelope, where an outer-band tag signals a stretched deviation, not a trade.
This is the advanced instalment of the VWAP cluster, written for a reader who already knows that VWAP is cumulative price times volume divided by cumulative volume. If that construction is new, read the plain-English build first in what the VWAP indicator is and why institutions benchmark to it. This page instead attacks the reference-frame mechanics: the single limitation baked into session VWAP, the one design change that removes it, and the statistical envelope that turns a bare line into context. The intraday reversion setup and its trend-day failure are handled separately in VWAP reversion on Indian equities, and this guide hands that detail to it deliberately.
1. The session anchor, and the limit built into it
Every standard VWAP line carries an invisible instruction: reset the cumulation to zero at the session open. On Indian equities that reset fires at 09:15 IST each trading day. From that instant the indicator sums price times volume across the day's prints and divides by cumulative volume, producing a single running average of where the day's business has transacted. It is a clean, honest measure of one thing, and it has one structural blind spot.
The blind spot is memory. Because the denominator restarts at every open, the VWAP line you see at 14:30 contains no information from yesterday, last week, or the move that created the level price is now testing. It is, by construction, an intraday-only object. Two consequences follow. First, VWAP becomes inertial as the session ages: the first hour's prints move the line freely, but by mid-afternoon the cumulative-volume denominator is so large that even violent moves nudge VWAP only fractionally. Second, and more limiting, you cannot carry it across days. Splice one session's VWAP onto the next and the line simply jumps and restarts. There is no continuous multi-day VWAP in the standard construction, because the daily reset forbids it.
That daily reset is not a bug. For an execution desk filling a single day's order it is exactly right. But for a swing trader asking a different question, what has everyone who bought this breakout actually paid, the reset destroys the answer before it can form. Anchored VWAP exists precisely to replace the reset with something a human chooses.
2. Anchored VWAP: fixing the cumulation to an event
Anchored VWAP was developed and popularised by Brian Shannon, CMT, of AlphaTrends, who set it out in his 2023 book Maximum Trading Gains with Anchored VWAP and presented the method to the CMT Association. The mechanical change is almost trivial and the interpretive change is large. As StockCharts puts it, anchored VWAP is "calculated using the same formula as traditional VWAP; the only difference is in the bars that are included in the calculations." Instead of the cumulation starting at the session open, you nominate a starting bar, and the sum runs from there.
for i = anchor bar to now, no daily reset
Set the anchor to the session open and you recover ordinary VWAP. Set it anywhere else and you get something the session version cannot express: the volume-weighted average price of everyone who has traded since that event. That single reframing is the entire idea. A line anchored to a major swing low tells you the average cost of every participant who has transacted since the market turned there. A line anchored to an earnings gap tells you the average cost of everyone positioned since the fundamentals repriced.
The reason this matters is behavioural, and it is why anchored VWAP so often acts as support or resistance. If the anchored line approximates the average entry of a cohort of participants, it approximates their collective break-even. Human beings manage positions around break-even. When price falls back toward an anchored VWAP drawn from a low, the buyers created by that move are, on average, roughly flat, and they tend to defend the level, adding demand: the line reads as support. When price rallies back to an anchored VWAP drawn from a high, the buyers trapped at that peak are, on average, roughly flat at last, and many sell to escape, adding supply: the line reads as resistance. StockCharts notes the corollary that where several anchored VWAP lines from different events converge, "that can indicate an especially strong area of support or resistance," because multiple cohorts share a break-even there.
The power and the weakness are the same fact: the anchor choice is the whole art. The line is only meaningful if the bar you chose marks a real change in market psychology, a decisive high or low, a genuine catalyst, a true breakout. Anchor to a random mid-range candle and the cohort it defines is not a cohort at all, and the line has no reason to be defended. The same chart yields as many anchored VWAP lines as there are starting bars, and only a few of them mean anything. This subjectivity is not a flaw to hide; it is the honest cost of the flexibility. Reading a chart well enough to pick the anchor that matters is upstream judgement, and that upstream work is exactly what the method we teach is built around.
3. Session VWAP versus anchored VWAP, side by side
The two are the same arithmetic aimed at two different questions. Fixing the distinction in a table prevents the most common confusion, that anchored VWAP is a fancier VWAP. It is not fancier; it answers something session VWAP structurally cannot.
| Dimension | Session VWAP | Anchored VWAP |
|---|---|---|
| Anchor | The session open, fixed by the clock, resets at 09:15 IST | A bar you choose: swing high or low, gap, breakout, year-to-date |
| What it shows | Average transacted price of today's business only | Average price paid by everyone who traded since the chosen event |
| Time span | One session; cannot be carried across days | Continuous from the anchor: days, weeks or months |
| Primary use | Intraday execution benchmark and intraday value line | Multi-day support and resistance via cohort break-even |
| Chief weakness | Amnesia: no memory beyond the open | Subjectivity: only as good as the anchor you pick |
4. Standard-deviation bands: a volume-weighted envelope
A VWAP line, session or anchored, tells you the mean but not the spread. Standard-deviation bands add the spread. They are plotted at plus or minus one, two and three standard deviations of price around VWAP, forming a volatility envelope that widens when trading disperses and tightens when it clusters. Conceptually they resemble Bollinger Bands, with one important difference: Bollinger Bands centre on a simple moving average, whereas VWAP bands centre on the volume-weighted average and the deviation itself is volume-weighted. A large move that printed on heavy volume widens the band more than a thin-volume wick of the same size.
= Σ (Vi · Pi2) ÷ Σ Vi − VWAP2
Bandk = VWAP ± k · σ for k = 1, 2, 3
Read the formula in words. First take the volume-weighted average, VWAP. Then measure how far each print sat from it, square that distance, weight it by the volume that traded there, and average across all prints: that is the volume-weighted variance, and its square root is the standard deviation. Each band is simply VWAP plus or minus one, two or three of those standard deviations, recomputed on every new print. The algebraic second line is the identity most platforms actually compute, the volume-weighted mean of price-squared minus the square of VWAP, which is faster to maintain incrementally.
What the bands measure is dispersion of trade around the mean, and that is all they measure. The single most important thing to internalise is what an outer-band touch does and does not mean. A tag of the plus-two or minus-two band says price has reached a statistically stretched distance from the volume-weighted average. It does not say the move is exhausted, that a reversal is due, or that a trade exists. On a rotational day that stretch frequently relaxes back toward VWAP; on a strong trend day price will ride the outer band for hours while VWAP trails far behind it. The band is a description of stretch, a piece of context, never a signal by itself. The detail of exactly how the reversion read fails on a trend day belongs to the reversion article, and this page leaves it there on purpose.
| Band | What it marks |
|---|---|
| VWAP (centre) | The volume-weighted mean: the fair-value reference the envelope is built around |
| ± 1σ | The inner band where the bulk of trading clusters: routine two-way rotation around value |
| ± 2σ | A meaningfully stretched deviation from the mean: uncommon, but common enough on a trend day that it is context, not a reversal cue |
| ± 3σ | A statistically rare excursion: usually an information event or a genuine break of equilibrium, not a deeper fade |
Read the bands as zones, not lines. The percentages a textbook attaches to one, two and three standard deviations assume a normal distribution. Real intraday returns on liquid Indian large-caps have fatter tails than the normal curve, so outer-band excursions occur somewhat more often than the textbook figure implies. The qualitative ordering holds, inner is routine, outer is stretched, but do not treat the bands as precise probability boundaries.
5. Reading it honestly, and the limits that bite
Put to work, this toolkit has two legitimate jobs and a set of honest caveats that decide whether it helps or misleads. The first job is as a reference for value. Session VWAP marks where today's business has transacted; anchored VWAP marks where a chosen cohort sits relative to break-even; the bands frame how far price has stretched from either. Read that way, the line is a map of who is comfortable and who is trapped, which is genuine context for a discretionary decision made on other grounds.
The second job is execution benchmarking. Session VWAP is the standard yardstick institutional desks use to judge fill quality: a buy filled below VWAP beat the day's average, a buy filled above it paid up. A serious retail trader can borrow the same discipline, comparing the price actually paid against the session VWAP at the time, to audit whether entries are systematically chasing. The line grades execution regardless of whether the trade itself worked.
Now the caveats, stated plainly because the omission of them is what turns a useful tool into a trap.
- Garbage anchor, garbage line. Anchored VWAP is only as good as its anchor. Choose a bar that marks no real change in psychology and the line describes a cohort that does not exist. There is no algorithm that rescues a bad anchor; the judgement is the tool.
- It is backward-looking. Every VWAP is a running average of trades that have already printed. It summarises the past accurately and forecasts nothing. It tells you where participants are, not where price is going.
- Trend rides the band and leaves VWAP behind. On a strong directional day price clings to one outer band while VWAP trails far below or above it. Any value or mean-reversion reading of the line fails exactly then, when the market is trending hardest. Recognising that regime before trusting the line is the difference between using the tool and being used by it, and the specifics of that failure are covered in the reversion article.
None of this is a reason to discard VWAP. It is a reason to hold it at the correct altitude: a volume-weighted reference frame that is powerful for reading value and benchmarking execution, subjective in its anchored form, and silent about the future. Used with those limits acknowledged, it is one of the most information-dense lines on a chart. Used as an oracle, it is one of the most reliable ways to get carried out.
6. Where this sits in the curriculum
VWAP appears at increasing depth across the programme, and this cluster mirrors that. The plain construction, why institutions benchmark to it, and how to read intraday displacement live in the VWAP indicator explainer. The intraday reversion setup, its regime filter, and the trend-day failure mode live in the reversion article. This page owns the reference-frame layer: anchored VWAP, the standard-deviation envelope, and the discipline of choosing an anchor. From there the natural next steps are the volume tools it pairs with, the volume profile deep dive for where volume concentrated by price rather than by time, and market depth in India for the order-book layer that VWAP summarises after the fact. The progression is deliberate: read the mean, then read the distribution around it, then read the book that produced both.
Learn the judgement upstream of the line
Anchored VWAP is only as good as the anchor you choose, and choosing it well is a reading skill, not an indicator setting. That reading skill is the spine of the curriculum: six stages, thirty volumes, from market structure to systematic execution, ₹14,999 to ₹1,49,999 with lifetime access and a 7-day refund window.
Take the free diagnostic →Frequently asked questions
What is the difference between VWAP and anchored VWAP?
+Standard VWAP resets at the session open, so its cumulation of price times volume starts fresh every morning and describes only the current day. Anchored VWAP uses the identical formula but starts the cumulation from a bar you choose, a swing high or low, an earnings gap, a breakout, the year's start, so it runs across days from that event. Session VWAP answers where today's business has transacted; anchored VWAP answers what everyone who traded since a specific event has paid on average.
Who invented anchored VWAP?
+Anchored VWAP was developed and popularised by Brian Shannon, CMT, of AlphaTrends. He set out the method in his 2023 book Maximum Trading Gains with Anchored VWAP and presented it to the CMT Association. The underlying volume-weighted average price is decades older and used across institutional execution desks, but the idea of anchoring that cumulation to a chosen market event, rather than the session open, is Shannon's contribution.
Why does anchored VWAP act as support or resistance?
+Because it marks the average price paid by every participant who has traded since the anchor event, it approximates the collective break-even of that cohort. When price falls back to an anchored VWAP drawn from a low, buyers from that move are near break-even and tend to defend the level, which can produce support. When price rallies back to an anchored VWAP drawn from a high, trapped buyers from that peak are near break-even and often sell into the recovery, which can produce resistance. It is a behavioural tendency, not a rule.
How do you choose the anchor for an anchored VWAP?
+You anchor to a bar that marks a genuine change in market psychology: a major swing high or low, an earnings or news gap, a decisive breakout bar, an index rebalance date, or a calendar reset such as year-to-date. The logic is that price action before that event reflects a different regime and should be excluded. The anchor choice is the whole art and the honest weakness: the same chart yields different anchored VWAP lines depending on where you start, so a poorly chosen anchor produces a line with no behavioural meaning.
What are VWAP standard-deviation bands?
+They are bands plotted at plus or minus one, two and three standard deviations of price around the VWAP line, forming a volatility envelope. The deviation is volume-weighted: a large move on heavy volume widens the band more than a thin-volume wick. They resemble Bollinger Bands but are centred on the volume-weighted average rather than a simple moving average. The bands measure how dispersed trading has been around fair value; they describe stretch, they do not by themselves signal a trade.
How are VWAP standard-deviation bands calculated?
+First compute VWAP as cumulative price times volume divided by cumulative volume. Then compute the volume-weighted variance as the sum of volume times the squared distance of each price from VWAP, divided by total volume, which is equivalent to the sum of volume times price squared over volume, minus VWAP squared. The square root of that variance is the standard deviation. Each band is VWAP plus or minus a chosen multiple, one, two or three, of that standard deviation, recomputed on every new print.
Does price touching the outer VWAP band mean it will reverse?
+No. An outer-band touch means price is a statistically stretched distance from the volume-weighted mean, nothing more. On a range day that stretch often relaxes back toward VWAP, but on a strong trend day price can ride the outer band for hours while VWAP trails far behind. Treating the outer band as an automatic reversal signal is the classic error. It is context about dispersion, not an instruction, and the reversion detail belongs to a separate discussion of trend-day failure.
Can VWAP be used across multiple days?
+Standard session VWAP cannot: it resets at each open, so carrying a single session's VWAP line across days is meaningless because the cumulation restarts. Anchored VWAP is the multi-day construct. By starting the cumulation from a chosen historical bar and never resetting, it produces one continuous line spanning weeks or months, which is precisely the gap in daily VWAP that anchoring was designed to close.
What are the main limitations of anchored VWAP?
+Three. First, it is only as good as its anchor: a poorly chosen starting bar yields a line with no behavioural meaning, so garbage anchor means garbage line. Second, it is backward-looking, a running average of what has already traded, not a forecast. Third, in a strong trend price rides one side and leaves VWAP behind, so a value or mean-reversion reading of the line fails exactly when the market is trending. It is a reference for context, not a standalone signal.
Sources
- Origin of anchored VWAP. Brian Shannon, CMT, Maximum Trading Gains with Anchored VWAP (AlphaTrends.Net Publishing, 2023), ISBN 979-8986868004, and his AlphaTrends writing, establish that anchored VWAP anchors the cumulation to a chosen event and reads as the cost basis of participants since that event. alphatrends.net
- Calculation and support/resistance. StockCharts ChartSchool states that anchored VWAP uses the same formula as traditional VWAP with only the included bars changed, that anchors are set to significant highs, lows, earnings or news, and that converging anchored VWAP lines can mark strong support or resistance. stockcharts.com
- Standard-deviation bands. Charting references such as the SierraChart Studies Reference define the VWAP standard-deviation bands and the volume-weighted variance, the sum of volume times price squared over volume minus VWAP squared, used to plot the envelope. sierrachart.com
- Price-data lag for education. A SEBI circular of May 2026 replaced the January 2025 three-month rule with a uniform 30-day lag for the use of price data in educational content, effective 1 July 2026; illustrative examples on this page use no live prices and name no securities.