Anchored VWAP for Swing Trading: Anchors, Bands, and Bias

Video: Anchored VWAP Explained: My Favorite Indicator For Swing Trading
Creator: Verified Investing
Original Video: Watch the original video
Difficulty: Intermediate

The Big Idea

Anchored VWAP begins its volume-weighted price calculation at a selected event rather than resetting every session. Verified Investing demonstrates anchoring around meaningful highs and earnings-related candles, then observing the central average and deviation bands as potential reaction areas.

The useful question is what the chosen starting point represents. A different anchor produces a different reference, so anchor selection is part of the method, not a cosmetic adjustment. The source’s examples and personal preference do not establish a universal support-and-resistance edge.

What the Average Measures

A bar-based anchored VWAP is cumulative price × volume divided by cumulative volume from the anchor. With bar price p and volume v, AVWAP = sum(p × v) ÷ sum(v). The price source and aggregation interval affect the approximation.

For an original example, a $100 bar with 1,000 units and a $104 bar with 3,000 units produce a weighted average of $103, not the unweighted $102. This summarizes the recorded price-volume observations; it does not identify each holder’s cost basis or institutional inventory.

Session VWAP and anchored VWAP share the weighting idea but differ in their reset policy. A session-reset version can be unsuitable for a multi-session question; that does not make ordinary VWAP inherently meaningless. TradingView’s anchored drawing tool (paid link) lets the user select where calculation starts.

Choose an Anchor Without Retrofitting

The creator uses significant candles, such as an earnings-related high. Other research conventions might use a dated event, a known swing, or a gap. Define the rule and record when the anchor became knowable.

An earnings timestamp can be known at the event. A final all-time high may only be recognized later. Drawing an average back from that eventual high does not mean it was available as a live decision reference on the high’s opening tick.

Avoid dragging anchors until historical reactions look persuasive. Keep rejected anchors and failed reactions in the review. If the rule is “latest confirmed swing high,” include its confirmation delay and the moments when a later swing replaces it.

Interpret Bands as Measurements

The source discusses one-, two-, and three-standard-deviation bands and associates them with chop, ordinary reactions, and extreme movement. Those are his observational descriptions, not guaranteed behavior or measured probability coverage.

Band calculations depend on the implementation. A volume-weighted standard deviation of bar prices differs from an unweighted deviation, a return-volatility measure, or tick-based trade variance. Bands move as new observations arrive.

Being two deviations above an anchored average does not establish a 95% probability of returning. Market paths are not independent draws from a fixed normal distribution, and a persistent trend can stay outside a band. The source acknowledges that extension can continue, even though other phrases imply frequent or inevitable return.

Use the Level Inside a Complete Plan

Before a trade, specify whether the line is context, a trigger area, an invalidation reference, or a target. A touch alone is incomplete. A mean-reversion entry and a trend-continuation retest are different hypotheses.

For example, if an illustrative long entry is $102 with a $100 structural stop and a $106 target, nominal reward/risk is 2:1. An average at $103 does not make that trade valid; it may instead be an intervening reference. Define the actual reaction and execution rule before reviewing the result.

The video does not supply a complete universal entry, stop, sizing, and exit algorithm for every band. Do not infer one from a series of attractive chart bounces.

Pine Script: Manual Anchor With Transparent Bands

Show Pine Script — manual anchored VWAP

//@version=6 // Guide: https://candlechasers.com/anchored-vwap-swing-trading-anchor-bias/ // Volume-weighted moments of chart-bar HLC3, not tick-by-tick trade variance. indicator("CandleChasers - Manual anchored VWAP research", overlay=true) anchor = input.time(timestamp("01 Jan 2026 00:00 +0000"), "Anchor timestamp") source = input.source(hlc3, "Bar price source") bandMultiplier = input.float(2.0, "Band standard-deviation multiplier", minval=0.1) var int firstLoadedTime = time var bool started = false var float weight = 0.0 var float mean = na var float m2 = 0.0 bool begins = time >= anchor and not na(time[1]) and time[1] < anchor if barstate.isfirst and time == anchor begins := true if barstate.isconfirmed if begins started := true weight := 0 mean := na m2 := 0 if started and not na(volume) and volume > 0 and not na(source) nextWeight = weight + volume if weight == 0 mean := source m2 := 0 else delta = source - mean nextMean = mean + volume / nextWeight * delta m2 += volume * delta * (source - nextMean) mean := nextMean weight := nextWeight deviation = weight > 0 ? math.sqrt(math.max(0, m2 / weight)) : na plot(started and weight > 0 ? mean : na, "Anchored VWAP", color.blue, linewidth=2) plot(started and weight > 0 ? mean + bandMultiplier * deviation : na, "Upper research band", color.orange) plot(started and weight > 0 ? mean - bandMultiplier * deviation : na, "Lower research band", color.orange) var table status = table.new(position.top_right, 1, 1) if barstate.islast message = anchor < firstLoadedTime ? "Anchor precedes loaded history: load more data" : not started ? "Waiting for anchor bar to close" : weight <= 0 ? "Need positive reported volume" : "Confirmed-bar weighted-price bands; not trade signals" table.cell(status, 0, 0, message, text_color=chart.fg_color)

This original v6 indicator accumulates volume-weighted moments of a chosen chart-bar price from a timestamp. Its default source is HLC3, and its bands use weighted population standard deviation of those bar prices. This is a research implementation, not a claim of exact equivalence to every TradingView drawing-tool setting.

The first chart bar opening at or after the timestamp starts the calculation when it closes. Align the timestamp to the intended bar. If the timestamp falls inside a bar, that bar is excluded and the next opening qualifies. If the anchor predates loaded history, the script stays blank and requests more data rather than silently using an incomplete calculation.

Updates occur on completed bars, with no future pivots, higher-timeframe requests, or backdated trade signals. The open bar shows the last committed level rather than an intrabar calculation. Missing or zero-volume bars are skipped; that can make the result incomplete when the feed lacks activity. Use standard candles and the instrument’s appropriate volume data. Changing timeframe, session coverage, or the source can change the result.

This is an indicator and produces no Strategy Tester trades. Its chronological calculation does not prevent a human from choosing an anchor with hindsight. The code has not been compiled in TradingView here.

Validate Both the Math and the Trading Hypothesis

First check a small hand-calculated weighted example, then compare settings with the platform’s reference tool. In the two-bar example above, the weighted variance is 3 and deviation is approximately 1.732; a two-deviation band spans roughly $99.54–$106.46. That verifies this formula, not a profitable strategy.

For trading research, compare a fixed event anchor with a separately defined confirmed-swing anchor; compare central-line and band reactions under the same entry rules; and compare trending versus ranging periods. Include untouched anchors, failures, and sustained extensions. Keep costs and risk budgets consistent and reserve later data for evaluation.

Final Takeaway

An anchored average is only as interpretable as its starting rule and data. Select the anchor before judging the reaction, understand the band calculation, and test a complete trade model instead of assuming every excursion returns to the mean.

Risk Disclosure

This article is educational, not financial advice or a recommendation to trade or buy a service. Trading can cause substantial losses, particularly with leverage. Stops and limits do not guarantee fills or prices. Hypothetical examples and backtests depend on data, timing, costs, spreads, slippage, liquidity, gaps, market conditions, and discretionary assumptions. Historical performance does not guarantee future outcomes. Validate the risk controls and execution assumptions before considering live use.

CandleChasers independently interprets this source and is not affiliated with or endorsed by Verified Investing. Product promotions and performance opinions belong to the creator; the examples and indicator are CandleChasers’ own educational work.