Most VWAP scalping guides teach one rule: above VWAP, only buy; below VWAP, only sell. It sounds clean, it is easy to code, and it throws away good trades. A study by Oyamori's quant desk of 19,452 one-minute structure breaks found the binary above/below VWAP rule made no measurable difference to results — the gate had no effect worth trading on. The problem is not VWAP. The problem is treating a fair-value line as a direction signal.

This guide uses VWAP the way it actually works: as a grading tool. Price action picks the direction, VWAP grades how easy the trip will be, and ATR sets the stop — with the exact M5 setup, the cost math, and a worked EUR/USD example.

A VWAP scalping strategy trades pullbacks to the session VWAP on M5: wait for price to pull back to the volume-weighted average price during a trending London or New York session, enter on the first M5 rejection candle in the trend direction, and place the stop 0.5 ATR beyond the wick. VWAP grades the trade; it does not decide it.

Key takeaways

  • VWAP is the volume-weighted average price paid since the session opened. It is a fair-value reference, not a direction signal.
  • The binary above/below rule failed a 19,452-trade test. Let price action pick direction; let VWAP grade the trade.
  • Trade the M5 VWAP bounce only on trend days, only in London or New York, and only when the distance from entry to VWAP is at least as large as the amount you risk.
  • Stop = 0.5 ATR beyond the wick. Target = 1.2R or the previous session extreme, whichever is closer.
  • A 0.7-pip round-trip cost on an 8-pip stop needs a 49.4% win rate to break even at 1.2R; on a 5-pip M1 target it needs 57%.

What VWAP is, in one minute

VWAP — the volume-weighted average price — is the average price of every trade since the session opened, weighted by volume. Big trades count more than small ones. On MetaTrader 5 the session VWAP resets at the session open (or at midnight, depending on the indicator setting) and draws one line that moves slowly through the day. Price above the line means the average buyer today is in profit; price below means the average buyer is underwater.

Three things VWAP does for a scalper: it filters bias loosely (a market holding above VWAP all session is one buyers are defending — context, not a signal); it marks the reversion point (price stretches away from VWAP and returns to it, because VWAP is where the most volume changed hands); and it gauges friction (if VWAP sits between your entry and your target, the trip is harder; if VWAP is the target, the trip is the trade).

What VWAP does not do: tell you whether the next candle goes up or down. A trending market can sit above VWAP for hours and still stop out every buyer who chased the first touch. The line grades trades; it does not take them.

Why the binary rule fails

The rule every beginner guide repeats — above VWAP only buy, below VWAP only sell — was tested by the Oyamori quant desk on 19,452 bearish one-minute structure breaks across 10 names over 13 months. Following it made no difference to results, and the study had enough power to detect an effect of 0.057R — a real null result, not a small sample.

The reason is simple. VWAP side is mostly a day-level variable: on an up day almost every signal sits above VWAP, on a down day almost every signal sits below it. The rule does not filter bad trades — it filters trades on days when the market already moves against the signal, which is exactly when a counter-trend scalp off VWAP can pay. It confuses where price sits relative to the daily average with where price is going next.

Replace the binary rule with a grading framework. Before every VWAP scalp, ask four questions:

  1. Which way are the peaks and dips pointing? Higher highs and higher lows mean longs; lower highs and lower lows mean shorts. This is the direction — VWAP does not vote.
  2. Am I at a clear level? The VWAP touch must land on a level marked before the session — a prior day high or low, the Asia range edge, a round number. A VWAP touch in mid-air is noise.
  3. Is there enough room before VWAP? The distance from entry to VWAP must be at least as large as the amount you risk. If the stop is 8 pips and VWAP sits 4 pips away, the trade is skipped.
  4. Is the session trending or chopping? If price has already crossed VWAP three times today, the line is a magnet for whipsaws. Skip the bounce and wait for a trend day.

A trade that passes all four is a VWAP bounce. A trade that passes three is a maybe. A trade that passes two is a coin flip with a magic line drawn on it.

The cost math first

A VWAP bounce is a small-target trade, and small targets die on costs. These are labelled assumptions for a typical ECN-style account — check your own costs before trading.

SetupStopTarget (1.2R)Round-trip costCost as % of targetBreak-even win rate
EUR/USD M5 bounce8 pips9.6 pips0.7 pip7.3%49.4%
EUR/USD M1 bounce5 pips6 pips0.7 pip11.7%51.8%
EUR/USD M1 tight scalp5 pips5 pips0.7 pip14.0%57.0%
XAUUSD M5 bounce$1.50$1.80$0.3418.6%55.6%
XAUUSD M15 bounce$3.00$3.60$0.349.3%50.5%

Round-trip cost is spread plus commission: 0.3 pip spread + 0.4 pip commission equivalent on EUR/USD, and a $0.30 spread + $0.35 commission per 0.1 lot on gold (about $0.34 per ounce). The break-even win rate is the win rate at which expectancy is zero after costs: net loss divided by net win plus net loss.

Two conclusions. First, the M5 bounce is a trade and the M1 tight scalp is a donation: the same 0.7-pip cost needs 49.4% on M5 and 57% on M1. Second, gold needs wider targets — a $1.80 target needs 55.6% to break even, a $3.60 target on M15 needs 50.5%. If your target is smaller than the cost allows, no entry rule saves you. Size the target to the instrument, not to your patience.

The setup: M5 VWAP bounce, step by step

Timeframe hierarchy: H1 for bias (is the day trending, and which way?), M5 for structure and the trigger (the rejection candle). Learn the setup on M5, where the cost table works, before touching M1.

Indicators: session VWAP (reset at session open), ATR(14) on M5, EMA 9/21 for trend confirmation. Nothing else.

Pairs and sessions: EUR/USD and GBP/USD in London (07:00–10:00 GMT) and the London–New York overlap (13:00–17:00 GMT). XAUUSD on M15 in the same windows. Never trade the bounce in Asia: volume is thin, VWAP is flat, and the "bounce" is usually the first touch of a range that will chop you.

The rules:

  1. Confirm a trend day. Price broke the Asia range cleanly and is holding on one side of VWAP. Higher highs and higher lows for longs; lower highs and lower lows for shorts.
  2. Mark your levels before the session — prior day high/low, Asia range high/low, round numbers. The VWAP touch must land within 2 pips of a marked level.
  3. Wait for the pullback. Price returns to session VWAP and prints a rejection candle on M5 — a wick at least twice the body, closing back in the trend direction.
  4. Check the room: entry to VWAP must be at least as large as your stop, and entry to the next structure level at least 1.2R away, or the target is too close to bother.
  5. Enter on the open of the next M5 candle after the rejection candle closes. Never enter on the wick — the wick is a promise, the close is the confirmation.
  6. Stop: 0.5 × ATR(14) on M5 beyond the rejection wick — 2.5 pips beyond the wick when ATR reads 5 pips.
  7. Target: 1.2R, or the previous session high/low if that is closer. Take the whole position off at the target — partials on a 1.2R trade are a second trade you did not plan.
  8. Risk 0.25%–0.50% per trade. Daily loss limit 2%. Three losses in a row: stop for the session.

Worked example — EUR/USD, London session

Example — not live prices. Assumptions are labelled so you can check the arithmetic.

Account: $10,000. Risk 0.5% = $50. Pair: EUR/USD. London, 08:30 GMT. ATR(14) on M5 reads 5 pips; session VWAP sits at 1.1020 and the marked Asia range high is 1.1022.

Price pulls back from 1.1045 to VWAP and prints an M5 rejection candle with a low of 1.1018 and a close of 1.1028. The grading questions: peaks point up, the touch lands on the marked Asia high, and the day is trending — but the wick is 10 pips, two full ATRs. The bounce already printed its profit before you could enter. Skip the trade. A wick that long means you are late, not early.

Second chance, 09:15 GMT. Rejection candle: low 1.1030, close 1.1038. Entry 1.1038, stop 1.1027 (1.1030 minus 2.5 pips), stop distance 11 pips, target 1.2 × 11 = 13.2 pips → 1.1051. The previous session high is 1.1055, so the target is reachable before structure.

Position size: lot size = dollar risk ÷ (stop distance × pip value). Pip value is $10 per 1.0 lot on EUR/USD. Lot size = $50 ÷ (11 × $10) = $50 ÷ $110 = 0.45 lots. Round down to 0.45.

Cost check: net win = 13.2 − 0.7 = 12.5 pips; net loss = 11 + 0.7 = 11.7 pips. Break-even = 11.7 ÷ 24.2 = 48.3%. The trade is worth taking above 48.3%.

The point is not the winning trade — it is the skipped one. The first pullback failed the grading questions because the wick was too long. A trader who takes every VWAP touch takes that trade, gets stopped out, and blames VWAP. The line did not fail. The grading did not happen.

When NOT to trade the VWAP bounce

  • Chop days. Price has crossed VWAP three or more times before midday. The line is a magnet now, not a reference. Switch to a range setup or stand aside.
  • Pre-news chop, Tuesday to Thursday. The 30 minutes before a red-folder release (CPI, NFP, FOMC, retail sales) are mean-reverting noise around multiple levels. A "bounce" there is the first touch of a whipsaw.
  • The first 15 minutes of London. VWAP has barely formed — it is an average of a handful of trades. Wait until the line has at least an hour of volume behind it.
  • When the room fails. Entry to VWAP is smaller than the stop, or entry to the next level is smaller than 1.2R. The trade is a bad price, not a bad setup.
  • When ATR is collapsing. ATR(14) on M5 under 2 pips on EUR/USD means the market is asleep. The bounce will print a 3-pip target and the spread will eat it.
  • Around scheduled news. Never hold a VWAP bounce through a release. The spike reprices VWAP itself, and the stop you placed 0.5 ATR away becomes a market order in a 20-pip gap. See our guide to trading US retail sales for the release-day rules.

Five mistakes and the fixes

  1. Trading every VWAP touch. Fix: the four grading questions. Three passes is a maybe, two is a skip.
  2. Using VWAP for direction. Fix: peaks and dips decide direction. VWAP grades the trip. The 19,452-trade test is the evidence.
  3. Fixed-pip stops. Fix: 0.5 ATR beyond the wick. A 10-pip stop is too wide for EUR/CHF in Asia and too tight for GBP/JPY at the London open.
  4. M1 before M5. Fix: learn the setup on M5, where the cost table works. M1 doubles the break-even win rate for the same setup.
  5. No cost check. Fix: run the break-even table before the session. If the target cannot carry the cost, the setup does not exist today.

Frequently asked questions

What is a VWAP scalping strategy?

A VWAP scalping strategy trades short pullbacks to the session volume-weighted average price, usually on M5. The trader waits for price to return to VWAP during a trending session, enters on a rejection candle in the trend direction, and uses an ATR-scaled stop. VWAP is a fair-value reference and a target, not a buy-or-sell signal on its own.

What is the best timeframe for VWAP scalping?

M5 is the best starting timeframe for VWAP scalping on forex, with M15 for gold. M1 doubles the break-even win rate for the same setup because the spread is the same while the target is half the size. VWAP only makes sense intraday, because it resets each session.

What are the best VWAP settings for scalping?

Use the standard session VWAP that resets at the session open, with no bands for the basic bounce setup. Add ATR(14) on M5 for the stop distance and EMA 9/21 for trend confirmation. VWAP bands (+1 and +2 standard deviations) are for the mean-reversion version — fading an overextension back to VWAP — not for the trend-continuation bounce.

Is VWAP good for scalping?

VWAP is useful for scalping as a grading tool and a target, but not as a direction signal. A 19,452-trade study found the binary above/below VWAP rule made no measurable difference to results. The edge comes from combining VWAP location with price action, ATR-scaled stops, and a cost check — not from the line alone.

How do you set a stop loss with VWAP?

Place the stop 0.5 × ATR(14) on M5 beyond the rejection candle's wick. On EUR/USD with a 5-pip ATR, that is 2.5 pips beyond the wick. Never use a fixed pip stop with VWAP: volatility changes through the session, and a fixed stop is right only by accident.

Does VWAP work on gold?

Yes, but on M15 rather than M5, because gold's spread is wider in dollar terms. A $1.80 gold target with a $0.34 round-trip cost needs a 55.6% win rate to break even; a $3.60 target on M15 needs 50.5%. Size the target to the instrument, and use the same grading questions as on EUR/USD.

Sources consulted

Educational material only — not financial advice. Forex and gold trading with leverage can produce substantial losses. Backtest and demo-trade any setup before risking real capital.