VWAP and Volume Indicators Tested on Gold and EURUSD: 1 of 28 Runs Passed, and It Was the Volume Spike, Not the VWAP Line

Type “vwap forex indicator mt4” or “forex volume indicator mt4” into Google and the suggestions keep coming, yet nobody ranking for them shows a test. VWAP is the line institutions are said to trade around, so the idea goes that price respects it. I built it, put four common rules on it and counted what they paid on the same bars I use for every test on this blog.

What is VWAP, and why does forex volume matter here?

VWAP is the volume weighted average price: from the start of the session, every bar’s typical price, (high + low + close) / 3, is weighted by how much traded on it. A bar with 3,000 contracts pulls the line ten times harder than a bar with 300. Bands around it are usually drawn at 1, 2 or 3 volume weighted standard deviations.

That needs real volume, and here the two markets split:

On MT4 the Volumes column for EURUSD is tick volume: how many times your broker’s price changed in that bar. It moves roughly with activity, but it is your broker’s count, not money traded, and it differs from broker to broker. So on EURUSD I tested VWAP as what it becomes without volume: the plain average price since the session start. On gold I used the real thing.

How did I test the VWAP rules?

• Cross: buy when a bar closes above VWAP after closing below it, sell on the opposite cross.
• Pullback: price closed above VWAP for 12 bars in a row, then a bar dips to touch VWAP and closes back above it. Buy. Mirror for sells.
• Band fade: a bar closes outside the 2 standard deviation band. Trade back toward VWAP, target VWAP itself, stop 1 ATR.
• Volume spike (gold only): bar volume more than twice the median of the previous 50 bars and a body bigger than half an ATR. Trade in the direction of the bar.

VWAP resets at 00:00 UTC. Two exits for each rule: stop 1.5 ATR and target 3 ATR closed after 48 bars, and a plain close after 12 bars. One trade at a time. Cost: $0.30 an ounce on gold, 1 pip on EURUSD. Gold data runs from June 16 to August 26, 2026: gold fell from $4,351 to $4,059 in the first half, touching $3,969 on July 17, then rose to $4,675 in the second. EURUSD from June 3. The 15-minute charts are built from the 5-minute bars. A rule passes only if it makes money after cost, makes money in both halves (before and after July 15), and beats the 95th percentile of 300 runs of random entries with the same exit and the same number of trades. The method is the same one behind my rule versus random entry test.

Which VWAP rules made money on gold?

Dollars per ounce. One of the two exits is shown per rule; all 16 gold runs are in the results file. Random p95: what the luckiest 5% of random runs with the same exit made.

The VWAP line itself did not give an edge in any form I tried. Crossing it lost on both charts. The band fade on M15 made +$0.99 an ounce and stayed positive in both halves, which looks good until you see that a random entry with the same target and stop made up to +$1.94. A rule that cannot beat random timing is not telling you anything about the market.

Why did the VWAP pullback lose the most?

Because of what the 12-bar condition selects. Price stays above VWAP for an hour on M5 mostly when gold is trending. The first touch of VWAP after that is often not a dip but the turn: the line has caught up because the move stopped. The pullback on M5 won 25.4% with a 1:2 stop and target, and lost $5.95 an ounce a trade, almost ten times worse than the median random entry at -$0.61. Its worst run of losses cost $958.70 an ounce.

With the 12-bar exit instead of the ATR stop it lost only $0.42, which says the entries were not so much wrong as badly timed for a tight stop. How much the stop size alone changes a result is the subject of my fixed versus ATR stop test.

What did the volume spike do that VWAP did not?

On gold M15 a bar with more than twice the usual volume and a real body was followed, on average, by more movement in the same direction over the next 3 hours. +$3.08 an ounce a trade after cost, +$1.14 in the first half and +$4.21 in the second, profit factor 1.41, worst run of losses $118.60 an ounce. That beat the 95th percentile of random runs, +$2.34.

Two honest limits. First, almost all of it came from buys: +$7.17 an ounce a trade on buy spikes, -$0.35 on sell spikes. Buy spikes paid in both halves, +$2.47 a trade on 22 trades while gold was falling and +$9.75 on 40 while it rose, but most of the money came in the rising half. Second, the same rule on M5 lost $0.87 with the ATR exit and $0.67 with the 12-bar exit. On the 5-minute chart a volume spike is too often a single news print that reverses.

When does gold actually trade?

This matters for VWAP. Anchored at 00:00 UTC, the line spends the Asian morning built from thin volume and then gets dragged hard by 13:00 to 15:00 UTC. A rule that treats VWAP at 06:00 and at 14:00 as the same thing is treating two different lines as one. My gold scalping session by session test found the same split in how far gold moves by hour.

How did EURUSD do?

Without volume, worse. All 12 EURUSD runs lost after the 1-pip spread. The cross on M5 lost 1.22 pips a trade on 396 trades; the band fade on M15 lost 1.26. The cross on EURUSD M15 was the only rule above its random band, at -0.08 pips a trade, and it still lost money. On EURUSD M5 the spread was larger than anything VWAP added, as it was for most rules in my spread and swap cost test.

What I would actually do with this

• Do not trade the VWAP line on its own. 0 of 24 VWAP line runs passed, on either market or chart.
• If you use volume, use it on gold or another exchange market, and on M15 rather than M5. The spike rule is the only thing here that passed, and only there.
• On EURUSD, treat the MT4 volume bars as an activity gauge. They tell you when the market is busy. They do not tell you where the money traded.
• Read VWAP by session. The line at 06:00 UTC and at 14:00 UTC is built from very different volume.

How to set VWAP up in MetaTrader and trade the volume spike step by step is in my VWAP indicator for MT4 guide. Whether VWAP is better than a plain moving average is in VWAP vs moving average.

Frequently asked questions

Does VWAP work in forex?

Not as a trading signal in my test. On EURUSD there is no traded volume to weight, and all VWAP rules lost after a 1-pip spread. On gold, which has exchange volume, none of the VWAP line rules beat random entries.

Is there real volume in forex on MT4?

No. Spot forex has no central exchange. MT4 shows tick volume, the number of price changes your broker sent in each bar, which differs from broker to broker.

What is the best VWAP strategy for gold?

None of the VWAP line rules passed. The rule that did pass used volume directly: a bar on gold M15 with more than twice the median volume of the last 50 bars, traded in its direction and closed after 3 hours, made +$3.08 an ounce a trade after cost.

Does the VWAP pullback strategy work?

It was the worst rule tested: -$5.95 an ounce a trade on gold M5 with a 1.5 ATR stop and 3 ATR target, 25.4% wins.

What time does gold have the most volume?

13:00 to 14:00 UTC in this data, around US data and the New York open: a median of about 1,000 contracts per 5-minute bar, against about 150 in the Asian morning.