Variations
Engine running · data updated 41s ago

← Dashboard

VWAP + EMA 9 crisscross

NAS100 · 3m chart · strategy s20_vwap_ema · 4 variations

Trades NAS100 on 3-minute candles when a fast 9-candle average crosses the day's average traded price (VWAP), and gets out when it crosses back.

What it is about

VWAP (volume-weighted average price) is the average price paid since the start of the session, with busier candles counting more. Big traders use it as the day's 'fair price'. Here it restarts at the 09:30 New York stock-market open. When a fast EMA 9 (an average of the last 9 candles, weighted to the most recent) moves above the VWAP, buyers have taken control of the day so far; below it, sellers have.

The strategy simply follows the cross and stays in until the EMA crosses back. Most crosses are noise and end in small losses (only about 1 trade in 5 wins), but on a strong one-way 'trend day' a single cross can run for hours. Looking back, the default version made all of its profit on trend days and lost on all other days.

Nobody knows in advance which days will trend, but the first one or two crosses after the open turned out to be a useful clue: later crosses happen mostly on choppy days. Variations C and D use that.

How it trades

Chart & market
NAS100 (the Nasdaq 100 index CFD) on 3-minute candles. Buys and sells by default.
Setup
VWAP of the typical price (average of high, low and close) weighted by tick volume (the number of price updates, the volume measure CFD data provides), restarting at 09:30 New York each day. EMA 9 of the closing price.
Entry
The EMA 9 crosses above the VWAP (it was at or below on the previous candle): buy at that candle's close. Crosses below: sell.
Stop loss
1 x ATR(14) from the entry (ATR = average candle size). Trades whose stop would be under 15 points (about 10 x the spread) are skipped.
Exit
At the close of the candle on which the EMA crosses back through the VWAP, or at the stop, whichever comes first. Always closed at the day's last candle (16:15 New York, 13:00 on US early-close days).
Filters
One trade at a time. Versions C and D only take the first 2 or the first 1 crosses after the 09:30 open.

NAS100 at any hour of the trading day (18:00 to 16:15 New York); the VWAP always counts from the most recent 09:30 open. Very active: about 60 trades a month for the default, 30 for buys only, 25 and 12 for the first-cross versions. Intraday only.

What to expect
  • Low win rate: 22-27% depending on the version. Expect long strings of small losses between the big trend-day winners.
  • Default, backtest 2019-today at zero cost: 5,577 trades, +856.0R, largest drawdown about 64.1R; every year positive.
  • Drawdowns are large compared with the risk per trade (32.7R to 64.1R, i.e. 33 to 64 losing trades' worth).
  • Trades last from a few minutes to most of a session; nothing is held past 16:15.
Risks and caveats
  • VWAP uses tick volume, which differs from broker to broker. On Pepperstone data (the paper-trading broker) the default made only about half as much as on the backtest data from July 2024 to Oct 2026, with only 48% identical signals.
  • The edge lives on trend days; a run of choppy months loses steadily.
  • The 'first N crosses' limit (C and D) was chosen by looking at 2019-26 results; paper trading has to confirm it.
  • The video's rules as stated lost money on every market, and the working version failed on US500 and gold: it is specific to NAS100.
  • With up to 60 trades a month, costs matter a lot: the results are before spread and slippage, so check your broker's spread in the calculator.

Where it comes from

From a trading video Victor shared on 2026-10-18 (S&P 500 options on a 3-minute chart). About 68,000 settings were tested; what held up was NAS100 on 3-minute candles with a VWAP restarting at the 09:30 open, a minimum stop and the 'exit when the EMA crosses back' rule, with 86% of neighbouring settings profitable in 2019-23.

Main variation: s20-D ★ Main

Live result
−1.8R
13 closed · 0 open · zero cost
Backtest total
+443.0R
1,136 trades · +0.39R a trade · win 27%
Backtest max drawdown
−32.7R
total ÷ drawdown 13.6
Years won
8/8
12.2 trades / month
Backtest R per year (zero cost), s20-D
0.0R50.0R100.0R150.0R20192020202120222023202420252026

The variations

Each variation is the same strategy with a different approach, trading its own paper account. Click one to see exactly what it changes.

s20-DFirst cross only★ Main

The strictest version: only the first cross after the 09:30 open, so at most one trend-day bet per day.

s20-ABoth directions, every cross

The default: every cross of the EMA 9 through the VWAP is traded, up or down, at any hour of the trading day.

s20-BBuys only

Only crosses above the VWAP are traded. NAS100 rose over most of the test years, so buying is the side that has worked more often.

s20-CFirst 2 crosses of the day

Both directions, but only the first two crosses after the 09:30 open. Early crosses tend to set the day's direction; later ones mostly come from choppy days that lose.

Variations side by side

Live result of all 4 together: −23.8R (zero cost).

Live paper account (R, zero cost)Backtest 2018 – today (R, zero cost)
s20-D★ MainVWAP + EMA 9 crisscrossNAS1003m−1.8R1323%–+443.0R+0.39R8/812.227%−32.7R
s20-AVWAP + EMA 9 crisscrossNAS1003m+4.4R8818%–+856.0R+0.15R8/859.922%−64.1R
s20-BVWAP + EMA 9 crisscrossNAS1003m−15.9R4418%–+517.3R+0.18R8/830.423%−58.5R
s20-CVWAP + EMA 9 crisscrossNAS1003m−10.7R2715%–+653.7R+0.28R8/825.424%−52.5R

Main = the variation with the most backtest R per R of max drawdown (2018 – today, zero cost). It is picked on that history, so the live accounts are the real test.