9 EMA Strategy Backtest: NQ Futures, 5- and 15-Minute Bars, 2020 to 2026

00The verdict in three lines

• A YouTube video on the 9 EMA (“Master the 9 EMA: The Only Moving Average You Need?”, BKTraders) describes how prop traders use it: as a trend filter, as dynamic support with the VWAP on NASDAQ 5- and 15-minute charts, and as a 9/20 “break and hold” cross on 5-minute NASDAQ charts. We tested the last two, which have entries and stops, on NQ futures from 2020 to September 2026, at $15.76 a round trip.
• The retrace to the 9 EMA lost. On 5-minute bars it made +$1.40 a trade before costs and -$150,013 after, about 6 trades a day; on 15-minute bars it made money until 2022 and then gave it all back, ending at -$90,072.
• The 9/20 cross ended slightly positive, +$13,021, but 2022 made more than the whole result, it lost in validation and the holdout, and it did not beat the control. All three Buried.

01What we tested

The video's setups in our words. It gives the entries and the stops; it gives no profit target, so the exits below are ours, fixed before the run.

• Retrace to the 9 EMA with the VWAP (E9-VWAP): long when the previous bar closed above both the VWAP and the 9 EMA, the 9 EMA is rising, and the bar dips to the 9 EMA and closes back above it and above the VWAP; stop one tick below the VWAP, as in the video. Exit after the first bar that closes back below the 9 EMA (the video says a fall below it means the trend has changed), at the stop, or at the session close. Shorts: the mirror. Tested on 5-minute and 15-minute bars.
• 9/20 break and hold (E9-X20): long when the 9 EMA crosses above the 20 EMA on 5-minute bars, stop one tick below the 20 EMA, as in the video; exit at the opposite cross, which opens the opposite trade, at the stop, or at the session close. Shorts: the mirror.
• The trend filter (buy only above a rising 9 EMA, sell only below a falling one) is built into the first setup; on its own it has no entries or exits to test.
• All: New York session (09:30 to 16:00), session VWAP, EMAs of the session bars carried from day to day, entries at the next minute's open, one contract, flat at the session close. EMA crosses had already been tried in many variants on the site, so the development t-statistic bar is 2.5.

02Data and costs

NQ (E-mini Nasdaq-100) futures, NinjaTrader's back-adjusted 1-minute history with real contract volume, 1,735 New York sessions from January 2020 to 25 September 2026. Costs: $15.76 per round trip (commission and one tick of slippage on each fill), as in every NQ test on the site; a flip pays two. Periods: development 2020 to 2023, validation January 2024 to March 2025, holdout April 2025 to September 2026, already used by earlier NQ tests. Engine check: a self-test with hand-worked cases and an independent re-implementation that matched every trade of all three variants (e9_selftest_result.txt).

03Results

Dollars per contract. Gross is before costs.

Cumulative net dollars by month. * Holdout already used before this test.

By period

The 15-minute retrace and the 9/20 cross made money in development and lost in both later periods.

04Year by year, before and after costs

2022, a year of large daily moves in the Nasdaq, carried the 9/20 cross: +$85,748 that year against +$13,021 over the whole test.

05Long, short, and how the trades ended

Fewer than three trades in ten won in any variant. The 9/20 cross was stopped out in 69% of its trades: on 5-minute bars the 20 EMA sits close to the price at the moment of a cross.

06Is it the signal or chance?

The moved-entry control: every real entry moved by one random offset of whole sessions, at least 5% of them, to the same minute after the open in another session, same direction, same stop distance, then managed by the same exit and costs; 500 runs with fixed seeds. The bar is the 95th percentile of the average net result per trade.

07The gates

08Verdict by variant

Costs (5-minute retrace): slightly positive before costs, negative after. No proven edge: negative before costs (15-minute retrace), or slightly positive after costs but with no significance, a loss in validation and the holdout, and no advantage over the control (9/20 cross). Not tested: other exits, the overnight session, other markets.

TWhat this shows, and what to take into your own tests

What the results show

• Buying a retrace to the 9 EMA above the VWAP (and selling the mirror) made +$1.40 a trade before costs on 5-minute bars; with about 6 trades a day, $15.76 a round trip turned it into -$150,013.
• On 15-minute bars the same setup made money from 2020 to 2022 and lost from 2023 on, ending at -$90,072: a short backtest on the early years would have looked good.
• The 9/20 cross ended slightly positive (+$13,021 in almost seven years), but 2022 alone made +$85,748, it lost in validation and the holdout, and it did not beat the control.
• None of the three entries picked better moments than the same trades placed in other sessions.

What remains unknown

• Other exits: the video gives no profit target, so we fixed one (a close back through the 9 EMA, or the opposite cross). Other exits would be a new test.
• The overnight session and other markets (gold, EUR/USD), which the video also mentions.

Why it failed

A 9-period average on 5-minute bars follows price closely, so price touches it constantly and crosses it often. The retrace signal came about 6 times a day and made almost nothing before costs; the 9/20 cross caught a few large trend days, mostly in 2022, but its stop under the 20 EMA was so close that 69% of the trades were stopped out.

What you can apply to your own tests

• Popularity is not an edge: a level everyone watches can still be no better than chance once you trade it mechanically.
• Check every year, not only the total: one strong year (here 2022) can carry a whole result.
• If a source gives no exit, any exit you choose is part of the strategy; write it down before testing.

A note on the numbers: drawdown figures are measured on closed trades unless a table says it includes open losses. Open losses inside a trade are not counted in a closed-trade drawdown, so real account risk is higher, especially for systems with no stop. Backtests are not live trading.

FAQCommon questions

Does the 9 EMA strategy work?

Not in our backtest on NQ futures from 2020 to 2026. The retrace to the 9 EMA with the VWAP ended at -$150,013 on 5-minute bars and -$90,072 on 15-minute bars per contract after costs; the 9/20 cross ended at +$13,021, with losses in 2024 and 2025.

Is the 9 EMA a support and resistance level?

Price touches it often, which is why it looks like support on a chart. In our test, buying the touch above the VWAP and selling it below won 27% of the time and did no better than the same trades placed in other sessions.

How was the 9 EMA strategy tested?

On NQ futures 1-minute data from 2020 to 2026, built into 5- and 15-minute bars of the New York session, with $15.76 per round trip and rules written and dated before the data was run, costs included on every trade, results split into development, validation and holdout periods, a random-entry control as the benchmark, and every variant reported. Read the protocol.

09Check our arithmetic

Code provenance: our own code, written from the standard EMA and VWAP definitions and the setups as fixed in the specification. Nothing from the video's authors is used or published.

10Data notes

Holes. Removing the trades open across a known hole in the NQ file changes nothing: E9-VWAP-M5 1, E9-VWAP-M15 0, E9-X20-M5 2 trade(s) removed, every verdict unchanged.

Back-adjustment. The file is back-adjusted, which shifts older prices by a constant at each roll; the EMAs, the VWAP and the stops move with the prices, so results in points are unchanged.

Session. The EMAs use the regular session's bars only, as on a chart showing that session; a chart that includes the overnight hours would give slightly different EMA values at the open.

11Corrections

22 Oct 2026, navigation: added a link to the new guide comparing trading costs across all our tests. No result changed. 22 Oct 2026, navigation: added a link to the new guide comparing trend following and mean reversion across all our tests. No result changed.

+More tests and guides

• Vwap strategy backtest
• Ema 9 21 crossover backtest
• Ema 9 20 crossover backtest
• Ema ribbon strategy backtest
• Trend following vs mean reversion: this test next to the others of its style
• Free prop firm challenge calculator
• What all the backtests found so far
• The backtesting protocol
• How to backtest a trading strategy without fooling yourself
• Backtesting glossary