Supertrend on MT5 vs MT4: MT5’s Extra M10 and M20 Charts Lost Where MT4’s M30 Made Money

My Supertrend settings test ran 16 combinations on EURUSD 5-minute bars and every one of them lost, even before the spread. It ended on a hint: the timeframe moved the result more than the settings did. That is where MT5 comes in, because the one practical thing MT5 gives a Supertrend trader that MT4 does not is more timeframes.

MT4 has 9 timeframes. MT5 has 21, adding M2, M3, M4, M6, M10, M12, M20, H2, H3, H6, H8 and H12. My MT4 vs MT5 comparison found that the MT5-only timeframes finished in the middle, never on top, for the rule it tested. This is the same question asked of Supertrend.

Is Supertrend different on MT5 and MT4?

The calculation is not. Neither platform ships Supertrend as a built-in indicator, so on both you install a custom version, and the good ones implement the same formula: a band of ATR times a multiplier around the bar’s midpoint, which flips when price closes through it. An MT4 file will not load on MT5 and vice versa, so you need the version written for your platform — installation for MT5 is covered in how to add indicators to MT5.

What does differ between versions is how the ATR inside it is averaged. MetaTrader’s own Average True Range uses a simple average of the true range, and many MT4 and MT5 Supertrend files call it. TradingView’s Supertrend uses Wilder’s smoothing, and some MetaTrader versions copy that. If your Supertrend and someone else’s disagree on the same chart, this is a common reason.

How were the timeframes tested?

EURUSD: 16,926 5-minute bars, June 3 to August 26, 2026, 1 pip per trade — the cost from my spread and swap test. Gold: 13,736 5-minute bars of COMEX gold futures, June 16 to August 26, 2026, $0.30 per trade, the same data as my gold session-by-session test. Each file was rebuilt into M10, M15, M20, M30, H1, H2, H3 and H4 by combining 5-minute bars that fall inside the same clock period, the way MetaTrader draws them.

The rule is the one from the settings test: Supertrend 10/3, always in the market, reversing on every flip, direction taken on the closed bar, ATR with Wilder’s smoothing as in the settings test. Any timeframe with fewer than 30 trades is marked and not ranked.

Which timeframe worked for Supertrend on EURUSD?

* Fewer than 30 trades — shown for completeness, not ranked.

From M5 to M30 every step up improved the result: -2.21, -1.49, -1.36, -0.22, +1.62. The two MT5-only timeframes, M10 and M20, sit exactly where their bar length puts them — between the MT4 timeframes on either side. Neither one found anything the next longer MT4 timeframe did not find better.

M20 is the interesting near-miss. It was positive before the spread at +0.78 pips per trade, and 128 trades gives it a real sample, but the 1-pip cost took it to -0.22. M30 cleared the cost with room to spare.

Then H1 broke the pattern and lost 8.67 pips per trade over 47 trades. Longer bars make Supertrend flip less, but each wrong flip on H1 costs far more, and three months is not long enough for a handful of H1 trends to pay for the reversals.

Is EURUSD M30 a fluke?

Possibly — 80 trades is a small base. Split into two halves on July 15, 2026, it held:

M30 was positive in both halves, though barely in the first. M20 lost in the first half and made money in the second, which is what a result without an edge looks like. Compare that with M5, where not one of 16 settings was positive in the second half.

Does the same hold on gold?

* Fewer than 30 trades — not ranked.

Gold makes the MT5 case weaker still. Of the ranked timeframes, the two that made money — M15 and M30 — are both on MT4. The two MT5-only ones, M10 and M20, both lost.

But gold’s winners are less solid than EURUSD’s M30. Gold M30 lost $5.59 per trade in the half when gold fell and made $11.56 in the half when it rose, on 26 trades each; M15 lost $2.91 and then made $4.24. Like every gold Supertrend setting in my gold settings test, the result followed the direction of the month.

Does ATR smoothing change the result?

The simple-average ATR flipped more often in every row — 46 to 72 more trades on M5 — most likely because a simple average drops each old bar out of the calculation all at once, where Wilder’s smoothing lets it fade. On M5 that meant more spread paid and a worse result each time. On M30 it went the other way: 6 more trades and a better result, +2.16 pips per trade against +1.62.

Over all EURUSD M5 bars, the two versions of 10/3 pointed in different directions 7.54% of the time. That is roughly one bar in thirteen where two traders with “the same” Supertrend see opposite signals. Before comparing results with anyone else, check which ATR your file uses.

What I would do on MT5 or MT4

• Do not switch platforms for Supertrend. The MT5-only M10 and M20 lost on both markets; the best timeframe on both was M30, which MT4 already has.
• Move Supertrend off M5. On EURUSD, M5 lost 2.21 pips per trade — only H1 did worse among the ranked timeframes — and M30 was the only positive one.
• Do not assume H1 is safer. EURUSD H1 lost 8.67 pips per trade — worse than M5.
• Know your ATR smoothing. It changed direction on 7.54% of bars and moved 14/4 from -1.41 to -2.18 pips per trade. If you also use ATR for stops, my ATR trailing stop guide covers that indicator.
• Treat gold results as the month’s direction until you have tested a falling market too.

For using Supertrend as a filter rather than a system, see my ADX vs Supertrend comparison and the Parabolic SAR vs Supertrend test. The original setup guide is my Supertrend indicator for MT4.

Frequently asked questions

Is Supertrend better on MT5 than MT4?

No. The formula is the same, and the extra MT5 timeframes did not improve it. With 10/3, EURUSD M10 and M20 — MT5 only — lost 1.49 and 0.22 pips per trade, while MT4’s M30 made +1.62. On gold, MT5’s M10 and M20 both lost while M15 and M30 made money.

What is the best timeframe for Supertrend?

In this test, M30 on both EURUSD and gold. EURUSD M30 made +1.62 pips per trade over 80 trades and was positive in both halves. Gold M30 made +$2.98 per trade over 52 trades but lost in the half when gold fell. EURUSD M5 lost 2.21 pips per trade, and H1 did worse still at -8.67.

Is Supertrend built into MT5?

No. Neither MT4 nor MT5 includes Supertrend among its standard indicators. You add it as a custom indicator, and you need the file written for your platform, because MT4 and MT5 indicators are not interchangeable.

Why does my Supertrend look different from TradingView?

Most often because of the ATR smoothing. TradingView uses Wilder’s smoothing; MetaTrader’s own ATR uses a simple average, and many MetaTrader Supertrend files use that. In this data the two versions of 10/3 disagreed on direction on 7.54% of EURUSD 5-minute bars.

Should I use Supertrend on M5 for scalping?

Not on EURUSD, on this evidence. Every one of 16 settings lost on M5, including before costs, and the default 10/3 lost 2.21 pips per trade. The same rule on M30 was positive.