Backtest 1 Feb – 12 Oct 2026 · 41.6 M real ticks
Voltus EA is a fully automated trading system for MetaTrader 5. It watches, decides when each 5-minute candle closes, and risks the same amount on every trade.
01 · Architecture
The entity arranges itself in layers. Each one has a single job and never decides for the others.
Reads price tick by tick and waits for the 5-minute candle to close. It never acts mid-candle.
One decision per closed candle, systematic and with no discretionary input, long or short.
Recomputes size at the moment of every entry from current volatility. Initial stop, dynamic stop and a maximum time in the market.
02 · Intelligence
Voltus is not built on a clever prediction but on a method: the criterion is fixed before looking, and the code delivers the verdict.
Success threshold and predictions written before measuring, sealed with a hash.
The test set is set aside and used only once.
A function checks the values against the thresholds and returns PASS or FAIL.
Each decision stores the seal of the previous one, failures included.
The lab
The EA is what was left after trying to improve it in every reasonable way. Almost everything failed, and that is why we show it.
Tested
Survived and is in the EA
| What was tested | Variants | Result |
|---|---|---|
| Trade management | 100+ | The dynamic stop survives. No fixed profit target passed |
| Exit horizon | 150+ | A fixed time exit survives |
| Entry filters | 80+ | 0 survivors: they cut drawdown but cut profit more |
| Signal adjustments | 50+ | One was kept |
| Sizing and risk | 300+ | Equal risk on every entry survives |
| Concurrency and exposure | 20+ | A position cap per instrument survives |
| Market regime models | ~20 | None is used to trade; only to monitor |
| Automated search | 4,172 | 0 survivors |
| Machine learning | several families | 0: a random control did as well |
| Execution: EA vs. simulation | 73,096 decisions | No differences |
Overlapping windows inflated a sample 200 times and flipped the sign of a result. It was voided.
A time-unit error produced an absurd result. It was caught and fixed.
One finding was fully retracted: the real cause was a print format.
Measured on candles, a mechanism acting on a price level looked up to 300 times better than on ticks. Since then everything is validated on ticks.
Figures rounded down: per-category counts are approximate.
03 · Voltus EA
An Expert Advisor for MetaTrader 5 that follows short-term momentum on 24/7 synthetic indices. These are its rules and its evidence, each figure with its level.
One MetaTrader 5 Strategy Tester report, month by month. * October to the 12th.
BACKTEST · 1 FEB – 12 OCT 2026 · 100 % REAL TICKS
One instrument from the universe, fixed risk per trade and no compounding: profit does not grow with the balance. Each following tab is one month of this same report.
26,771 trade pairs compared one by one between the simulation and the program.
Dynamic-stop exits match in 47 % of cases; the gap is worth −2.4 % of profit and goes against the configuration, not in its favour.
BACKTEST · 1 FEB – 12 OCT 2026 · 41,579,848 REAL TICKS · 1 INSTRUMENT
MetaTrader 5 Strategy Tester report of the candidate configuration: 100 % real ticks, 2,000 USD initial deposit and fixed risk per trade.
A single instrument from the universe, with fixed risk per trade and no compounding: profit does not grow with the balance and does not depend on the deposit.
The full period includes the windows the system was developed on: it is not an out-of-sample test.
Result per trade is highest in February and March and more moderate from July; September and October (to the 12th) close negative. That is why it is shown month by month.
04 · Access
One plan, one price. The entity condenses: what remains is access.
Stripe confirms the subscription.
You receive your licence key and a personal, expiring download link by email.
Install the EA in MetaTrader 5 and enter the key; it binds to your account.
$299USD / month
14-day money-back guarantee: if it is not what you expected, we refund the first month.
Billed monthly in USD through Stripe. Cancel anytime. Past results, simulated or not, do not guarantee future results.
FAQ
MetaTrader 5 and an account with access to synthetic indices. The market is open 24/7, so we recommend a VPS. The program recovers its positions after a restart.
Yes. We show them on video on our social channels. This site publishes the MetaTrader 5 Strategy Tester backtests on real ticks, each with its period and conditions.
Because that is where the edge is. We publish the validation method, the figures with their conditions and what went wrong, which is what lets you judge the result.
It is the unit of risk: what you lose if a trade hits its initial stop. You decide how much money 1R is on your account.
Yes, and they are normal for this system. In the February–October backtest the hit rate was 41.49 % with a profit factor of 1.59: the average win (36.25 USD) is more than twice the average loss (16.12 USD). The worst streak was 24 losses in a row, for −444.62 USD.
In the February–October backtest it made 15,287 trades on one instrument, with an average duration of 21 minutes 52 seconds. It trades both directions evenly: 7,684 longs and 7,603 shorts.
Each licence covers 1 MetaTrader 5 account. More accounts need more licences, within the limit of 50.
If within the first 14 days it is not what you expected, we refund the first month and the licence is deactivated.
It can. Synthetic indices are generated by the provider and their behaviour has changed before. We monitor the edge continuously. There is no guarantee of future performance.
Contact
Questions about the licence, installation or the evidence. We reply by email.