An MT5 account through our partner link
With gold, USDJPY and bitcoin. Tight spreads matter: the test above shows what wider ones cost.
Session Break · Gold, yen and bitcoin breakout EA for MetaTrader 5
Largest drawdown in this test 17.0%
MT5 Strategy Tester · 1 September 2022 – 31 October 2026 · 100% real ticks · fixed spread plus slippage on every fill · tested 5 November 2026.
A backtest, not a live account. Past drawdown is not a loss limit.
$47,267 at the end of Oct 26 (+373% since the start)
Our chart of the MT5 backtest from 1 September 2022 – 31 October 2026: balance at the end of each month at 0.5% risk per trade, from the report’s deal list, after realistic slippage. Positions grow with the account, so later months move more dollars. Backtest, not a live account; not a forecast.
How it trades
Session Break waits while gold, the yen and bitcoin each go through a calm stretch of their day, trades the side that breaks out and is flat again before the evening.
One position per market at a time. A losing trade is never added to, and size never grows after a loss.
Sent to the broker with the position, so it holds even when your terminal is offline.
Trades close on the day they open. Normally nothing is held overnight, and nothing over the weekend.
Gold, the yen and bitcoin each have their own quiet hours, so the trades spread across the day.
After an unusually large day on gold or the yen, Session Break skips the next breakout there.
Low, Medium or High: a fixed share of the current balance per trade. The size grows with the account and shrinks after losses.
The backtest
Bitcoin brought about 53% of the profit. All three markets ended positive.
One run of Session Break (RX_SessionBreak 1.00 (release file)) with gold, USDJPY and bitcoin on one H1 chart in the MetaTrader 5 Strategy Tester, mode “Every tick based on real ticks”, $10,000 deposit, leverage 1:500, 0.5% risk per trade from the current balance. Run by us on 5 November 2026.
Price history: real ticks. Gold and yen from IC Trading’s own feed since April 2026 and, before that, real ticks of BlackBull Markets and RoboForex; bitcoin from Binance. The spread is fixed for the whole period at the average we measured at Session Break’s own entry and exit times (gold $0.168 over 1,823 trades, yen 0.0117 over 1,572), commission included; bitcoin $6.
The tester fills orders without slippage. We subtracted slippage from every entry and every exit of the same trades: gold $0.30 per ounce, yen 0.005, bitcoin $6. All figures on this page are after that deduction; the original MT5 charts below are before it.
Monthly and yearly results are the change of the balance, rebuilt from the report’s deal list. A backtest is not a live or demo-forward result. Different brokers, deposits and start dates produce different results.
Risk & spread
Three risk levels, three spreads, nine separate tester runs. Higher costs cut the result, they do not break it.
As tested, Low: +45% a year, largest drawdown 17.0%, Monte Carlo 95% 25.2%.
Tester run: 2026-11-05-sb100-fix1-r05-2209
Every result is its own MetaTrader 5 tester run on the same real ticks, 1 September 2022 – 31 October 2026, $10,000 start, tested 5 November 2026. Each spread is fixed for the whole period. “As tested” is the average spread we measured at the moments Session Break actually trades (gold $0.168, yen 0.0117), not a quiet-hours best case. On top of the spread every entry and exit pays realistic slippage: gold $0.30 per ounce, yen 0.005, bitcoin $6. Where a level asks for less than the broker’s smallest lot, Session Break skips that trade instead of risking more, so the 0.5% runs have a few trades fewer. The risk is a share of the current balance, so sizes grow with the account; at the high level the totals reach position sizes where real fills would likely be worse. Backtests, not live accounts; past drawdown is not a loss limit.
| Spread | Risk per trade | A year | Largest drawdown | Monte Carlo 95% | Profit factor |
|---|---|---|---|---|---|
| As tested | 0.5% per trade | +45% | 17.0% | 25.2% | 1.28 |
| As tested | 1% per trade | +106% | 31.3% | 44.2% | 1.29 |
| As tested | 2% per trade | +169% | 53.4% | 65.7% | 1.3 |
| 1.5 × | 0.5% per trade | +34% | 16.8% | 27.9% | 1.22 |
| 1.5 × | 1% per trade | +76% | 30.9% | 49.0% | 1.24 |
| 1.5 × | 2% per trade | +143% | 53.5% | 69.9% | 1.29 |
| 2 × | 0.5% per trade | +32% | 16.0% | 28.4% | 1.21 |
| 2 × | 1% per trade | +69% | 31.7% | 50.3% | 1.22 |
| 2 × | 2% per trade | +132% | 57.7% | 72.3% | 1.26 |
Limits
Read these before you choose a risk level.
Every level a separate tester run, slippage on top.
About 53% of the profit. Our bitcoin ticks start in September 2022, so earlier years are not tested.
Weakest calendar year in this test: 2022 with +3.6%. Several months in a row can end below the previous high.
No live or demo-forward record yet. Start on demo and compare your own fills with the test.
MT5 report
The charts MetaTrader 5 saved with the report of the 0.5% run. They show the tester’s raw fills; the figures above subtract slippage from them.
Saved by MetaTrader 5 with the report, before slippage. The x-axis counts deals, not days.
Session Break runs on desktop MT5. You choose the account and keep the terminal connected on weekdays.
Talk through the requirementsWith gold, USDJPY and bitcoin. Tight spreads matter: the test above shows what wider ones cost.
Gold’s smallest position sets the floor: about $4,500 at 1% and $2,500 at 2%, at today’s gold price.
Algo Trading on, on one chart only, terminal connected Monday to Friday. Hosting is a separate cost. The phone app is for monitoring.
Your next step
Tell us your risk level and broker. You get the partner link and we answer your questions first.
Open an MT5 account through that link and send us your account number.
We send you the EA and go through the setup with you. Any chart, Algo Trading on.
Direct message to @RoboXpertAI
Message us on Telegram. You open an MT5 account through our partner link and send us the account number; we set it up with you and send you the EA.
No licence fee: access runs through our partner-broker arrangement, and we receive broker rebates on referred trading activity. Spreads, commissions, swaps and VPS hosting can still cost money.
No. One position per market at a time, each with its own stop-loss from the start. Session Break never adds to a losing trade and never increases the size after a loss.
Normally not. Every trade is closed on the day it opened. Bitcoin can stay open overnight on a few US holidays when the gold market closes early.
Start with 0.5%. In the test it gave the smallest drawdowns. 1% earns more but roughly doubles the swings (largest drawdown 31.3%); 2% reached 53.4% at its worst point.
It depends on the risk level, because gold’s smallest position is fixed. At today’s gold price that is about $8,500 at 0.5%, $4,500 at 1% and $2,500 at 2%. Below that Session Break skips gold trades it cannot size small enough, rather than risk more.
Not yet. The figures on this page are backtests with real ticks and fixed spreads. A backtest is not a forecast.
Versions
Newest first. The EA you get is always the newest version.