EA 06
COSMOS
- Symbol BTCUSD
- Timeframe M5
- Strategy Type Japan Original Scalping
- Risk Level Low
TYO SCORE
71/100
Moderately evidenced
- Profitability 11.2/16
- Drawdown control 11.3/16
- Consistency 16/16
- Sample size 12.8/16
- Robustness 11.6/16
- Evidence quality 8/20
A measure of how well-evidenced this backtest is — not a prediction of future results. See the scoring model for how each component is calculated.
03 Performance snapshot
- Profit Factor 1.80
- Maximal Drawdown 19.80% (2,028.18 USD)
- Total Trades 4,144
- Win Rate 95.08%
- Sharpe Ratio 0.71
- Expected Payoff 0.47 USD
The same deal list, resampled 1,000 times with a recorded seed (20260820). This explores how fragile the backtest record is — it adds no information about the future.
Order shuffle — path risk
The compounded final return does not depend on trade order, so re-ordering isolates one question: how bad could the drawdown have been if the same trades had arrived differently?
- Observed (actual order)
- 8.5%
- Median of 1,000
- 3.9%
- 90th percentile
- 5.49%
- 95th percentile
- 6.04%
- 99th percentile
- 7.22%
- Worst of 1,000
- 8.33%
Missed trades
Each trade independently skipped with the stated probability — the offline-terminal / rejected-entry question.
Observed (actual order): Final return +19.3% · Max drawdown 8.5%
| Final return p5 | Median | p95 | DD p95 | |
|---|---|---|---|---|
| 10% missed | +8.7% | +18.0% | +21.8% | 8.88% |
| 20% missed | +5.4% | +16.1% | +21.8% | 9.1% |
Cost stress
Every trade return reduced by a fixed cost — a parameterised slippage-and-spread stress, not a calibrated broker model. The unit is this system’s own median absolute per-trade return (0%).
| Cost per trade | Final return | Max drawdown | |
|---|---|---|---|
| 10% of median trade | 0% | +18.3% | 8.5% |
| 25% of median trade | 0% | +16.8% | 8.51% |
| 50% of median trade | 0% | +14.3% | 8.51% |
Per-trade returns are fractions of the balance before each close; drawdowns are measured on trade closes, so intra-trade floating declines are invisible here.
Every simulation resamples the same historical record under stated rules. None of these numbers is a forecast.
No Shadow Forward evidence has been collected yet.
This site shows only records that exist as verifiable data. Nothing is estimated or simulated in their place.
When a verifiable record lands — a public signal page or an investor-access statement — it appears here and raises the Evidence Quality score.
Forward evidence has not yet been published for this strategy.
This site shows only records that exist as verifiable data. Nothing is estimated or simulated in their place.
When a verifiable record lands — a public signal page or an investor-access statement — it appears here and raises the Evidence Quality score.
Live evidence has not yet been published for this strategy.
This site shows only records that exist as verifiable data. Nothing is estimated or simulated in their place.
When a verifiable record lands — a public signal page or an investor-access statement — it appears here and raises the Evidence Quality score.
04 Equity curve
05 Drawdown analysis
- Reported maximum drawdown
- 19.80% (2,028.18 USD)Equity (includes open positions)
- Closed-trade drawdown
- 8.5%
- Worst window
- 2018-11-27 → 2018-11-28
- Recovered
- 2026-02-022624 days
- Longest losing streak
- 7
06 Monthly returns
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec | Year |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2018 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 0 | +4% |
| 2019 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | +2% |
| 2020 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | +1% |
| 2021 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | +2% |
| 2022 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | +2% |
| 2023 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | +1% |
| 2024 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | +3% |
| 2025 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | +3% |
| 2026 | 0 | 0 | 0 | Not published | Not published | Not published | Not published | Not published | Not published | Not published | Not published | Not published | +1% |
Each cell is that month’s return as a percentage of the balance it started with. Green is a gain, red a loss. Empty cells had no closed trades.
07 Yearly performance
| Year | Return | Total Trades | Win Rate |
|---|---|---|---|
| 2018 | +3.8% | 595 | 89.08% |
| 2019 | +2.1% | 425 | 92% |
| 2020 | +0.9% | 418 | 94.98% |
| 2021 | +1.6% | 353 | 98.02% |
| 2022 | +1.7% | 520 | 96.35% |
| 2023 | +1.5% | 552 | 94.2% |
| 2024 | +2.7% | 567 | 98.24% |
| 2025 | +2.8% | 569 | 97.36% |
| 2026 | +0.8% | 145 | 99.31% |
08 Trade statistics
Direction
- Long trades
- 2,036
- Short trades
- 2,108
- Closed trades
- 4,144
Outcome
- Win rate
- 95.08%
- Loss rate
- 4.92%
- Longest winning streak
- 253
- Longest losing streak
- 7
Size
- Average win
- 1.11
- Average loss
- -11.91
- Largest win
- 878.69
- Largest loss
- -241.69
- Payoff ratio
- 0.09
09 Risk profile
Observed in backtest
Measured from the trade record, not declared by the developer.
- Maximum concurrent positions
- 10
- Entries opened while another was live
- 645
- Closed by stop loss
- 100%
- Longest losing streak
- 7
Declared by the developer
Not yet published. Martingale, grid, averaging and position-scaling behaviour will be stated here once documented.
10 Parameters
| Parameter | Default | Description |
|---|---|---|
InpLots |
0.01 | |
maxposition |
100 | |
MaxSpread |
20.0 | |
Slippage |
300 | |
MM_ONOFF |
false | |
CODE3_ONOFF |
true | |
MagicNumber3 |
19877468 | |
CandleCountB |
3 | |
CandleCountS |
3 | |
Rikaku3 |
0 | |
Songiri3 |
0 | |
Trail_ONOFF3 |
true | |
TrailStartPips3 |
8 | |
TrailPips3 |
5 | |
Martin_ONOFF3 |
true | |
martin_point3 |
2.0 | |
Nanpin_ONOFF3 |
true | |
Nanpin_Pips3 |
45 | |
CODE_ONOFF3 |
true |
14 Research transparency
- Data Source
- BTCUSD
- Backtest Period
- 2018.01.01 – 2026.03.28
- Broker / Server
- Raw Trading Ltd
- Modeling Quality
- 15,160,026 ticks · 767,240 bars
- Initial Deposit
- 10,000 USD
Every figure on this page comes from the MetaTrader Strategy Tester report for this system. Nothing is modelled, smoothed or estimated. Fields the report does not state are left blank rather than filled in.
15 Evidence
Evidence
8/20
How far this system’s verification actually goes. A stage is marked verified only when the underlying data exists in this repository.
- ✓ Backtest Verified
- — Out of sample Pending
- — Walk forward Pending
- ✓ Monte Carlo Verified
- — Shadow forward Pending
- — Forward test Pending
- — Live Pending
The optimisation window of this system is not documented, so no period can be proven out-of-sample. The split-stability analysis in the Research Log is a stability check, not OOS validation.
Research timeline
AHow to access this EA
TYO has not assigned this strategy to an access plan. The research and evidence above are complete regardless; where the strategy is published on MQL5, that remains the way to obtain it.
The account and download layer is not live yet, so nothing is downloadable from this site.
17Disclaimer
Trading foreign exchange, CFDs and cryptocurrency carries a high level of risk and may not be suitable for every investor. You can lose more than your initial deposit. Leverage magnifies both gains and losses.
TYO develops and publishes automated trading software. We do not provide investment advice, portfolio management or trading signals, and nothing on this site is a recommendation to buy or sell any instrument.
No Expert Advisor, algorithm or optimisation guarantees profit. Any performance figure shown on this site is historical and does not guarantee, imply or predict future results.
Backtest results are simulations. Live results differ due to spread, slippage, commission, liquidity, execution speed, requotes, broker conditions and market regime changes.
You are solely responsible for evaluating whether any system is appropriate for your circumstances, and for any decision you take using it. Seek independent licensed advice where appropriate.