Who else is building analysis pipelines outside of MT4/MT5? (Multi-Year Strategy Backtest Data)
I’ve spent the last few months moving my strategy analysis pipeline entirely outside of MT4/MT5 environments to get cleaner historical backtesting data. Most standard EAs look great on a 1-year curve but fall completely off a cliff when volatility shifts.
| Year | Starting Bal ($) | Ending Bal ($) | Max Drawdown (%) | Win Rate (%) |
|---|---|---|---|---|
| 2023 | 10,000 | 14,200 | 8.4% | 54.2% |
| 2024 | 14,200 | 21,800 | 11.1% | 52.8% |
| 2025 | 21,800 | 34,500 | 6.5% | 56.1% |
| 2026 (YTD) | 34,500 | 39,100 | 9.2% | 51.5% |
Here is the raw performance data of a custom liquidity-tracking script over a multi-year horizon, tracking drawdowns honestly:
The biggest issue I’m running into right now with python-based execution pipelines is handling broker-side latency and variable spreads during high-impact news drops on XAUUSD.
For those home-brewing your own standalone trading platforms or data scrapers: What are you using to accurately model real-time slippage and spread widening in your backtests? Are you relying on custom Claude/OpenAI scripts to clean your historical M1 data or handling it manually?