Automated Telegram Signal Backtester

Problem

Manually backtesting Telegram signal providers was extremely time-consuming — each provider required opening TradingView, locating every signal from Telegram history, replaying the trade manually against price data, and recording the outcome. With dozens of signals per provider, this took 2–4 hours per provider, making it impractical to evaluate multiple providers quickly.

Old Way

Previously, the process was entirely manual: Open a signal provider's Telegram channel Read each signal one by one Switch to TradingView, find the exact date/time on the chart Visually check whether price hit TP or SL first Record the result in a spreadsheet manually This was slow, error-prone, and limited how many providers could realistically be evaluated.

New Method

I built the Telegram Signal Backtesting System (tg_backtester) — a full-stack automated tool. Here's how it works step by step: 1.Login — Authenticate with Telegram API credentials; session is saved for future use 2.Configure — Set target channels, trading symbols, SL/TP rules, and backtest parameters 3.Run — App connects to Telegram and pulls historical messages from selected channels 4.Parse — Messages are scanned for valid BUY/SELL signals with entry, SL, and TP. AI (Groq/OpenRouter) handles non-standard or unusual signal formats Backtest — Each valid signal is replayed against real historical price data via MetaTrader 5, determining whether TP or SL was hit first 5.Results — Dashboard displays win rate, total pips P&L, equity curve, and per-symbol breakdown. Every trade is logged in the History tab 6.Signal Brain — A separate module for manually pasting signals to test parser accuracy and train the AI on channel-specific formats 7Tech stack: Python backend (localhost:8000), React dashboard (localhost:3000), integrated with Telegram API, Groq (Llama3),

Impact

Time saved: Backtesting a signal provider dropped from 2–4 hours to minutes or seconds Scale: Multiple providers can now be evaluated in a single session instead of one per day Accuracy: Removes human error from manual chart-reading and result recording Faster decisions: Provider selection is now data-driven — win rate, pip performance, and consistency are immediately visible Better integrations: Faster evaluation means we can onboard higher-quality signal providers into our trading workflow much sooner