Signal extraction, not screenshots
Messages are parsed into instrument, direction, entry or entry zone, stop-loss and targets. Follow-ups (“TP1 hit”, “move SL to breakeven”, “cancel the order”) are linked back to the signal they belong to.
Private beta · manually approved
Measure what a Telegram signal channel actually delivers.
Backtester.app reads the posts of a trading channel, turns its signals into testable trades, and simulates every entry, stop-loss and target against historical minute-level bid/ask candles. You get an honest report — including the signals that never became trades.
Registration is open. Access is activated manually on Telegram by @backtester_admin.
Most “signal performance” claims can’t be checked.
Win rates are quoted without the losing calls, the unfilled orders or the spread. This tool starts from the channel’s own messages and a fixed, published set of execution rules, then shows every outcome — including the ones that never became trades.
Historical candles, real spread
Trades are simulated on 1-minute bid/ask candles from Dukascopy. A long buys at the ask and sells at the bid; a short does the opposite. Slippage and commission are explicit settings, stored with each run.
Nothing is guessed
A message the parser cannot read confidently is reported for review, never traded. Statistics cover resolved trades only, and every unresolved bucket is shown next to them.
From a channel post to a report
The pipeline is deterministic and each step is inspectable in the finished report.
01
Fetch
Only the selected channel and only the chosen period, from 1 July 2025 onward.
02
Parse & link
Signals and their follow-ups are extracted, with explicit ambiguity flags.
03
Simulate
Entries, stops, partial targets and updates are applied to historical candles.
04
Report
Aggregate metrics plus a temporary trade-level report with an expiry.
What the report measures
Definitions are fixed before a run, so two channels are compared on the same basis.
Win rateWins ÷ resolved trades. Breakeven trades are reported separately.
ExpectancyAverage net result per resolved trade, in pips and in R.
Profit factorGross profit ÷ gross loss, after spread, slippage and commission.
DrawdownPeak-to-trough decline of the run’s equity curve, in R and account currency.
Target exitsMultiple targets split the position; each is closed independently.
Stop updatesBreakeven moves, partial closes and cancellations take effect from the update message.
Unresolved outcomesNot filled, cancelled, open at the end of data, out of range, needs review.
Data coverageThe exact date window, instruments and candle timeframe used are recorded.
Assumptions are stated, not hidden
Backtest results depend on execution assumptions. These are the ones this engine makes — and they are printed in every report.
Fills and gaps
Market orders fill at the next candle open. A gap through a level fills at the open, never at an invented price. Limit orders expire after the configured validity window and are reported as not filled.
One-minute resolution
If a stop and a target are touched inside the same minute, the intrabar order is unknowable. The run uses an explicit policy — conservative (stop first) by default — and reports how many trades it affected.
Costs
Spread is measured per minute from the bid/ask data. Slippage and commission are configured per run. Swap, broker execution quality and liquidity limits are not modelled.
Your channel’s content is not stored as a database
Raw messages and extracted signals exist only in an isolated per-job workspace while the backtest runs. They are deleted when the job finishes, fails or is cancelled, and a cleanup process removes anything left behind by an interrupted job.
What is kept
- Your account, activation status and permissions
- Usage counters for the daily beta limits
- Job status, aggregate metrics and report metadata
- Administrative audit records — never message text
What is deleted
- Raw Telegram messages and follow-up links
- Extracted signals and per-trade simulations
- Per-job databases, logs and temporary files
- The downloadable report, at its stated expiry
Beta access
Registration is open, but the service is not: every new account starts as pending and cannot run backtests until it is activated.
How to get in
- Register an account with your email.
- Message @backtester_admin on Telegram and mention the email you registered with.
- An administrator reviews and activates the account, usually by hand.
- Sign in again — the dashboard unlocks the backtest form.
There is no payment flow. Beta access is approved manually and can be revoked if the service is abused.
Default beta limits
| Item | Limit |
|---|---|
| Distinct Telegram channels | 1 per day |
| Historical months validated | 1 per day |
| Queued or running jobs | 2 per user |
| Concurrent jobs | 1 per user |
| Features | Standard deterministic parsing |
| Daily reset | 00:00 UTC |
Need more? Ask @backtester_admin for full access; limits are configured per account.
What this is not. Backtester.app is an independent research tool. It is not
investment advice, not a broker, and not affiliated with Telegram or with any channel it
measures. Backtests are historical estimates: they cannot account for future market
conditions, and a positive historical result is not a guarantee of future performance.