Subscribe
EN

Backtest

Test trade messages or a built-in strategy on real past prices, with the luck test and the settings finder. No sign-in needed: your backtests are kept for this connection, five a day. With a subscription, the backtest also replays your own channels with your own account's rules. Subscribe or open your copier.

Backtest trade messages

Pasted messages are read like a channel with the default settings; a message is one line or several, starting with its time, for example "2025-09-01 14:03 BUY XAUUSD SL 3600 TP 3625 or as Telegram copies it: Gold VIP, [01.09.2025 16:03] BUY XAUUSD SL 3600 TP 3625". The messages are replayed on real past prices as the copier would have traded them. Buys fill on the ask, sells on the bid; inside a bar the stop is hit before the target; a gap fills at the price, not the level. Every result says where its prices came from and how much of the period they cover; a period without past prices says so instead of showing a result. "Find the best settings" tries up to 24 settings (targets, break-even, hours, stop width) on the first 80 % of the messages, checks the chosen one on the last 20 % and counts the luck of the search.

Backtest a built-in strategy

The settings each strategy can take
  • asia_range_breakout: buf = 0.0 / 0.5, last_entry = 10:00 / 12:00, exit = 16:00 / 20:00
  • asia_sweep_reversal: stop = 0.5 / 1.0, last_entry = 11:00 / 15:00
  • asian_session_meanrev: n = 8 / 16 / 32, k = 1.5 / 2.0
  • comex_open_range_breakout: buf = 0.0 / 0.25, exit = 13:30 / 16:00
  • donchian_h1: n = 24 / 48 / 96 / 192, exit_div = 2 / 4
  • ea_inside: no settings
  • ea_london: no settings
  • ea_orb: no settings
  • ea_pdhr: no settings
  • ea_trend: no settings
  • ea_volbrk: no settings
  • ema9_vwap_atr: chan = 1.5 / 3.0, sl = 1.5 / 2.5, tp1 = 1.0 / 2.0, trail = 2.0 / 3.0
  • equity_stress_haven: k = 2.0 / 3.0
  • h4_trend_m15_pullback: fast = 10 / 20, dip = 25 / 35
  • intraday_seasonality: start = 18:00 / 19:00 / 20:00, end = 02:00 / 03:00 / 08:00, short_leg = False / True
  • london_to_ny_momentum: k = 0.0 / 0.5 / 1.0, exit = 16:00 / 20:00
  • m15_shock_reversal: k = 2.5 / 3.5 / 4.5, hold = 1 / 2 / 4 / 8
  • macro_residual_reversion: z = 1.0 / 1.5
  • metals_dollar_confirm: z = 1.5 / 2.0, hold = 120 / 240
  • news_0830_momentum: k = 1.0 / 2.0 / 3.0, hold = 2 / 4 / 8 / 16
  • nr7_break: rr = 2 / 3
  • ny_session_long_control: no settings
  • one_model_ridge: h = 4.0 / 8.0
  • open_range_expansion: k = 0.5 / 0.7, sl = 0.4 / 0.6
  • overnight_carry: rr = 2 / 3
  • overnight_carry_strong: k = 1.5 / 2.5
  • pdh_breakout_regime: hold = 32 / 64, k = 1.0 / 2.0
  • pm_fix_reversal: k = 0.5 / 1.0 / 1.5, hold = 2 / 4 / 8
  • prev_day_extreme_fade: hold = 4 / 8 / 16 / 32
  • range_bollinger_fade_h1: n = 24 / 48, k = 2.0 / 2.5, er_max = 0.2 / 0.35
  • silver_leads_gold: z = 2.0 / 3.0
  • squeeze_breakout_h1: q = 0.5 / 0.65 / 0.8, hold = 6 / 12 / 24
  • tsmom_daily: lookback = 10 / 20 / 40 / 60, mode = long_short / long_flat

The strategies are the research desk's own rule sets, each with the idea it tests; most were rejected by the judge, and the page says which. They run on the reference bars (Baroq data, up to the end of 2025): a position decided at a bar's close fills at the next bar's open, buys on the ask and sells on the bid, with slippage and commission. "Find the best settings" tries the strategy's own settings, chooses on the first 80 % of the bars and checks once on the last 20 %.

Your backtests

No backtest yet.

One backtest runs at a time; the last 20 are kept. A result is never a promise: it is what the rules would have done on those prices.

Demo accounts only by default: copying to a real-money account stays blocked unless the site owner enables it. Results shown here are simulated and hypothetical (fills at the stated price, no spread, slippage or gaps), so a real account would do worse. Nothing here is investment advice.