Fortuna Coil fires only when a range break is confirmed by abnormal volume inside a compressed-volatility regime — the exact combination that survived when 1,800+ strategies didn't.
An automated research pipeline generated 1,800+ candidate strategies. Each one faced pre-registered parameters, lookahead audits, realistic costs, and data it had never seen. Almost everything died — including our own first two flagships. This survived.
Hypothesis, universe, and parameters locked in writing before any test ran. No tuning after peeking at results — the discipline that kills false discoveries.
An automated check confirms no future data leaks into signals. Every input uses completed bars only. The result never repaints.
Taker fees, slippage, and funding charged on every trade. ~24bp round trip — not the fantasy zero-fee backtests most sellers quote.
Tested only on data excluded from development — then tracked on a further unseen forward window. Sharpe held flat.
Anyone can quote a Sharpe ratio. We show the period, the universe, the cost model, and the shrunk forward expectation — because a seller who hides those is telling you something.
| Market & timeframe | Binance USDT-M perpetuals · top-10 alts · 1D bars |
| Out-of-sample window | Jan 2023 – May 2025 · forward-checked to Jul 2026 |
| Cost model | ~24bp round trip · taker + slippage · funding applied |
| OOS Sharpe ratio | ≈ 1.27 annualised |
| Profit factor / trades | 2.50 / 21 |
| Honest forward expectation | 0.4 – 0.7 Sharpe at ~2 signals / month |
Yes, one of those is a losing trade. Anyone selling you a system without losses is selling you a fantasy.
7-day money-back guarantee on every tier — email hello@fortunalabs.net, no questions asked.
No. Every input uses completed bars only — verified by an automated lookahead audit. Signals appear at bar close and never move.
The validated evidence is Binance crypto perpetuals on daily bars. It runs on any symbol and timeframe in TradingView, but we make no claims outside the tested setting — and we say so up front.
About two per month per chart at canonical settings. Deliberately selective. If you want more frequency, the honest path is re-validation on your market — not loosening the filters.
Because a strategy selected from a large candidate pool deserves statistical shrinkage — and because pretending otherwise is how this industry scams people. If a seller quotes their best backtest as "what you'll get," walk away.
No. This is software and educational material; it doesn't consider your objectives, situation, or needs. Trading involves substantial risk of loss. Full disclosure below.
7 days, no questions asked. Email hello@fortunalabs.net.
7-day guarantee · source included · lifetime updates