Active strategies — overview
Updated: June 2026. CryptoQuantix runs three quantitatively validated strategies on 4 years of multi-cycle data (Jun 2022 → Jun 2026: bear, bull and bear again), with backtests executed on the actual production code — not a reimplementation — using realistic costs (0.20% roundtrip) and no lookahead.
🔒 The complete operational specifications (exact entry/exit rules and validated parameters) are proprietary and available under a commercial license agreement — see license.
All strategies implement the BaseStrategy plugin interface
(scan / execute_entry / manage_positions) and receive an
injectable data provider: the exact same code runs live and in
backtest, eliminating by construction the entire class of "the backtest
does one thing, live does another" bugs.
The three strategies
| Strategy | Category | Validation (BTC, 4 years) |
|---|---|---|
| Trend Breakdown | Macro-gated two-sided tactical: breakdowns in bear phases, breakouts in bull phases | short +22 bps/trade PF 1.26 · long +68 bps PF 1.53 |
| Funding Squeeze | Contrarian deep-bear capitulation specialist (funding crowding) | +74 bps/trade PF 2.65 · ETH +64 bps PF 1.82 |
| Macro Core | Regime-following core position with disciplined exit and vol-targeting | +315%/4y vs +136% buy&hold, maxDD 24.7% |
Live instances: 5 (Trend Breakdown and Funding Squeeze on BTC+ETH, Macro Core on BTC). Each side of each strategy is active ONLY on the market and macro phase where it passed validation.
Portfolio (4-year equity simulation, all instances together): +491% with 21.5% maxDD, Calmar 2.61, worst year 0.0% — after adopting vol-targeting on the core position (tested and rejected on the data: fractional Kelly and drawdown de-risking).
Automatic 3-level gating
The system decides on its own what it is allowed to trade, and when:
| Level | Horizon | Mechanism |
|---|---|---|
| Macro | days-months | primary-trend phase filter: each side is structurally inactive in its adverse phase |
| Regime | hours | TREND/RANGE/COMPRESSION/EXPANSION classification with per-strategy rules |
| Performance | rolling | a strategy underperforming live gets disabled automatically |
Live guardrails: rolling profit factor below threshold after a minimum trade sample → disable and re-validate; drawdown beyond 1.5× the backtest maximum → disable immediately.
The validation process (why the numbers can be trusted)
Before deployment, every strategy passes a strict pipeline:
- Backtest on the real code — the data provider is injectable, the strategy code is identical between live and simulation
- 4 years, multi-cycle — the edge must survive bear, bull and transitions, not a single lucky phase
- Realistic costs (fees + slippage) and no lookahead
- Temporal in-sample / out-of-sample
- Robustness to neighboring parameters — no edge that only lives on one magic combination
- Minimum profit factor on the full sample
The rejected strategies (transparency)
The same pipeline rejected eight legacy strategies on the data — they remain in the codebase, disabled, with their verdicts documented:
| Strategy | 4-year verdict |
|---|---|
| Volume Breakout | PF 0.42-0.74 — no edge, negative in every phase and year |
| Mean Reversion (VWAP fade) | PF 0.28-0.53 |
| Liquidation Squeeze | by the time the cascade is visible, the move is over |
| Imbalance Scalp | fee-bound across 29k+ trades — zero gross edge |
| NY Brings | negative every single year (718 trades) |
| W/M Formation | non-structural edge (does not replicate cross-asset) |
| Smart Money | testable components already falsified |
| Iron Condor | options — out of project scope |
A system that publishes what does NOT work is a system whose method you can verify.
License and access
The code is source-available under a dual license: free for noncommercial use (including trading your own personal capital), while any commercial use requires a paid license — which includes access to the complete operational specifications of the strategies and the full validation reports.
Contact: lantoniotrento@gmail.com