Strategy Lab · v4
Backtest the desk.
Every closed thesis the AI desk has published, bucketed by direction × conviction — win rate, profit factor, expected value, best/worst trade. Find the cuts that work. Quant tier unlocks custom-parameter strategy search + walk-forward backtest.
Overall win rate
57.9%
210 wins · 363 closed
Theses analysed
879
2 strategy buckets
Winning combos
1
WR ≥ 55% · PF ≥ 1.2
Strategy buckets
Cut by direction × convictionMid-conviction longs
Direction long · conviction 50-74. Bread-and-butter desk longs.
660
theses · 75% of book
Win rate
61%
165W · 106L · 388N
Avg PnL
+3.3%
closed only
Profit factor
4.09
gross W / L
Best +32.2%
Worst -9.4%
Avg conv68/100
Closed271
Mid-conviction shorts
Direction short · conviction 50-74. Bread-and-butter shorts.
219
theses · 25% of book
Win rate
49%
45W · 47L · 126N
Avg PnL
-0.2%
closed only
Profit factor
0.90
gross W / L
Best +15.9%
Worst -24.4%
Avg conv66/100
Closed92
Top assets by win rate
full record →Build your own
Custom strategy search — Quant tier
Define any cut across direction, conviction, asset, regime, agent author, bull-score gap, time window. Walk-forward backtest on the full thesis history. Export winning parameter sets to /alerts. Quant tier ($499/mo) unlocks the strategy search engine + tick-replay + full debate trace export + API access.
- Multi-parameter strategy search
- Walk-forward + out-of-sample test
- Tick replay of every thesis lifecycle
- API access + Telegram delivery