Public · Live · Immutable

Track record

Every thesis generated by our AI desk, publicly logged. Outcomes marked when targets hit, stops trigger, or theses expire. No retroactive edits.

See AI vs Buy-and-Hold benchmark

Win rate

all-time

54.5%

▲ 114W▼ 95L— 383N

Expected value / trade

+1.66%

math: WR × avg_win + LR × avg_loss

Profit factor

2.35

gross win / gross loss

Avg win

+5.28%

114 wins · max win not capped

Avg loss

-2.69%

95 stop-outs

Live calls · progress to target

All open theses

Live · 3 open calls tracked against mark price

HBARshortOn track
entry 0.08875target 0.08730
78% to target0% to stop+1.27%
SOLshortBuilding
entry 75.000target 71.500
-15% to target27% to stop-0.71%
VIRTUALshortBuilding
entry 0.73500target 0.69000
-21% to target27% to stop-1.28%

Equity curve · cumulative PnL%

peak+175.2%·final+135.5%

Monthly performance · since first thesis

3 months · 592 closed1 green2 redΣ +135.55%

Each cell is sum of per-trade % outcomes that closed in that calendar month (UTC). Cell intensity scales with magnitude of monthly PnL; tooltip on hover shows trade count + win rate. No retroactive edits — this is the same outcome timeline the equity curve plots, just rolled up.

Year	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Σ year
2026

By asset · where AI is good (and where it isn't)

30 assets with ≥1 closed thesis

Asset	Total	Closed	Win rate	WR bar	Avg PnL	Drill
INJ	39	8W/7L	53%	n=15	+1.69%

Conviction calibration

When the desk says conviction 70, do trades actually win 70%? This chart plots realized win rate per conviction band. Bubbles on or above the gold dashed line = well calibrated. Below = overconfident.

Total theses

595

generated since launch

Closed

209(35%)

hit target or stop

Expired neutral

383(64%)

didn't move enough — under tuning

Recent theses

Auto-refresh every 30s

Asset	Side	Conviction	Bull / Bear	Status	Outcome	PnL	When
HBAR	SHORT	68	25 / 76	active	open	+1.55%	06-03 07:47