MONSTER

A trading desk that teaches itself.

Seventeen agents learn from market, alternative and news data every night, must pass out-of-sample exams before they may trade, and are benched the moment they stop earning it. Everything below is the live paper record - unedited, updated hourly. How it works and the plan →

Return since 2026-10-21
+0.35%
Equity (paper)
$100,349
Drawdown from peak
-0.21%
Closed trades
25
Agents trading
2 of 16
Days live
17
Account equity, daily, since launch. As of 2026-11-08T16:02+00:00 UTC.

Agents

AgentStatusScoreSizeTradesReturn
mixed-low benched -0.81 0%4 +0.49%
mixed-medium benched -2.44 0%2 +0.00%
mixed-high benched -3.12 0%1 +0.75%
stocks-low trading -0.08 38%7 +0.23%
stocks-medium benched -0.43 0%6 +0.57%
stocks-high trading 0.66 93%3 -0.14%
crypto-low benched -3.77 0%1 +0.33%
crypto-medium benched -3.18 0%0 +0.00%
crypto-high benched -3.52 0%1 +0.86%
mixed-super council benched -2.93 0%0 +0.00%
stocks-super council benched -1.55 0%0 +0.00%
crypto-super council benched -3.68 0%0 +0.00%
rogue rogue benched -9.00 0%0 +0.00%
rogue-daily rogue benched -9.00 0%0 -%
rogue-contra rogue benched -9.00 0%0 -%
rogue-ranker rogue benched -9.00 0%0 -%

Score = out-of-sample Sharpe-style score on unseen data, averaged over the agent's own markets, 100 untraded S&P 500 names and a 20-year daily history. Below -0.25 an agent is benched and keeps learning without trading. Size = position size relative to full, rising with the score.

Follow the desk on your own paper account. Create a free Alpaca paper account, link it here, and the desk's trades are mirrored onto it at the risk level you choose - no real money, nothing to install. Create an account   How linking works →