67Sign in
67
quant.com
multi-strategypaper-firstwalk-forward validatedappend-only audited

Every alpha is 6-7.

Multi-strategy paper-first algorithmic trading. Backtested, walk-forward validated, append-only audited. Probabilistic by construction.

Open dashboardSee the bots
the bots

Five strategies, one ledger.

5 agents · paper-first
AI agent
Momentum / Trend
momentum

Trend-following on US large caps — daily, regime-aware.

Win
rate
Total return
+28.4%
Sharpe
1.42
Max DD
-12.1%
Trades
412
AI agent
Mean Reversion
mean_reversion

Short-window reversal on US ETFs — daily, oversold/overbought.

Win
rate
Total return
+14.6%
Sharpe
1.11
Max DD
-8.2%
Trades
287
AI agent
Congress Copycat
congress

Copies recent buys from allowlisted U.S. lawmakers.

Win
rate
Total return
+19.3%
Sharpe
0.94
Max DD
-15.4%
Trades
96
AI agent
News Sentiment (FinBERT)
sentiment

Long names with positive rolling news sentiment, intraday.

Win
rate
Total return
+22.1%
Sharpe
1.23
Max DD
-10.3%
Trades
538
AI agent
Cross-Sectional Momentum
xs_momentum

Long top-decile / short bottom-decile ranks on a US universe.

Win
rate
Total return
+31.2%
Sharpe
1.31
Max DD
-18.1%
Trades
642
Illustrative metrics. Live tear sheets, public T+ 1 delayed, will publish at bot.67quant.com.
01 / strategies

Run them in parallel.

Momentum, mean-reversion, congress trades, sentiment, cross-sectional — each with its own capital allocation, risk caps, schedule, and audit trail.

02 / research

An AI layer that proposes.

A three-agent pipeline mines public AI-trading content and writes structured findings into the same database the trader runs on. LLMs propose. Python disposes.

03 / safety

Three gates before live.

Paper-first by default. Two environment variables, ≥30-day Sharpe ≥ 1.0, and an orchestrator startup recheck stand between you and a real account.