idea
What this journal does
Not trade calls. A notebook of which rules hold up and where they broke.
I pick names with chart rules, let the machine buy and sell, and I write up the seats that broke. The trading engine and the brokerage stay off this site. What you get here is method and a record.
Rule name: Breakout book breakout_rr, momentum book momentum_spec — used only in the tables below.
Philosophy
Chart setup → minimum reward/risk → many small stops, fewer large targets. Size by 1R (1% equity per trade). Scan a broad liquid universe rather than mega-cap only — breakouts are sparse on the largest names.
Layout (local only)
The trading engine, broker adapters, and live loops stay on the private automation host. This site publishes summaries of methodology, paper logs, and backtest notes only.
| Area | Role |
|---|---|
strategies/ |
Turtle, breakout, pullback templates |
rr_setup.py |
Closed pullback_rr / breakout_rr rules |
rr_backtest.py |
Historical simulation |
paper_loop |
Breakout paper state |
momentum_paper |
Parallel momentum bucket |
Two parallel paper books
Breakout (breakout_rr)
| Field | Value |
|---|---|
| Setup | breakout_rr |
| min_rr | 2.0 |
| Gates (default) | RS>SPY · sector>SPY |
| Risk | 1% equity / trade |
| Universe | ~150 liquid (S&P-style) |
| Max positions | 8 · 15% notional cap / name |
Momentum (momentum_spec)
| Field | Value |
|---|---|
| Setup | momentum_spec |
| Rules | close>SMA20>SMA50 · near 60d high or Donch20 · 5d≥+5% or vol≥1.5× · ATR%≤8% |
| Stop / target | ATR×2.5 · min_rr 1.5–2.0 |
| Pick | One top-ranked name per cycle in live-style runs |
| Alerts | Peak exit (−20%), gap −5%, SMA20 break |
What is not published here
- Full multi-strategy vote scanner loop
- Hermes / web UI integrations
- Live broker execution paths
- Raw signal JSON, portfolio state, or Telegram wiring