parabolic-short-trade-planner

---
name: parabolic-short-trade-planner
description: Use when the user wants a Qullamaggie-style Parabolic Short watchlist, pre-market conditional short plans with borrow/SSR gating, or intraday trigger evaluation from 5-min bars.
category: Technical
version: 1.0.0
tools: []
---

# Parabolic Short Trade Planner

Generate Qullamaggie-style Parabolic Short watchlists and conditional pre-market plans for US equities. This skill never sends orders — it produces analysis for a human to review against their own broker before entry.

## Three Phases

**Phase 1 — Daily screen.** Pull end-of-day bars and company profile data, apply hard invalidation rules (mode-aware — a safer large-cap mode vs a looser small-cap "classic" mode), then score survivors on five factors weighted 30/25/20/15/10: moving-average extension, acceleration, volume climax, range expansion, and liquidity. Assign an A/B/C/D grade from the combined score.

**Phase 2 — Pre-market plan.** Take Phase 1 candidates, filter to a minimum tradable grade (default B), check short-borrow availability, and evaluate SEC Rule 201 short-sale restriction (SSR) state from the prior day's close. Without live borrow data, mark every candidate watch-only pending manual borrow confirmation. Render three trigger plans per candidate: opening-range-low (ORL) break, first red 5-minute candle, and VWAP failure — each with an entry-hint and stop-hint formula rather than a baked-in price, since the trader computes exact levels at trigger time.

**Phase 3 — Intraday trigger monitor.** On a repeating cadence (typically every ~60 seconds during the first 30 minutes of the session, then every ~5 minutes after), walk each plan's trigger state machine forward using live 5-minute bars: `armed` → `triggered` / `invalidated`. When triggered, resolve concrete entry price, stop price, and share count from the plan's sizing formula. Each run replays the full session from open, so re-running mid-session reproduces the same state deterministically.

## Reviewing a Plan Before Entry

Check three things per ticker before acting:
- `plan_status`: `actionable` (manual gates can be cleared) vs `watch_only` (hard blocker — no borrow, or SSR active)
- Blocking reasons — must ALL be resolved before pulling the trigger
- Advisory reasons — heads-up only, e.g. manual locate required (always applies to shorts), too-early-to-short warnings, or a recent-earnings-catalyst warning when the last earnings report was within the catalyst window (default 10 trading days) — flag the move as event-driven rather than pure technical exhaustion

## Earnings Awareness

Two checks, both computed once per run: a hard exclusion when the *next* earnings date is within a blackout window (default 2 calendar days forward — do not short into an earnings gap risk), and a soft advisory warning when the *last* earnings date is within the catalyst window (default 10 trading days back).

## Scope

Do not use this for long-side momentum screening (that's a separate concern), for signals below 5-minute granularity (Phase 3 only evaluates 5-min bars), or as a live order router — it detects and sizes, but a human fires the trade.

## Mode Selection

Use the safer large-cap mode by default — it applies stricter market-cap and average-daily-volume floors and is the conservative starting point. Switch to the looser small-cap mode for classic small-cap blow-off tops, which relaxes market-cap/ADV floors and raises the 5-day rate-of-change threshold required to qualify, reflecting how much more violently small caps can move.

## Risk Management Context

Before treating any candidate as actionable, confirm short-borrow inventory is actually available (hard-to-borrow names may be unshortable or carry punitive fees) and check whether SSR is active — under Rule 201, once a stock drops 10%+ intraday, short-sale restrictions apply for the rest of that day and the next, meaning shorts can only be entered on an uptick. Both borrow inventory and SSR state materially change whether a plan is truly tradable, independent of how good the technical setup looks.

## Output

Phase 1: graded watchlist (A–D) with the five factor scores. Phase 2: per-candidate plan with the three trigger setups, plan_status, and blocking/advisory reasons. Phase 3: per-plan live state (armed/triggered/invalidated) with resolved entry/stop/shares once triggered.