stockbee-20pct-study

Stockbee 20% Study

Build a daily event study of US equities that moved +20% or -20% over a defined window. Convert large movers into structured study records, classify the catalyst and chart context, update forward outcomes, and summarize recurring patterns for research.

This skill is a research, model-book, and setup-fluency workflow. It does not generate buy/sell signals, place orders, or output broker execution instructions.

When to Use

• User wants to run a Stockbee-style daily 20% mover study
• User asks which stocks moved +20% or -20% today, this week, or over a configurable lookback window
• User wants to backfill historical 20% movers and study what happened next
• User wants to identify continuation, reversal, exhaustion, or theme-cluster patterns
• User wants to build a model book of explosive winners, major failures, and failed low-quality pops
• User wants edge hints for downstream strategy research rather than immediate trade signals

Prerequisites

• Python 3.9+
• FMP API key for live US universe scans, or offline OHLCV JSON via --prices-json
• Optional structured news/catalyst JSON for higher-quality catalyst classification
• Recommended market regime artifact from market-regime-daily
• Recommended local state path: state/stockbee/20pct_study_events.jsonl

Workflow

Step 1: Scan for 20% Movers

Run after the US market close, or against the latest complete daily bar in an offline OHLCV file.

python3 skills/stockbee-20pct-study/scripts/run_20pct_study.py scan \
  --fmp-universe \
  --max-symbols 300 \
  --as-of 2026-06-28 \
  --lookback-days 5 \
  --min-abs-return-pct 20 \
  --min-price 5 \
  --min-dollar-volume 20000000 \
  --include-down-movers \
  --state-file state/stockbee/20pct_study_events.jsonl \
  --output-dir reports/

Use offline data instead of FMP:

python3 skills/stockbee-20pct-study/scripts/run_20pct_study.py scan \
  --prices-json data/us_daily_ohlcv.json \
  --as-of 2026-06-28 \
  --lookback-days 5 \
  --include-down-movers \
  --state-file state/stockbee/20pct_study_events.jsonl \
  --output-dir reports/

Step 2: Enrich and Classify Events

Use structured catalyst data when available. The enrichment step is best-effort: if no news record is found, the event remains a price-only NO_CLEAR_NEWS study record.

python3 skills/stockbee-20pct-study/scripts/run_20pct_study.py enrich \
  --events-json reports/stockbee_20pct_events_YYYY-MM-DD_HHMMSS.json \
  --news-json data/catalysts_YYYY-MM-DD.json \
  --market-regime reports/market_regime_latest.json \
  --state-file state/stockbee/20pct_study_events.jsonl \
  --output-dir reports/

Step 3: Update Matured Forward Outcomes

Update 1-day, 3-day, 5-day, 10-day, and 20-day forward outcomes after enough future bars exist.

python3 skills/stockbee-20pct-study/scripts/run_20pct_study.py update-outcomes \
  --prices-json data/us_daily_ohlcv.json \
  --state-file state/stockbee/20pct_study_events.jsonl \
  --horizons 1,3,5,10,20 \
  --output-dir reports/

The update records close return, MFE, MAE, direction-adjusted continuation return, and outcome tags.

Step 4: Summarize Cohorts

python3 skills/stockbee-20pct-study/scripts/run_20pct_study.py summarize \
  --state-file state/stockbee/20pct_study_events.jsonl \
  --group-by direction,catalyst.label,technical_context.pattern_label,technical_context.close_quality \
  --min-sample 10 \
  --output-dir reports/

Treat rule_candidates and exported edge hints as research prompts. Require representative chart review, sample-size thresholds, and out-of-sample validation before changing trade rules.

Step 5: Historical Backfill

python3 skills/stockbee-20pct-study/scripts/run_20pct_study.py backfill \
  --from 2020-01-01 \
  --to 2026-06-28 \
  --prices-json data/us_daily_ohlcv.json \
  --min-abs-return-pct 20 \
  --include-down-movers \
  --state-file state/stockbee/20pct_study_events.jsonl \
  --output-dir reports/

Backfill records are marked CURRENT_UNIVERSE_BACKFILL_SURVIVORSHIP_BIAS by default. Add --survivorship-complete only when the supplied OHLCV includes delisted symbols and historical universe coverage.

Output Format

• stockbee_20pct_events_YYYY-MM-DD_HHMMSS.json — scan metadata and event records
• stockbee_20pct_daily_report_YYYY-MM-DD_HHMMSS.md — human-readable daily 20% study report
• stockbee_20pct_enriched_YYYY-MM-DD_HHMMSS.json — enriched event records
• stockbee_20pct_outcome_update_YYYY-MM-DD_HHMMSS.json/md — matured forward outcome update
• stockbee_20pct_cohort_summary_YYYY-MM-DD_HHMMSS.json/md — cohort statistics and rule candidates
• stockbee_20pct_edge_hints_YYYY-MM-DD_HHMMSS.yaml — edge-hint export for downstream research skills
• state/stockbee/20pct_study_events.jsonl — durable 20% mover model book

Resources

• references/methodology.md — 20% study methodology and review checklist
• references/event_schema.md — JSONL event record schema
• references/catalyst_taxonomy.md — catalyst and risk label definitions
• references/scoring_system.md — event quality and study priority scoring
• references/cohort_mining_rules.md — overfitting controls and sample-size rules
• scripts/run_20pct_study.py — CLI for scan, enrich, update-outcomes, summarize, and backfill

Shared FMP client

Live FMP access uses the bundled generated scripts/fmp_client.py, loaded only when an FMP run starts, so offline modes still work without requests. CLI key/environment precedence, the API-call budget and successful output shapes are preserved. Requests use stable endpoints only; a denied, empty or failed stable request does not fall back to retired v3 URLs. The shared client masks raw and URL-encoded key echoes, retries malformed JSON once, and retries a 429 at most once, after a 60-second wait, before stopping on a repeated rate limit.