Skip to content

Parameters & Configuration

Price Filters

  • --min-price / --max-price narrow universe to viable capital ranges.
  • Avoid illiquid penny stocks with --min-price 5 or 10.

Volume Filter (RSI Divergence)

  • --min-volume N ensures recent average volume (last bar) exceeds threshold.
  • Helps eliminate thinly traded noise setups.

RSI Settings

  • --rsi-period (default 14). Lower values (9) increase sensitivity; higher (21) smooth.
  • Consider longer period for weekly scans.

Swing Window

  • --swing-window controls local extremum detection width.
  • Larger window = fewer, more significant swing points.

Lookback

  • --lookback bars to search for last two qualifying swings.
  • Typical range: 50–80 for daily; can expand for weekly.

Aggregation Period

  • --period sets candle aggregation size: 1d, 1wk, 1mo.
  • Choose higher timeframe (1wk) for broader context; lower (1d) for tactical entries.

Historical Span

  • --lookback defines the data history window fetched: 1mo, 3mo, 6mo, 1y, etc.
  • Must be sufficient to compute indicators and run statistics (include enough prior runs).

Tolerance

  • Fixed RSI tolerance (0.5) ensures meaningful divergence.
  • Adjustable in divergence.py if needed.

Input Filter

  • --input-filter FILE loads tickers from a previous screener CSV to intersect signals.
  • Enables multi-factor workflows (e.g., green HA AND bullish divergence).

Output

  • --output FILE writes results to CSV.
  • For reproducibility, include parameters in filename (e.g., div_bullish_6mo_tol0.5.csv).

Delay (HA Screener)

  • --delay controls per-ticker pause for yfinance.
  • Reduce cautiously (risk rate limiting).

Quiet Mode

  • --quiet suppresses progress logging for cleaner automation output.
  • Momentum exhaustion: --type bearish --min-price 50 --min-volume 2_000_000
  • Reversal scan: --type bullish --swing-window 7 --rsi-period 21
  • Liquidity focus: --min-volume 5_000_000 --min-price 10
  • Extended HA runs: --min-run-percentile 90 --period 1d
  • Early HA runs for continuation: --max-run-percentile 25 --period 1d

Extended Terminology Guide (Merged)

Aggregation Period (--period)

Defines candle size: 1d, 1wk, 1mo (replaces earlier --interval).

Time Window (--lookback)

Historical span fetched: 1mo, 3mo, 6mo, 1y, 2y, 5y.

Explicit Date Range (--start / --end)

Overrides lookback when both provided; format YYYY-MM-DD.

Example Commands

stockcharts-screen --color green --period 1d --lookback 3mo --output green_reversals.csv
stockcharts-screen --color green --period 1wk --lookback 1y --output weekly_reversals.csv
stockcharts-screen --color green --period 1d --start 2024-01-01 --end 2024-12-31
stockcharts-screen --min-run-percentile 90 --period 1d --lookback 6mo --output extended_runs.csv
stockcharts-screen --max-run-percentile 25 --period 1d --lookback 3mo --output early_runs.csv

Common Mistakes

  • Mixing lookback with start/end.
  • Using too short lookback to detect color change.
  • Expecting run_percentile with insufficient history (use larger --lookback).
  • Using deprecated --interval instead of --period.

Multi-Timeframe Workflow

Weekly context → Daily execution → Optional charting.

Heiken Ashi Run Statistics

The screener derives distribution-aware metrics for trend maturity:

Field Description
run_length Consecutive same-color HA candles ending at latest bar
run_percentile Inclusive percentile rank of run_length among all historical runs in current dataset

CLI Filters: - --min-run-percentile PCT include only runs with percentile >= PCT. - --max-run-percentile PCT include only runs with percentile <= PCT.

Guidelines: - 90–100: Extended; evaluate for exhaustion. - 60–80: Mature; monitor for divergence. - 20–40: Emerging; possible continuation potential. - 0–10: Fresh or rare short reversal start.

Notes: - Percentile uses inclusive ranking (count of runs <= current / total * 100). - Expand --lookback to improve percentile stability. - Combine with --changed-only to isolate fresh flips vs continuations.