phanijapps/yf-signals

YF Signals

Overview

Generate reproducible technical signals from direct yfinance OHLCV pulls. Keep calculations transparent so conclusions can be audited and communicated clearly.

Workflow

1. Load cleaned OHLCV data from yf-data or pull directly with yfinance.
2. Compute core indicators: EMA(20/50/200), RSI(14), MACD(12,26,9), ATR(14), Bollinger(20,2).
3. Build a weighted signal score: trend + momentum + volatility + volume confirmation.
4. Run multi-timeframe checks (for example, 1d trend and 1h timing).
5. Emit machine-readable signals and analyst-friendly visual summaries.

Signal Schema

Return one row per symbol/timeframe with:

• trend_state: bullish | neutral | bearish
• momentum_state: rising | flat | falling
• volatility_state: expansion | contraction
• signal_score: -100..100
• risk_flags: list of rule breaches such as overbought or low-volume breakout

Collaboration Contract

• Coordination role: provide indicator parameters and validation expectations.
• Quant role: return signal tables, assumptions, and visualization artifacts.
• Research role: correlate large technical moves with external catalysts.
• Reporting role: summarize alignment/conflicts across indicators.

Decision Rules

• Require at least two aligned categories before making a directional call.
• Downrank signals when volume confirmation fails.
• Mark analysis incomplete when bars are fewer than longest_lookback + 20.

References

Read references/indicator-recipes.md for pandas formulas and fallback logic when pandas_ta is unavailable.