Not financial advice. All data from public sources and may be delayed or incomplete. Models shown are statistical descriptions of past behavior, not predictions. Provider: yahoo.
Not enough signal coverage to classify.
Direction comes from bias (mean of the four normalized signal scores): ≥ 58 reads bullish, ≤ 42 reads bearish, otherwise mixed.
Conviction comes from agreement + coverage: strong = agreement ≥ 75% and coverage ≥ 75%; medium = agreement ≥ 50%; otherwise low.
Bullish + strong → Strong Candidate · Bullish + medium/low → Watchlist Candidate · Bearish + strong → Avoid for Conservative · Bearish + medium/low → High Risk · Mid bias → Mixed Signals.
GBM simulation with calibrated drift/vol. Probability terminal price > today.
Mean of RandomForest + LogisticRegression on 7 TA features.
RSI · MACD histogram · ROC10 · ADX, averaged then rescaled.
StockTwits author tags + StockTwits/X/Yahoo News VADER, averaged.
The classification is a deterministic readout of the four directional signals below — not a prediction, not financial advice, and not a recommendation to trade. Each component is normalized to 0–100 and combined by the weights shown.
Insufficient data
0% → 0, 20%+ → 100. Negatives map to 0.
0% → 0, 25%+ → 100.
0% → 0, 25%+ → 100. Above-cost-of-capital territory.
<1 stressed, 1.5–3 healthy, >3 idle cash.
0 → 100, 2.5+ → 0 (inverted, lower leverage is better).
Insufficient data
−10% → 0, +30%+ → 100.
−20% → 0, +50%+ → 100.
Insufficient data
≤10 cheap (100), ≥40 rich (0).
≤10 cheap (100), ≥35 rich (0).
≤0.5 deeply cheap (100), ≥2.5 priced for perfection (0).
≤1 cheap (100), ≥6 rich (0).
Insufficient data
GBM simulation with calibrated drift/vol. Probability terminal price > today.
Mean of RandomForest + LogisticRegression on 7 TA features.
RSI · MACD histogram · ROC10 · ADX, averaged then rescaled.
StockTwits author tags + StockTwits/X/Yahoo News VADER, averaged.
Moderate risk
0% → 100, 60%+ → 0 (annualized).
0% → 100, −60% → 0.
0% → 100, −6% → 0 (worse 1-day downside = lower).
−1 → 0, +2 → 100.
Each sub-score is a transparent weighted mean of normalized inputs (raw values shown below the headline). Mappings are deterministic piecewise functions — expand any card to inspect every driver.
≤1 cheap (100), ≥10 rich (0).
≤6 cheap (100), ≥20 rich (0).
Distance from 1 — exactly 1 → 100, |β−1|≥1.5 → 0.
