METHODOLOGY
# METHODOLOGY VERSION 2.0 — EFFECTIVE 2026-08-20 (SCORECARD SCHEMA V2)
# STOCKVERA scores a curated US equity universe with a deterministic, regime-adaptive model. Every signal is produced by the same code path: screen → score → rank → publish. No discretion. No LLM. Records carry their generating model version and scorecard schema version so they remain interpretable under the rules then in effect.
# SCORING CONTRACT (V2)
# FOUR CATEGORIES, REGIME-ADAPTIVE WEIGHTS. CATEGORY TOTAL = 100 BEFORE ADJUSTMENTS.
| CATEGORY | DEFAULT | BULL | CAUTIOUS | VOLATILE/BEAR |
|---|---|---|---|---|
| VALUE | 25 | 18 | 28 | 30 |
| QUALITY | 25 | 17 | 27 | 30 |
| TECHNICAL | 30 | 35 | 22 | 15 |
| RISK | 20 | 30 | 23 | 25 |
P/E, PEG, DCF discount, P/B, P/S. Missing inputs score 0 and are labeled as unavailable, not as measured zeroes.
ROE, profit margin, FCF yield, debt/equity, current ratio. When no fundamental data arrives, the bucket shows NO DATA and the quality gate is neutral.
Eight sub-factors: momentum, trend alignment, volume flow, RSI, MACD, Bollinger position, ADX, relative strength. Sub-factor maxima scale with the regime technical budget.
Beta, ATR%, risk/reward vs support and 1.5×ATR stop, and 20-day drawdown.
# DETAIL-PAGE DISPLAY BUCKETS
The signal detail page shows SIX buckets: VALUE, QUALITY, MOMENTUM, TREND, TECHNICALS, and RISK MGMT. MOMENTUM and TREND are broken out of the eight technical sub-factors for readability — the six display buckets do not change the underlying four-category computation. Bucket maxima follow the versioned contract: under BULL weights, for example, MOMENTUM and TREND display out of 6 rather than 5.
# TOTAL SCORE TRANSFORM
▸ CATEGORY TOTAL (0-100) + SECTOR ADJUSTMENT (±2) + CONFLUENCE BONUS (−5…+5) + QUALITY BONUS (−3…+3), CLAMPED TO 0-120.
▸ CONFLUENCE COUNTS HOW MANY OF 11 FACTOR GROUPS (4 CATEGORIES + 8 TECHNICAL SUB-FACTORS, MINUS 1 OVERLAP) SCORE ABOVE THEIR MIDPOINT.
▸ THE TOTAL IS A RELATIVE MODEL SCORE, NOT A PROBABILITY AND NOT A RETURN FORECAST. RANK IS CROSS-SECTIONAL POSITION; RISK CLASSIFICATION IS INDEPENDENT OF RANK.
# RISK CLASSIFICATION
LOW: quality A/B, value ≥ 15/25, risk ≥ 14/20, total ≥ 60. MEDIUM: total ≥ 45 with quality ≥ C. HIGH: everything else. A high rank can legitimately carry HIGH risk — risk is not derived from rank.
# ENTRY / EXIT FRAMEWORK
# ENTRY ZONE:
Nearest support floor to current price, capped at 2% wide; never implies an above-market fill.
# TARGET 1:
Nearest resistance (3-10% above) or 2.5×ATR; at least 4% above the buy zone and 2:1 reward-to-risk.
# TARGET 2:
Stretch: max(4×ATR, 1.02×resistance), capped at 15% above price; only shown if >3% above target 1.
# STOP LOSS:
Max of (0.995×support, 0.995×swing low, price−1.5×ATR), floored at 8% below price.
# GUARDRAILS (CURRENT CONFIGURATION)
HARD EXCLUSIONS — ENFORCED ON EVERY PUBLICATION PATH (STRICT SCREEN AND RELAXED FALLBACK)
■ Price within the tier's bounds: PRIMARY ≥$100 · MID $20-$100 · VALUE $2-$20
■ Minimum average volume: PRIMARY 1M · MID 500K · VALUE 300K shares/day
■ Minimum market cap (when known): PRIMARY $10B · MID $2B · VALUE $500M
■ ATR within tier range: PRIMARY 0.5-4% · MID 0.5-5% · VALUE 0.5-6%
■ RSI between 25 and 75
■ Earnings within the target window excluded
■ Universe exclusion list (ETFs, leveraged products, delisted/speculative names)
STRICT-SCREEN PREFERENCES — APPLIED IN THE STRICT PASS; MAY BE BYPASSED BY THE DISCLOSED RELAXED FALLBACK
WHEN THE STRICT SCREEN RETURNS NO ELIGIBLE SIGNAL FOR A TIER, THE RELAXED FALLBACK PUBLISHES THE BEST REMAINING CANDIDATE THAT PASSES THE HARD EXCLUSIONS ABOVE. EVERY FALLBACK SIGNAL CARRIES AN EXPLICIT "RELAXED FALLBACK" DISCLOSURE ON ITS DETAIL PAGE LISTING EXACTLY WHICH PREFERENCES WERE NOT MET. FALLBACK SIGNALS SKIP THE FULL QUALITY/STRUCTURAL SCREEN — ELEVATED RISK VERSUS STRICT-SCREEN SIGNALS.
■ 5-day change ≥ −3% (close vs close five trading sessions earlier — the same interval used in every '5-day' explanation)
■ Price above its 50-day average (when the average is available)
■ Volume ratio ≥ 0.8× the 20-day average
■ Quality grade floor per tier: PRIMARY A/B · MID A/B/C · VALUE A/B/C
■ PRIMARY tier only: beta ≤ 1.8 · P/E positive and ≤ 50
THE $5 MINIMUM PRICE DESCRIBED IN THE LEGACY METHODOLOGY APPLIED TO THE OLD GENERIC SCREEN. UNDER V2, SUB-$5 TICKERS ARE ELIGIBLE ONLY IN THE VALUE TIER (FLOOR $2) — FOR EXAMPLE THE 2026-09-03 GRAB SIGNAL AT $3.53 WAS PUBLISHED UNDER THE VALUE TIER RULES THEN IN EFFECT.
# UNIVERSE & DATA
▸ CURATED LIST BUILT FROM S&P 500 CONSTITUENTS PLUS A FIXED SET OF ADDITIONS, MINUS A SPECULATIVE/MEME EXCLUSION LIST. THE LIST IS STATIC BY DESIGN — THIS IS A DISCLOSED SOURCE OF SURVIVORSHIP BIAS IN RESEARCH TESTS.
▸ PRICES: YAHOO FINANCE ADJUSTED OHLCV (PRIMARY), FMP HISTORICAL EOD (FALLBACK). FUNDAMENTALS: FMP. EARNINGS: NASDAQ CALENDAR. SENTIMENT/REGIME: CNN FEAR & GREED + VIX.
▸ MISSING FUNDAMENTALS ARE LABELED UNAVAILABLE AND SCORE NEUTRALLY — THEY ARE NEVER DISPLAYED AS MEASURED ZEROES.
# LEGACY METHODOLOGY (PRE-2026-08-20 RECORDS)
HISTORICAL RECORDS CREATED BEFORE 2026-08-20 USED AN 8-FACTOR, 100-POINT SCHEME (MOMENTUM 0-20, TREND 0-20, VOLUME 0-15, RSI 0-10, MACD 0-10, FUNDAMENTALS 0-15, REL. STRENGTH 0-5, RISK/REWARD 0-5), A $5 MINIMUM PRICE, AND A 500K-SHARE VOLUME FLOOR. THOSE RECORDS REMAIN INTACT AND ARE LABELED "LEGACY SCORECARD" ON THEIR DETAIL PAGES. THEY ARE NOT RERUN UNDER THE NEW MODEL.
# DISCLAIMER
STOCKVERA is algorithmic research only. Not financial advice. Not a recommendation. Past performance ≠ future results. Markets involve loss risk. Do your own research. Consult a licensed advisor. No liability for losses from acting on this data.