Fair Value Screener Forward-Return Rank
Cross-sectional rank of every fair-value snapshot using our proprietary ensemble scoring model. Updated daily 05:30 ET and every 30 min during market hours. Guide
Algorithmic ranking, informational only. See Terms §17.
What this screener does. Every stock with a fair-value snapshot is tagged as LONG (undervalued) or SHORT (overvalued). A proprietary ensemble model then predicts forward return at 21/42/63 days and ranks every stock cross-sectionally within today's universe. The strongest 5% on each side, after risk filters, are flagged Active.
Column reference:
| Column | What it means |
|---|---|
| Symbol | Stock ticker. Click for the full quote page. |
| Price | Latest closing price. |
| Fair Value | Blended fair-value estimate from 8 methods (DCF, DDM, four peer-multiple comparables, market anchor, options-implied expected price). What the model thinks the stock is worth. |
| Dev % | (Fair Value − Price) / Price. + = undervalued (LONG candidate); − = overvalued (SHORT candidate). |
| Conf % | Confidence in the fair-value estimate (0–100). Reflects how many of the 8 methods produced an estimate and how tightly peer multiples agree. |
| Dir | LONG = undervalued (model says buy). SHORT = overvalued (model says sell short). gated = signal was suppressed by a risk filter. |
| Expected α | The model's forward-return point estimate vs SPY over the selected horizon. E.g., +7.5% at 21d = the model expects this stock to beat SPY by 7.5% over the next 21 trading days. |
| Blended z | The ranking metric. Combines the forward-return model with a fundamental-conviction score, then standardized (z-scored) within today's universe. Larger positive = stronger LONG; larger negative = stronger SHORT. |
| Rank % | Percentile of Blended z within today's same-direction universe (LONGs are ranked among LONGs, SHORTs among SHORTs). LONG ≥ 95% or SHORT ≤ 5% are the strongest signals. |
| Active 21/42/63 | Active flag per horizon. 🟢 = active LONG pick · 🔴 = active SHORT pick · ⚫ = not strong enough · ⚪ = suppressed by risk filter. Active = top/bottom 5% AND all filters pass. |
| Mom | 12-1 momentum: trailing 12-month return excluding the most recent month (classic Jegadeesh-Titman). Captures medium-term trend. Strong positive momentum tends to persist; the deeply negative bucket can mean-revert. |
| DTC | Days to Cover = short interest ÷ average daily volume. How many trading days it would take all current shorts to buy back. Low DTC = few shorts (no consensus to short); high DTC = heavy short pressure already in place. |
| β | Beta vs SPY. 1.0 = moves with the market; >1 = amplified moves (more volatile); <1 = dampened. High beta on the short side is dangerous in bull regimes. |
| Q | Quality Score (0–6). Composite of ROE, gross margin, FCF margin, growth consistency, debt load. 6 = compounder; 0 = struggling/cyclical. High-quality names are filtered out from SHORT signals (they tend to keep grinding up). |
Risk filters (the "gates"). A signal can be ranked highly but still suppressed by one of these direction-specific filters:
- LONG side: Drops names in cyclical sectors (oil/gas, petroleum, hotels, insurance brokers) that systematically value-trap, and names in the "dead zone" of mild-negative momentum (−10% to +5%) which empirically underperforms.
- SHORT side: Drops high-quality compounders (Q ≥ 6), high-beta names (β > 2.0 — catastrophic to short in bull regimes), low-DTC names (DTC < 2 — no organized short pressure means they keep ripping), and high-momentum names (Mom > 20% — AI-darling problem).
Backtest performance (active picks, walk-forward, S&P 500, May 2024 – Feb 2026): 21d LONG mean alpha vs SPY +13.4% · 63d LONG +34.9% · 21d SHORT mean P&L +2.7% · 63d SHORT +6.8%. Gross of transaction costs and borrow. Realistic net likely 30–60% lower. Past performance is not indicative of future results.