For Deep-value hunters
Cigar butts without the asbestos.
Net-net, sub-tangible-book, single-digit P/E plays - but with the triple-fraud check overlaid so you do not step on accounting landmines while reaching for the deep-value cigar.
· Reviewed by Javier Sanz, ValueMarkers Founder
The pain we solve for deep-value hunters
A stock at 0.6x book is either a steal or a fraud. The screener that finds it does not tell you which. We do - Piotroski + Altman + Beneish on every cheap stock.
Must-haves we built in
- P/B < 1.0 AND P/E < 10 AND ROIC > 0
- Altman Z-Score > 1.8 (not in distress)
- Beneish M-Score < -1.78 (not manipulating earnings)
- Piotroski F-Score >= 5 (improving fundamentals)
- Insider buying in last 6 months (positive signal)
VM features tailored to you
- Deep-value screener with Z-Score floor
- Net-net (NCAV > MarketCap) screener
- Glass-box DCF with bear-case sensitivity
- Greenblatt Magic Formula - original + Piotroski-filtered version
How we filter deep-value hunter candidates
Deep-value investing is the highest-reward and highest-trap quadrant of fundamental analysis. ValueMarkers structures the workflow around three independent fraud-and-distress filters layered on top of cheapness. First, Altman Z-Score above 1.8 - this 5-factor bankruptcy-prediction model penalizes negative working capital, retained-earnings erosion, and excessive leverage. Names in the "distress zone" (Z < 1.8) statistically go bankrupt at 5-10x the rate of safe-zone names. Second, Beneish M-Score under -1.78 - this earnings-quality test flags managements potentially inflating revenue or capitalizing operating expenses. The cohort of deep-value stocks is disproportionately exposed to manipulation because management is often trying to hide structural decline. Third, Piotroski F-Score above 5 - the 9-point test favors deep-value names whose fundamentals are improving (margins expanding, leverage falling, share count stable). Joseph Piotroski's original 2000 paper showed that filtering low P/B stocks by F-Score above 5 doubled excess returns versus an unfiltered low-P/B screen. We replicate that filter in real time on 44,722 stocks across 73 exchanges.
Building the screen step by step
Start broad: P/B < 1.0 AND P/E < 12 AND market cap > $200M (excludes microcap manipulation risk). Layer Altman Z-Score > 1.8 to exclude likely-bankrupt names. Layer Beneish M-Score < -1.78 to exclude likely-manipulator names. Layer Piotroski F-Score >= 5 to keep only deep-value names with improving fundamentals. The resulting universe typically contains 40-120 names. Then sort by ROIC descending - the best deep-value picks are companies temporarily mispriced but with above-average historical capital efficiency. Buy a basket of 15-30 names, sized equal-weight, rebalanced quarterly. Historical academic literature suggests this construction outperforms unfiltered deep-value by 4-7% annualized.
Common mistakes deep-value hunters make
Three errors recur in deep-value investing. (1) Skipping the Altman filter and ending up with 20% of the basket in actual bankruptcy candidates - returns get destroyed by one or two zeros. (2) Concentrating into a single sector that is uniformly cheap (financials in 2009, energy in 2016, real estate in 2023) - macroeconomic correlation overwhelms the per-name edge. Always cap sector exposure at 25%. (3) Failing to size positions equal-weight - a deep-value strategy generates excess return on the median name, not on outliers, so equal-weighting prevents one bad pick from dominating the basket.
Case study: OPI
Office Properties Income (OPI) traded at 0.4x book and 6% yield in mid-2023. The screener flagged Altman Z-Score at 1.3 (deep distress), Beneish M-Score at -1.5 (manipulation flag), and Piotroski at 3. All three filters said skip - and by Q1 2024 OPI cut the dividend by 55% and the equity dropped another 60%. Deep-value cohort dodged this name entirely.
Case studies illustrate how the ValueMarkers screen flagged this name historically; they are research examples, not investment recommendations. See our full disclaimer.