How It Works
The Valyra Framework
A systematic 6-pillar scoring model for Indian listed equities — adapted from Graham-Buffett-Munger principles and calibrated for India-specific market realities.
Philosophy
Warren Buffett and Charlie Munger evolved from Graham's pure “cigar butt” value investing into something more nuanced: buying high-quality businesses at fair prices and holding them for the long term. The key insight: time amplifies compounding in quality businesses, making entry price less critical than business quality.
India calibration is the standout design decision. Most frameworks blindly apply US thresholds (P/E < 15, P/FCF < 20) to Indian markets. India trades at a structural premium — ~7% real GDP growth, large domestic consumption story, and promoter-led business structures create a unique context. The Valyra model is calibrated for these realities, not ported from the S&P 500.
Valyra is a quantitative screener, not a qualitative judgment tool — which is its strength and its limit. The model scores what is measurable. The 25–60 screened candidates are the right place to apply qualitative judgment: can I understand this business in 10 years? Does it have pricing power I can feel, not just measure?
Universe & Screening
How 500 stocks become 25–60 scored candidates — 6 hard gates, all must pass
Starting universe: Nifty 500 (~500 stocks). Pre-filter removes PSU banks, complex NBFCs, market cap <₹500 Cr, pledge >30%, and ASM/ESM surveillance list stocks. Gate sequencing is deliberate — earnings quality runs first to eliminate accounting-game companies before valuation or moat is even evaluated.
Earnings Quality Gate
Positive OCF; OCF/PAT > 0.80
OCF-based, not PAT-based. This is the right first gate — it eliminates accounting-game companies before scoring begins. Working capital manipulation inflates PAT while OCF stays honest.
Moat Proxy Gate
ROCE > 15%, ROE > 15% (5yr avg)
ROCE above cost of capital for 5 sustained years is the quantitative fingerprint of a moat. Single-year ROCE can be gamed; the 5yr average cannot.
Balance Sheet Fortress Gate
D/E < 1.0; Interest coverage > 4x
Ensures the business can survive a credit cycle downturn. India's periodic credit crunches (IL&FS, Yes Bank, NBFC crisis) routinely destroy leveraged companies with otherwise good operations.
Valuation Sanity Gate
P/FCF < 40; EV/EBITDA < 50
A loose ceiling, not a value filter — it removes egregiously overpriced stocks where even a perfect business cannot justify the entry price. India quality stocks trade at a structural premium; thresholds are calibrated accordingly.
Management & Promoter Gate
Promoter holding > 35%; pledge > 30% = hard exclude
Promoter skin-in-the-game is a meaningful signal in India's promoter-led market structure. Pledge above 30% is a hard exclude — not a penalty — because it indicates the promoter is under financial stress severe enough to risk a margin call cascade.
Revenue Growth Gate
Revenue CAGR > 10% over 5 years
India's nominal GDP grows ~11% annually. A company growing below this rate is declining in real terms. A stock can pass all 5 quality gates while growing at 4% nominally — which in India's inflationary context is real decline. This gate filters those out.
6-Pillar Scoring Model
Final Valyra Score = Σ(Pillar Score × Weight) × 10 → Range: 0–100
Weights reflect the Munger insight: moat quality compounds wealth; valuation determines the starting point but matters less over a long holding period. Weights sum to 100%.
Core metric: ROCE consistency, gross margin stability
Score = f(5yr ROCE avg, gross margin trend)
The primary driver of long-term compounding. Brand + distribution moats dominate Indian consumer markets. A business with a durable moat compounds wealth even at a fair price — moat deserves the highest weight.
Core metric: P/FCF vs. thresholds · EV/EBITDA vs. thresholds
Score = 60% × f(P/FCF) + 40% × f(EV/EBITDA) · DCF retained as qualitative overlay only
Entry price matters, but less than business quality over the long run. DCF is highly sensitive to terminal growth and discount rate — a 1% change in either swings intrinsic value 30–40%, making it unreliable as a scoring input. P/FCF and EV/EBITDA are more manipulation-resistant primary signals. DCF is retained for Margin of Safety % display as a sanity check.
Core metric: OCF/PAT ratio, margin trend, working capital days
Score = f(OCF/PAT 5yr avg, gross margin trend)
OCF/PAT above 0.80 is already enforced at Gate 1 — this pillar rewards consistency above the threshold. Working capital cycle is the most manipulable dimension of Indian P&Ls; OCF-based quality filters are more reliable than PAT-based ones.
Core metric: D/E, interest coverage, promoter pledge %
Score = 40% × f(D/E) + 30% × f(int. coverage) + 30% × f(pledge %)
Promoter pledge is baked in twice — here and in the penalty system — because it is an India-specific fortress risk indicator with a strong empirical track record as a predictor of distress (DHFL, ADAG group, Zee).
Core metric: Incremental ROCE, dividend consistency, equity dilution
Score = f(incremental ROCE, DPS growth, share count change)
Incremental ROCE is arguably the single best forward-looking moat indicator. A company with 25% ROCE on existing assets but only 10% incremental ROCE is shrinking its moat in slow motion — deploying capital at diminishing returns. Raised from 10% to 15% to reflect this.
Core metric: Promoter holding %, RPT transactions, auditor quality
Score = f(promoter %, RPT/revenue, auditor track record)
SEBI disclosures make governance partially measurable. Low weight because the worst governance failures are caught earlier — qualified audits (Gate pre-filter), high pledge (Gate 5 + fortress pillar), and SEBI actions (–10pt penalty).
Penalty Deductions
Applied post-scoring as deductions from Valyra total
| Trigger | Deduction |
|---|---|
| Promoter pledge > 20% | –5 pts |
| Promoter pledge > 30% | Hard exclude (Gate 5) |
| Qualified audit opinion | –8 pts |
| SEBI / ED action (any of last 5yr) | –10 pts |
| Auditor change (last 2yr) | –4 pts |
| High RPT (> 10% of revenue) | –3 pts |
| FII exodus (3 qtrs) + corroborating signal¹ | –2 pts |
| FII exodus (3 qtrs) standalone — macro/index-driven | Flagged · no deduction |
¹ FII corroborating signals: elevated promoter pledge (>10%), recent auditor change, or weak OCF/PAT (<0.80). Standalone FII selling is often macro-driven — EM rotation, MSCI rebalancing, currency hedging — and should not penalise a fundamentally sound business caught in a broad selloff. The penalty only fires when FII exit co-occurs with a company-specific weak signal, indicating informed money may be responding to something they have spotted.
Limitations & Caveats
Valyra is a quantitative screener, not a qualitative analysis tool. Management quality, industry disruption risk, competitive dynamics, and macro factors are not fully captured. The 25–60 screened candidates are the starting point for qualitative judgment — not a buy list.
Promoter pledge data lags by ~3 months (quarterly BSE filings). The last known value is displayed.
IND AS adoption (2017) creates discontinuity with pre-FY17 data. All ratio calculations use FY17 onwards.
Banks and NBFCs are excluded. Their financials (NIM, GNPA, CAR) require a different scoring model and cannot be evaluated with ROCE-based gates.
OCF data from yfinance can be unreliable; Screener.in is used as primary source for fundamental data.
DCF intrinsic value and MoS% are displayed for reference only. A 1% change in discount rate or terminal growth assumption swings intrinsic value by 30–40%. Treat as a directional sanity check, not a precise target.
Revenue CAGR gate is skipped gracefully when less than 5 years of historical data is available (new listings, restructured entities).
This tool does not constitute SEBI-registered investment advice. Always consult a registered investment advisor before making investment decisions.