01
Profitability
ROEFormula:
Net income ÷ EquityBands
- Score 0: ROE ≤ 5%
- Linear: 5% → 25% maps 0 → 20
- Score 20: ROE ≥ 25%
Capital efficiency. Buffett's stated quality threshold is sustained ROE above 15%.
Methodology
Roiver Score™
A 0–100 quality composite calibrated for retail value investors. We publish the full formula because no-black-box is the point.
Pedigree
Inspired by Joseph Piotroski's F-Score (financial strength), Edward Altman's Z-Score (bankruptcy risk), and the Buffett/Munger framework on quality compounders at fair prices. Calibrated for the modern retail self-directed investor — different thresholds than 1968-vintage formulas, same underlying logic.
Five components, each 0–20
Five components, each scored 0–20. If one can't be measured from a company's filings, the score is scaled from the ones that can (at least three are required). Each leg has linear bands — no magic curve-fits, no opaque weighting tricks. A weak leg drags the total; users see exactly which one. These are the standard components; banks, insurers and REITs use the sector models below.
02
Capital structure
Debt / Equity (inverse)Formula:
Total debt ÷ EquityBands
- Score 20: D/E ≤ 0.3 (fortress)
- Linear: 0.3 → 2.0 maps 20 → 0
- Score 0: D/E ≥ 2.0 (distressed)
Survives downturns. High leverage is one of the most consistent predictors of permanent capital loss.
03
Cash quality
FCF marginFormula:
Free cash flow ÷ RevenueBands
- Score 0: FCF margin ≤ 5%
- Linear: 5% → 25% maps 0 → 20
- Score 20: FCF margin ≥ 25%
Cash is harder to fake than earnings. Companies with strong FCF margins have optionality and survive bad years.
04
Valuation
Earnings yieldFormula:
Net income ÷ Market capBands
- Score 0: EY ≤ 3% (P/E > 33, expensive)
- Linear: 3% → 10% maps 0 → 20
- Score 20: EY ≥ 10% (P/E < 10, cheap)
The bond-equivalent of stock price. Price discipline matters even for great businesses.
05
Stability
Multi-year ROE consistencyFormula:
Coefficient of variation (stdev / mean) of ROE historyBands
- Score 20: CV ≤ 10% (very consistent)
- Linear: 10% → 80% maps 20 → 0
- Score 0: CV ≥ 80% (highly volatile)
Compounding requires consistent profitability, not one-off spikes. Most composite scores miss this leg.
Sector models
Banks, insurers and REITs are scored on measures suited to them. Debt-to-equity says little about a bank, whose deposits are its raw material, and depreciation makes a REIT's net income understate the cash its properties throw off. Each sector model still has five components, each scored 0–20 on linear bands, and the same tiers. A company scored on one is marked with its model (for example "Bank model") wherever its score appears.
The bands are calibrated, not guessed: each sector component puts the same share of that sector's profitable companies at 0 and at full marks as the matching standard component does across profitable companies generally. A typical profitable bank, insurer or REIT therefore scores about where a typical profitable company does, rather than its sector sweeping the top of the rankings.
Banks
| Component | Measure | 0 → 20 |
|---|---|---|
| Profitability | ROE | 5.5% → 0, 13%+ → 20 |
| Capital strength | Equity ÷ total assets | 8.5% → 0, 13%+ → 20 |
| Efficiency | Non-interest expense ÷ (net interest income + non-interest income) | 64%+ → 0, 51% or less → 20 |
| Valuation | Earnings yield | 6% → 0, 9.5%+ → 20 |
| Stability | Variation in ROE across years | 25%+ → 0, 4.5% or less → 20 |
Property & casualty insurers
| Component | Measure | 0 → 20 |
|---|---|---|
| Profitability | ROE | 8% → 0, 20%+ → 20 |
| Capital strength | Equity ÷ total assets | 16.5% → 0, 30.5%+ → 20 |
| Cash generation | Operating cash flow ÷ revenue | 14% → 0, 35%+ → 20 |
| Valuation | Earnings yield | 7% → 0, 14%+ → 20 |
| Stability | Variation in ROE across years | 55%+ → 0, 10% or less → 20 |
Life & health insurers
| Component | Measure | 0 → 20 |
|---|---|---|
| Profitability | ROE | 8% → 0, 20%+ → 20 |
| Capital strength | Equity ÷ total assets | 3% → 0, 13.5%+ → 20 |
| Cash generation | Operating cash flow ÷ revenue | 14% → 0, 35%+ → 20 |
| Valuation | Earnings yield | 7% → 0, 14%+ → 20 |
| Stability | Variation in ROE across years | 55%+ → 0, 10% or less → 20 |
REITs
| Component | Measure | 0 → 20 |
|---|---|---|
| FFO return on equity | FFO ÷ equity | 6.5% → 0, 23.5%+ → 20 |
| Leverage | Total liabilities ÷ total assets | 83.5%+ → 0, 47% or less → 20 |
| FFO margin | FFO ÷ revenue | 31.5% → 0, 67.5%+ → 20 |
| FFO yield | FFO ÷ market cap | 6.5% → 0, 14.5%+ → 20 |
| FFO stability | FFO variation across years | 32.5%+ → 0, 7% or less → 20 |
FFO (funds from operations) here is net income + depreciation & amortization. The standard definition also removes gains on property sales, but filers rarely tag those consistently, so ours is an approximation.
Negative shareholder equity
Some companies have negative shareholder equity, usually because buybacks have returned more than the book value. ROE and debt-to-equity have no meaning when equity is below zero, so these companies are scored on return on assets and liabilities ÷ assets instead.
When a component can't be measured
If a filing doesn't report what a component needs, that component is shown as n/a rather than scored as zero, and the score is scaled from the components that were measured.
| Components measured | Result |
|---|---|
| 5 or 4 | Scored and ranked |
| 3 | Scored, not ranked on leaderboards |
| Fewer than 3 | Not scored |
Score tiers
Excellent
75–100
Solid
50–75
Mixed
25–50
Weak
0–25
What it deliberately doesn't include
- ×Growth metrics. Growth is part of the thesis, not the quality signal. A high-quality slow-grower scores higher than a low-quality hyper-grower — which matches value-investing logic.
- ×Price momentum. Momentum signals reward chasing. Antithetical to value investing.
- ×Sector bands beyond financials and REITs. Banks, insurers and REITs have their own models; every other company, utilities included, is scored on the standard thresholds — so a utility will rarely hit ROE 25%.
Why we publish this
Most quality scores in this category — GuruFocus's GF Score, Morningstar's Quantitative Rating, others — are opaque black boxes. Users see a number with no way to interrogate it. Roiver Score is the opposite by design: every component on every stock page shows its rationale, and this page documents the entire formula. If we're wrong about a band, we want you to push back.
Roiver Score is a quality signal, not a recommendation. Educational use only — not investment advice.