Method
Distrust the data first. Then decide.
Market data gives you speed. The filings give you the answer. When they disagree, we side with the filing and write down the gap.
01
Where data vendors get it wrong
Revenue, earnings, GAAP EPS and operating cash flow are usually copied correctly. The trouble starts with the numbers vendors compute themselves. Cases we have hit in real research:
- Free cash flow with no stated definition, off by a multiple when rebuilt from the filing.
- A "forward P/E" built on next fiscal year’s EPS. For companies whose fiscal year is offset from the calendar, that can be a valuation almost two years out.
- Operating income with a trademark impairment added back, so it no longer matches GAAP.
- Unrealised gains labelled as "gains on sale of securities".
- The risks that matter most (off-balance-sheet guarantees, tax litigation, subsequent events) sitting only in the 10-Q or 8-K text, invisible in structured data.
02
What we check
| Area | What we look at | Where |
|---|---|---|
| Core figures | Revenue, operating income, net income, GAAP EPS, operating cash flow | XBRL, quarter by quarter |
| Earnings quality | Is net income propped up by investment gains or fair-value marks? | Other income notes in the 10-Q / 10-K; company adjusted EPS |
| Cash flow and working capital | Cash flow vs net income, receivable days, inventory mix | Liquidity and Capital Resources |
| Debt and capital allocation | Bond terms, buybacks and dividends, remaining authorisation | 8-K items 8.01 / 2.03, 10-Q |
| Off-balance-sheet obligations | Guarantees, purchase commitments, leases not yet commenced, VIEs, subsequent events | 8-Ks filed after the latest periodic report |
| Non-GAAP definitions | Did the definition change? Is history still comparable? | Non-GAAP notes in the press release |
| Concentration | Share of revenue or receivables from top customers | 10-K / 10-Q notes |
| Vendor-derived metrics | Free cash flow, forward P/E, Normalized Income and the like | Rebuilt from the filing and listed in the corrections table |
03
Scorecard
| Dimension | Main evidence | How it is scored |
|---|---|---|
| Fundamentals | Revenue growth, margin trend, ROE | 3 = in line with peers; the direction of growth and margins moves it up or down |
| Valuation | Current-year P/E vs peer median, PEG | The cheaper versus peers, the higher; an excessive PEG costs a notch |
| Momentum | Excess return vs the S&P 500, moving averages | Beating over most periods and above the 200-day line scores high |
| Earnings quality | Non-operating income share, FCF/net income, receivables trend, definition changes | Starts at 5; each problem found takes a point |
| Risk | Volatility, drawdown, customer concentration, leverage, off-balance-sheet items, litigation | Higher means lower risk; starts at 5 and deducts |
Every score cites a number; without enough data it is "not rated". The five dimensions are weighted into a suggested rating. Rules: Ratings.
04
What we don’t do
- Fake precision. If we cannot verify it, we say so; guesses are marked as guesses, estimates as estimates.
- Hide the lag. Market data is delayed and every report states its cut-off.
- Decide for you. Ratings are research opinions for all readers, not tailored to anyone’s situation.