Method

How QualityDip values a company

Three numbers per company, recomputed from public data after each close, and three rules for spotting a reversal. Simple published models, not predictions.

1. Market Value

The volume-weighted average daily close since the market last got thesis-changing news. Prices from before that news don’t count, because the information behind them is out of date.

“Thesis-changing news” is read straight from the company’s SEC 8-K filings:

• Item 2.02: earnings results
• Item 1.01: a material agreement
• Item 2.01: an acquisition or disposal
• Item 5.02: only when the CEO or CFO changes (board and other officer changes don’t count)
• Items 7.01 / 8.01: only when the filing reads as guidance, an outlook, or a deal. Debt offerings, dividends, stock splits and routine notices don’t count.

The average starts on the first full trading day after the filing was accepted (a filing after the open counts from the next day). Under 3 trading days it’s marked “forming”; the Ledger waits for at least 5. For the core watchlist each 7.01/8.01 filing is read and classified; for the wider universe the scanner goes by the filing’s item numbers, so the two can occasionally start from different filings.

2. Intrinsic value (DCF)

Every company we cover gets the same discounted-cash-flow model — today 759 of 1001 have a value.

• Starts from free cash flow per share over the last 12 months: operating cash flow minus capital spending (purchases of property, plant and equipment), divided by diluted shares.
• Growth for years 1–5 is the company’s own revenue growth over its last three fiscal years, clamped between 3% and 15% (6% if there’s less than two years of history). It then fades evenly to 3% by year 10.
• Discount rate 9.5%. Terminal growth 3% after year 10. The value is what those ten years plus the terminal value are worth today, per share.
• Free cash flow is counted after interest, so it’s treated as cash to shareholders; net cash or debt is shown on the stock page but not added or subtracted.
• Inputs come from the company’s own SEC filings (XBRL financial data), refreshed when a new 10-Q, 10-K, 20-F or 40-F lands. Each stock page names the filing and its date. For our nine core names the trailing figures come from Yahoo Finance and are checked against SEC.
• Foreign companies filing 20-F or 40-F reports use their IFRS figures from the latest annual report, converted to dollars at the latest exchange rate. For ADRs, the SEC share count has to match the listed ADRs at a standard ratio (for example two shares per ADR), or we show no value.

When the model doesn’t fit, the page says n/a and why, instead of guessing:

• Negative or near-zero free cash flow (heavy spending years): no value.
• Missing or out-of-date SEC data: a few companies haven’t filed machine-readable statements recently (or ever, for brand-new listings), or their share count can’t be matched to the listed shares. No value until they do. Where a company files capital spending under its own label, or only inside its latest annual report, we read that figure from the filing itself and say so on the stock page.
• Royalty/streaming companies and fund-consolidating asset managers: their spending is stream or fund investments that can’t be separated from acquisitions, so no value.

What the price implies

Next to each fair value estimate we run the same model backwards: the yearly free-cash-flow growth for years one to five (fading to 3% by year ten, exactly as in the DCF) that would make the value equal today’s price, or for banks and insurers the return on equity that would. It doesn’t change the estimate; it shows what the market is assuming, so a big gap reads as a question about growth rather than a typo.

Capital-spending surges. When capital spending jumps to 60% or more of operating cash flow and at least 25 points above its level three fiscal years earlier (Alphabet, Microsoft and Meta during the AI build-out, for example), trailing free cash flow isn’t representative of the business, so we show “estimate unreliable” with both numbers instead of a DCF.

Results we won’t stand behind. When the free-cash-flow DCF lands more than three times the price (over 200% upside) or below 15% of it (more than 85% downside), the inputs are saying more than the business: peak-cycle cash flow, heavy debt, customer or loan balances flowing through cash flow, or a fast grower with little cash flow yet. Those pages say “estimate unreliable” with the raw result and still show what the price implies. Banks, insurers and lenders use tighter bounds (above, +80% / −60%).

Banks, insurers and lenders: the Financials model (residual income)

Deposits, loans, premiums and client money run through a financial company’s cash flow statement, so free cash flow doesn’t measure what it earns. For banks, insurers, brokers and lenders the value column shows a Financials model (residual income) instead, marked “RI” in the tables:

• A standard residual-income model. Value per share = today’s book value per share + the present value of each future year’s residual income, the earnings above a 9.5% cost of equity: (ROE − 9.5%) × book value at the start of that year.
• Starting ROE = the average of two numbers, both net income to common shareholders ÷ average common equity (start and end of the period): the last 12 months (or the latest fiscal year for annual filers), and the average of the last five fiscal years (SEC filings). Capped between 0% and 30%. One unusually strong or weak year (a quiet catastrophe season, a tax-asset release) counts for half rather than carrying the whole estimate. Book value is shareholders’ equity from the latest 10-Q or 10-K, less preferred stock where it’s reported, divided by the shares outstanding on that filing’s cover page. When SEC’s data feed hasn’t picked up a filed report yet, we read the numbers from the filing itself. The cost of equity is 9.5% for every company, the same rate as the DCF.
• Book value grows each year by retained earnings: ROE × (1 − payout ratio). The payout ratio is common dividends plus net buybacks over the last 12 months (SEC cash-flow statement) ÷ net income, capped between 0% and 100%. Net buybacks = shares repurchased minus stock-based pay, because shares issued to employees are credited back to equity. Money handed back to shareholders stays theirs; it just stops compounding inside the company.
• Estimate unreliable: we show no number, and say why, when the company lost money in two or more of the last five fiscal years, when trailing ROE is more than 2.5× its five-year average, when the return on book averages above 30% (a fee or brand business that book value doesn’t capture: AXP, AMP and some brokers), for hybrid asset managers that consolidate an insurer (Apollo/Athene, KKR/Global Atlantic), or when the result is more than 80% above or more than 60% below the price.
• Book value excludes AOCI (since Nov 2, 2026) for US filers: unrealised gains and losses on bond portfolios sit in “accumulated other comprehensive income” and flow back into book as the bonds mature without passing through earnings, which a model that grows book only from earnings would otherwise never count. Book and every ROE use common equity excluding AOCI on the same basis.
• Fresh filings first. When SEC’s data feed lags a report the company has already filed (Citigroup’s 10-Qs; the Canadian banks’ quarterly 6-Ks; many 20-F annual reports), we read that filing’s own XBRL instead.
• The excess return fades evenly over 10 years toward the 9.5% cost of equity, but only halfway: a bank earning 19.5% starts at a 10-point excess and is at 5 points by year 10. From year 11 that residual income grows at 3% a year forever, valued at year 10 as residual income in year 11 ÷ (9.5% − 3%). A company earning exactly 9.5% on equity is worth its book value.
• The same rules apply to every financial; nothing is tuned per stock. Low-book businesses (brokers, asset managers, crypto and payments platforms) usually look expensive on it because their value sits in fees rather than book, which is why the five-year ROE check above exists. Models are estimates, not advice.

“Cyclical / peak cash flow — treat with care.” Stock pages and tables flag a value with this chip when the model shows more than 100% upside, or when the business is cyclical or rides a cash-flow peak (miners, energy and refiners, memory chips, payments and fintech). One strong year of cash flow can make any model look cheap.

Limits. A DCF is only as good as its inputs. Small changes in growth or the discount rate move the result a lot, one unusual year of cash flow can skew it, and fast growers usually look expensive on a model that caps growth at 15%. Read it as a reference point next to Market Value, not as a price target. Models are estimates, not advice.

3. FCF value

Free cash flow per share ÷ 4%: the price at which the stock would give you a 4% free-cash-flow yield.

Fair value range, key levels, bull vs bear, risk vs market

Fair value range. The base case is the DCF above (or the residual-income model for banks and insurers). The bear and bull cases rerun the same model with its two biggest inputs moved: the discount rate 1 point higher or lower and growth 2 points lower or higher (for financials, cost of equity ±1 point and return on equity ±2 points). It’s a fair value estimate, not a prediction or a price target. When a company has no DCF, we show no range.

Key levels. The price, Market Value, the DCF base case, an armed setup’s trigger and stop, and the 52-week low and high (daily closes), each with its distance from the price.

Bull vs bear. Up to five points each side, written by fixed rules from the numbers on the page: the DCF gap, Market Value gap, P/E against its own history and its sector’s median, free-cash-flow yield and P/FCF, EV/EBITDA, 12-month and 3-year sales growth, diluted EPS growth, gross and operating margin trends, return on equity and on invested capital, net cash or debt, debt/equity, interest coverage, free-cash-flow growth, buybacks or dilution, dividends and payout (all from SEC filings, see Fundamentals), analyst upgrades and downgrades among today’s notable voices, dip depth, RSI, the 200-day trend, distance to the 52-week high and low, the cyclical flag, earnings within 14 days, the quality screen, an armed or fired setup, beta, and dated investor quotes with their source. Business fundamentals lead: each side shows up to three fundamental points first, and when the specific rules find fewer than three, the company’s strongest (bull) or weakest (bear) metrics against the median of the sector peers we cover fill in (sales and EPS growth, margins, ROE, ROIC, free-cash-flow yield, P/E, price/book, EV/EBITDA, debt/equity, share count, dividend yield), each with both numbers. Price-based points (technicals, setups, beta, Market Value) are capped at two per side. Each rule has a strength score; we show the strongest first and fewer than five when the data doesn’t support more. No opinions, no invented numbers.

Risk vs market. One year (252 sessions) of daily returns: beta against the S&P 500 (SPY), correlation with SPY and with the Nasdaq-100 (QQQ), and annualized volatility next to SPY’s. Under 0.7 is low, 0.7–0.9 below market, 0.9–1.1 market-like, 1.1–1.5 above market, over 1.5 high. It describes the past year, not the future.

Fundamentals, the screener and “All fundamentals”

Every fundamental comes from the company’s own SEC filings (XBRL company facts), converted to US dollars at the latest exchange rate where a company reports in another currency. Trailing 12 months = the last fiscal year plus this year’s year-to-date minus last year’s (or the last four quarters). Market cap = price × the latest diluted share count (adjusted for the ADR ratio), checked against the listed market cap.

Valuation: P/E = market cap / net income to common; P/FCF = market cap / free cash flow (operating cash flow − capital spending); FCF yield is its inverse; EV/EBITDA = (market cap + debt − cash and short-term investments) / (operating income + depreciation and amortization); P/S and P/B use revenue and shareholders’ equity. P/E vs its own history compares today’s P/E with the average of the P/E at the last three fiscal year-ends (year-end close / diluted EPS from the latest annual report); we skip it when any of those is negative or above 100, or when the share count moved sharply (a split or merger). Profitability: gross, operating, net and FCF margins; ROE = net income / average equity; ROIC = operating income × (1 − effective tax rate, 0–35%, else 21%) / (debt + equity − cash). Growth: sales and FCF vs the 12 months a year earlier, sales growth per year over three fiscal years, diluted EPS growth in the last fiscal year. Balance sheet and cash returns: debt/equity, net cash, interest coverage (operating income / interest expense), dividends and buybacks paid in the last 12 months, and the diluted share count vs a year ago.

Ratios on a near-zero base (P/E above 300, ROE above 200%, margins below −200%, growth from almost nothing) are shown as n/m, not meaningful. Margins, EV/EBITDA, ROIC and debt ratios are not meaningful for banks, insurers and lenders. For banks and lenders, revenue = net interest income + noninterest income, because the plain revenue tag often covers only part of a bank’s revenue; when a bank doesn’t tag both and its revenue tag is less than twice its net income, revenue (and the ratios built on it) shows n/a. When a company doesn’t tag a number in its filings, we show n/a with the reason and never fill it in.

The lists on the front page

• Needs attention now: reversal signals and Ledger alerts that fired in the last five sessions; armed capitulation setups, nearest to their trigger first; and the quality names already in a dip that fell hardest on the last session.
• Top Picks: the Companies of Interest score, plus 12 points for a rule that fired in the last five sessions, 6 for a capitulation setup within 3% of its trigger, 3 for ROIC above 20% (ROE above 25% for banks) and 2 for a free-cash-flow margin above 20%. A name needs a 5% discount to Market Value or a live signal to qualify. A screen, not a recommendation.
• Investor Buzz: public comments by well-known investors on names in our lists, with the date and a link to the source. Their view, not ours.
• Companies of Interest: every quality company in a Deep dip or on Dip watch, ranked by its discount to Market Value (counted once Market Value has 5+ trading days) plus half its drop from the 52-week high.

Data sources

• Filings: SEC EDGAR (company submissions and the 8-K documents themselves).
• Cash flow and shares: SEC XBRL company facts for every covered name; for the nine core names, Yahoo Finance trailing-12-month fundamentals with SEC as the cross-check.
• Daily prices and SPY: Yahoo Finance daily closes.

When it updates

After the US market close on trading days. Market Value alerts (Rules v1) are judged only on a final daily close. Reversal Signal Rules v1: R1 is alerted intraday the moment price clears its trigger; R2 and R3 are judged on the final daily close.

Ledger rules

These are model estimates built from public data and may be wrong. Paper-tracked, not real trades. Returns exclude dividends, fees and taxes. Past results don’t predict or guarantee future results. Information only, not investment advice.