Which valuation method is used — and why

Not all companies can be valued the same way. A dividend-paying bank is fundamentally different from a high-growth software company, which is different from a property trust. We automatically select the most appropriate method based on what kind of business you're analysing.

Method A — Buffett Normalized (the default)

Used for most growth companies with a consistent earnings history. We project earnings forward 10 years at a conservatively discounted growth rate, apply an appropriate P/E multiple, and discount back to today at 15% per year. The "Buffett Normalized" part refers to our use of a long-run average earnings figure rather than a single year, following Warren Buffett's approach to measuring a business's true earning power.

Method B — 10 Cap (Owner Earnings)

Used for mature, low-growth companies where cash generation matters more than earnings growth. We calculate owner earnings — Warren Buffett's preferred measure of true cash profitability — and capitalise them at 10 times. This means you're paying 10 years' worth of true cash earnings for the business.

Method C — FCF Growth

Used for businesses primarily valued on their ability to generate and grow free cash flow — typically capital-light technology companies where earnings can be distorted by accounting choices like stock-based compensation or amortisation.

Method D — Dividend Discount Model

Used for dividend aristocrats — companies with 10 or more consecutive years of dividend growth. The intrinsic value is derived from the present value of a growing stream of future dividends. The margin of safety applied is 33% rather than 50%, because the regular income provides ongoing return regardless of the share price.

Method E — Justified P/B (Banks, insurers, and balance-sheet lenders)

Standard earnings-based models don't work well for financial companies because loans and deposits are both assets and liabilities in ways that make "invested capital" a different concept. We use return on tangible common equity (ROTCE) to derive a theoretically justified price-to-book multiple, then multiply by tangible book value per share.

This covers banks and insurers, but also brokerages and capital-markets firms, and — importantly — any consumer lender whose balance sheet is genuinely loan-funded (credit cards, auto loans, student loans), even when it's classified in a broader "financial services" category that also contains asset-light, fee-based businesses like payment networks. A card issuer that earns most of its revenue from interest on a loan book has the same "capex and cash flow don't mean what they normally mean" problem a bank has, even if it isn't technically a bank — we detect this by checking how much of a company's revenue actually comes from interest income, not just its industry label.

Method F — FFO-Based (REITs)

Real estate investment trusts are required to pay out most of their earnings as dividends and take large non-cash depreciation charges that suppress reported earnings. We use Funds From Operations (FFO) — earnings before depreciation — which better captures a REIT's true cash-generating ability.

One exception: mortgage REITs, which own loans and mortgage-backed securities rather than physical property, don't have the depreciation problem FFO exists to fix — their assets don't depreciate the way a building does. They're valued with Method E instead, the same way a bank is, since a portfolio of loans behaves much more like a bank's balance sheet than like a landlord's.