How the research is made

Methodology

Research is only as trustworthy as the process behind it. This page describes exactly how a Fundamenta report is produced, what we adjust, what we do not know, and how you can check us.

Last revised 4 August 2026 · Version 3.1

Primary sources

In order of precedence. Where two sources disagree, the higher one wins and the discrepancy is noted in the report.

Source	Used for	Precedence
Annual report / 10-K / 20-F	All audited financial data, segment detail, risk factors	1
Quarterly filings (10-Q, half-year)	Trend within the year, disclosure changes	2
Proxy statement (DEF 14A)	Compensation structure, incentive alignment, ownership	3
Earnings call transcripts	Management framing, questions management avoids	4
Investor presentations	Segment commentary and non-GAAP bridges	5
Third-party data providers	Market prices only. Never fundamentals.	-

Research process

Each report goes through the same six stages. The order matters: we do not look at a valuation until the business is understood.

Read the filings, unaided

The last three annual reports and four quarterly filings, before reading any outside commentary. Notes are taken on the document, not on a summary of it.

Rebuild the financials

Six years of income statement, balance sheet and cash flow are re-entered by hand into a common template, so definitions are identical across every company we cover.

Describe the business model

How the company earns a unit of revenue, what it costs to earn it, and what has to be true for that to continue. Written before any metric is calculated.

Test the competitive position

We look for barriers to entry that show up in the numbers – sustained returns above cost of capital, pricing that outpaces input costs, share that does not erode.

Build the valuation framework

Historical multiples, a reverse-engineered discounted cash flow, and a sensitivity table. The output is a range and a list of assumptions, never a target.

Second read and publication

A second analyst checks every figure against the filing it came from. Anything that cannot be traced to a source is removed before publication.

Financial analysis

The definitions we use throughout, applied identically to every company.

Metric	Our definition
Free cash flow	Operating cash flow less all capital expenditure. Not adjusted for stock-based compensation, which we treat as a real cost.
ROIC	NOPAT divided by (total debt + equity - cash - goodwill). Reported both including and excluding goodwill where the difference is material.
Operating margin	GAAP / IFRS operating income divided by revenue. Non-GAAP margins are shown alongside but never used as the headline.
Net cash / net debt	Cash and marketable securities less total debt, including capitalised operating leases.
Revenue CAGR	Compound annual growth over the stated window, using reported currency. Constant-currency growth is shown separately.
Shares outstanding	Diluted weighted average, with a separate line for the effect of stock-based compensation net of buybacks.

Valuation methodology

We publish valuation frameworks, not price targets. A framework states the assumptions, shows the arithmetic, and gives you a sensitivity table so you can substitute your own view.

Three tools, used in this order:

Historical multiple context. Where the current multiple sits against the company own five-year range – with a warning about whether the denominator is at a normal level.
Reverse-engineered DCF. Rather than forecast cash flows, we solve for the growth rate the current price implies. This makes the market assumption explicit and testable.
Scenario range. Conservative, base and optimistic cases with the single assumption that drives each identified by name.

Data limitations

Filings lag reality

An annual report describes a year that ended months ago. Everything in a report is historical by construction.

Segment data is management choice

Companies choose how to segment. Comparisons across companies with different segmentation are approximate and we say so.

Currency and standards differ

IFRS and US GAAP treat leases, R&D and goodwill differently. Cross-border comparisons carry a note where this matters.

We make mistakes

Corrections are published on the report itself with the date and what changed. We do not silently edit figures.

Conflicts of interest

Fundamenta is funded by reader subscriptions. We do not accept sponsored research, paid placement, affiliate commissions on brokerage products, or advertising from companies we cover.

Analysts disclose personal holdings in any company they write about, at the top of the report, and are prohibited from trading in a covered company for 72 hours either side of publication.

Found an error? Tell us. Send the report name and the figure you think is wrong — we will check it against the filing and publish a correction if you are right.