Why does the public page not show every internal step?

The public page is meant to explain the research principles and reader-facing checks. The internal routing, prompts, calibration controls, provider wiring, and private benchmarks are proprietary company assets.

Why is the report bear-case-first?

Putting the downside case early forces every recommendation to confront what can break the thesis before the positive narrative takes over.

How should readers use Fair value and Potential value?

Treat Fair value as the evidence-backed midpoint and Potential value as the favorable high endpoint, not as guaranteed target prices or personalized investment advice.

What changed in v3.2?

Versioned the methodology itself, so every report records the exact methodology it was produced under; Made accounting-quality checks source-aware before earnings are trusted; Extended long-horizon financial history and regulatory-filing coverage behind each analysis; Added per-ticker input-coverage scoring so thin evidence is caught before an analysis begins; Blocked silent substitution on valuation-critical fields; Added third-party fair-value cross-checks alongside the house range.

← Metodologia

§ Per-stock methodology v3.2

How every report is produced.

Every report follows a versioned research framework designed to separate business quality from stock attractiveness. We adapt the analysis to the company context, review the evidence quality, triangulate valuation through multiple lenses, stress the assumptions, and write the risks before the final synthesis. This page explains the public principles and reader-facing checks; the internal orchestration, prompts, calibration controls, provider wiring, and private benchmarks remain proprietary.

Report languages

Choose a language from the report page to read an available translation. Translation availability varies by report. If a translation is unavailable, the page identifies the original language.

Step 0

Business context

We first identify the company context: maturity, industry structure, cyclicality, balance-sheet strength, reinvestment needs, and the type of evidence that should matter most.

Step 1

Evidence gathering

We gather filings, financial statements, market data, peer context, and company disclosures, then normalize the inputs so the report is comparable and auditable.

Step 2

Return hurdle

We estimate the return hurdle a stock must clear using market context, balance-sheet risk, business durability, and company-specific uncertainty.

Step 2.5

Financial-quality review

Before trusting the earnings base, we review cash conversion, margin quality, capital intensity, balance-sheet resilience, and accounting signals.

Step 3

Valuation triangulation

Comparamos várias perspectivas independentes de avaliação para que nenhum múltiplo, fórmula ou narrativa determine o valor justo. Os cartões públicos dos modelos mostram os dados, o cálculo e as limitações de cada abordagem.

Step 4

Sensitivity analysis

We pressure-test the valuation against changes in growth, profitability, discounting, terminal assumptions, and downside scenarios.

Step 5

Bear-case-first risk review

The report explains what can break the thesis before it argues for upside, reducing the risk that a bullish story overpowers the evidence.

Step 6-7

Durability and monitoring

We connect business durability, competitive pressure, capital allocation, upcoming catalysts, and monitoring checkpoints to the valuation view.

Step 8

Decision overlay

The final rating separates business quality, valuation attractiveness, and risk so a great company is not automatically treated as a great stock.

Step 9

Publication and translation

Reports are published with versioned methodology markers, reader-facing assumptions, consistent numbers across locales, and release checks before they appear on the site.