Methodology
This page explains what the DecodeStocks evaluations rest on – which data is used, how individual metrics are placed and combined into scores, and what those scores explicitly do not say.
Where the data comes from
DecodeStocks uses published company data, regulatory filings and data from specialised financial data providers.
The available data differs by company, market and data type. Which analysis areas and which level of detail exist for a company therefore depends on that company — coverage and depth are not the same everywhere.
Prices: We obtain stock prices and daily closing prices from Twelve Data (twelvedata.com). The prices are not supplied in real time. From the price data we calculate, among other things, valuation metrics and scores. These calculations are made by DecodeStocks.
Financial reports: We obtain financial data from published company reports and further data sources. For US-listed companies we use in particular the reports published via SEC EDGAR for this purpose. As a supplement, and for companies in other markets, we obtain financial data from Twelve Data. Evaluations, conversions and scores are made by DecodeStocks.
For currency conversions we use the euro foreign exchange reference rates of the European Central Bank (source: ECB). From them we calculate rates between other currencies and average rates for reporting periods; these conversions are made by DecodeStocks, not by the ECB. The reference rates are available free of charge on the ECB website at www.ecb.europa.eu. For the following currencies, for which the ECB publishes no reference rate, we convert with daily closing rates from Twelve Data (twelvedata.com): Argentine peso, Chilean peso, Colombian peso, Kazakhstani tenge, Peruvian sol, Russian ruble, Taiwan dollar, UAE dirham and Vietnamese dong.
How the scores are formed
DecodeStocks places selected financial metrics according to a defined, rule-based methodology and combines them within several analysis areas into category scores.
Category scores run from 0 to 100. They show only how the metrics taken into account are placed within the DecodeStocks methodology.
The metrics score is calculated from the available category scores using a defined weighting. That weighting can be adapted to the company’s business model. Dividends are assessed as an area of their own and enter the metrics score only where the company profile assigns them a weight.
Many metrics are placed relative to a suitable peer group rather than against a fixed threshold. A narrower industry group is used where possible, and the parent sector otherwise.
Missing metrics are not substituted with 0 or with assumed neutral values. A score is calculated only where the methodology has enough data to work with. Where the dividend cannot be placed reliably while the core sections meet the methodological prerequisites, the score is still published, and the dividends entry then shows a “–”. That holds both where no dividend rating exists and where one was calculated that does not enter the total score, because the company's dividend status is not sufficiently established. Such a score is not comparable with scores that do include a dividend rating.
The metrics score condenses several analysis areas into one overall value. The category scores are additionally reported separately, so differences between the individual areas stay visible.
What a score does not say
A score assesses only the data taken into account within the DecodeStocks methodology. It is not a statement about future price performance, expected return or an appropriate purchase or sale price.
A higher score does not automatically mean a stock is “better” overall or more suitable for you. The scores are not a buy, sell or hold recommendation.
Where the required data for a company is missing, the value in question is not reported. A missing value is a statement about the available data, or about whether the methodology applies at all — not automatically a negative assessment of the company.
DecodeStocks does not provide personalised investment advice. Personal investment objectives, risk tolerance, capacity to bear losses and financial circumstances are not taken into account in the analyses shown.
Detail on individual metrics and scores
This methodology page explains how our scores are formed and how values are placed in principle. You will find the detailed calculation rules and the particulars of individual metrics directly in DecodeStocks, via the info icons where they apply.
That is also where you will find further information on the individual analysis areas, peer groups, missing values and particular calculation rules.
The same explanations, gathered and ordered by the areas of the metrics score, are on the page “Metrics explained”.
How the AI analysis is created
The AI analysis processes company data and evaluated documents from SEC EDGAR. The analysis lists these documents with their form, date and source link. This list identifies the evidence base, rather than substantiating each individual sentence.
The analysis is generated in German and English. The as-of date shows when its inputs were last checked. New publications may not be reflected in the analysis immediately.
AI-generated text can be incomplete or incorrect. Check important statements against the linked original documents. The analysis is not investment advice or a recommendation to buy, hold or sell securities.