How we score

Augur analysis brings together five separate assessments of a company. An AI model writes each assessment using a structured prompt and the inputs supplied to that section. The overall score is calculated from the section scores, rather than chosen by the model as a separate recommendation.

This describes the current five-dimension method. Older published analyses may have been produced under earlier versions. Read the date and context of an analysis before comparing it with another.

The five dimensions

Financial health

Can the company meet its obligations and support its operations? This section examines the balance sheet, cash generation, leverage and liquidity where those measures suit the business model. It considers the current position as well as the direction of change.

Growth

How quickly, consistently and on what basis is the business expanding? This section considers the available revenue and earnings history, comparability between periods and the quality of the reported growth. It distinguishes a meaningful growth rate from a distorted calculation on a negative or very small starting value.

Profitability

How effectively does the company turn revenue and capital into profit? For non-financial companies, this includes the relationship between gross, operating and net margins, returns on assets and equity, and available cash-flow evidence. The business model and the company's own history matter.

Business quality

What does the available evidence say about the company's competitive position and the forces affecting its business? This section uses dated, cited reporting rather than inferring competitive strength from financial ratios alone. Company disclosures can also be included where that evidence channel is enabled.

If the evidence is too thin to support an assessment, the current method can retain a neutral score and explain that the business could not be assessed. A neutral result in that situation is not evidence of an average-quality business.

Valuation

What does the current price imply relative to the reported business? This section uses available price or market-value inputs and supported earnings or book-value measures. It must not invent comparable companies, industry averages or a meaningful multiple when the underlying earnings or other denominator do not support one.

A valuation score is not a price target or an instruction to trade.

Scale and overall score

The active scoring contract allows section scores from 0.0 to 5.0, normally stated to one decimal place. Its reference points are:

• 5: exceptionally strong on that dimension.
• 3: adequate on that dimension.
• 1: materially weak on that dimension.
• 0: severe or distressed conditions supported by the inputs.

Higher scores indicate a stronger assessment on the named dimension. They are not probabilities, expected returns or buy and sell recommendations.

Each of the five dimensions has an equal 20% weight when all are assessed. A missing section or a section with no score is excluded from both the total and the available weight; it is not counted as zero. The remaining weights are normalised for the overall calculation.

The stored average can have more precision than the one-decimal score shown on a card. Methodology versions and the mix of assessed sections can affect comparisons.

Inputs and evidence

The financial sections use reported company financial information: annual reports and other available filings, extracted financial statements, ratios derived from those figures and, where supplied, interim results. Valuation also uses end-of-day market information.

The prompts require figures to be present in, or directly calculable from, the supplied inputs. They do not authorise invented forecasts, peer averages or unsupported explanations for changes.

Business Quality is the exception to the financial-data-only boundary: its assessment relies on dated, cited qualitative evidence. Risks, catalysts and the summary are then written from the completed sections; they do not add a sixth score.

Source documents are linked from each company's Documents tab. See data sources and update schedules.

Banks and other business models

Banks are not scored as if deposits were ordinary industrial borrowing. Financial health uses available capital and asset-quality measures. Profitability uses measures such as net interest margin, cost-to-income, return on assets and return on equity where supplied.

Industrial current ratios, debt-to-equity measures and the standard gross-to-operating-margin ladder do not apply to banks in the same way. Their absence is not, by itself, evidence of weakness.

For non-financial businesses, the analysis uses relevant balance-sheet, cash-flow, margin and return measures. The prompts also account for differences such as prepaid or subscription revenue and industry-specific economics. A measure that does not apply to a business is different from an applicable measure that is missing.

Updates and availability

The configured weekly analysis job checks for changed inputs. It does not rewrite every analysis each week, and processing limits mean that not every eligible company is updated in a single run.

Sections track their own inputs. For example, a price change does not by itself require a new growth assessment. The current valuation policy generally waits for a price move of at least 10% from the price recorded at the last valuation run, with separate handling when no usable prior price exists.

Always check the analysis and data dates. A scheduled check is not a guarantee that every report is current.

What "Soon" means

In the score bars, "Soon" appears when that view has not received a score for a dimension. For valuation, this does not mean a zero score or a promised publication date.

"N/A" means that a section is present but unassessed. Missing or unassessed valuation is not silently treated as neutral in the overall calculation.

Limitations

AI-assisted analysis can contain errors, omissions or misinterpretations. Extracted figures can be incomplete, source data can be delayed and company circumstances can change after publication.

Historical results and scores do not predict future performance. The service does not consider your objectives, financial situation or ability to bear losses. It is not investment advice. If something looks wrong, report it to us.