How ValScreen scores stocks
ValScreen turns available company fundamentals into a consistent, peer-relative research signal. This page explains the inputs, comparisons, missing-data rules, and limitations behind the score.
Methodology reviewed September 2026. Scores are research filters—not recommendations, price targets, or predictions.
The calculation
Four steps from raw data to a 0–100 score
Validate the inputs
Missing, non-numeric, and invalid ratio values are excluded before scoring.
Choose the peers
Each usable ratio is compared with the stock's sector when at least 30 usable sector observations exist. Otherwise, the supported universe is used.
Score each ratio
A percentile shows where the value sits among its peers. Direction is adjusted so a higher score always represents the stronger screen result.
Combine the results
Ratio scores form category scores, and available categories form the composite score. Unavailable inputs do not receive a zero.
The four scoring categories
The official model emphasizes valuation while still requiring context from profitability, financial strength, and cash generation.
Valuation
40%P/E, EV/EBITDA, P/B, and P/S. PEG is an optional input and is disabled by default.
Profitability
25%ROIC, return on equity, return on assets, operating margin, and gross margin.
Financial strength
20%Debt-to-equity, net debt/EBITDA, current ratio, and interest coverage.
Cash flow
15%Free-cash-flow yield, operating-cash-flow margin, and free cash flow/net income.
When an enabled ratio is unavailable, the other usable ratios in that category share its influence. If an entire category is unavailable, the remaining usable categories share its composite weight. This prevents missing information from being treated as poor performance, while the data-quality label keeps the reduced coverage visible.
Peer comparison and outliers
Different industries naturally carry different margins, debt levels, and valuation ranges. ValScreen therefore compares each ratio with sector peers when the peer set is large enough. If fewer than 30 usable sector observations exist for that ratio, the broader supported universe is used and disclosed.
Values beyond the 5th and 95th percentiles are limited to those boundaries during scoring. This reduces the ability of a few extreme observations to distort every company's relative position; it does not alter the raw value shown to users.
How PEG is handled
PEG compares a company's P/E ratio with its earnings-growth rate. It is disabled in ValScreen's default model because it is not meaningful for every business and is frequently unavailable or unsuitable.
A user may enable PEG in model settings. It is scored only when both P/E and earnings growth are positive. Otherwise, PEG is omitted and the remaining usable valuation ratios share its weight.
Data quality describes coverage—not company quality
A Complete label does not mean the business is strong or that every provider value is correct. It means the selected scoring model had complete usable coverage.
- Complete
- Every ratio enabled in the selected model and all four categories were usable.
- High
- Broad ratio coverage was available across all four categories.
- Medium
- Enough information was available to score the stock, with meaningful omissions.
- Low
- The stock was scoreable, but the result relies on relatively limited inputs.
- Insufficient
- Fewer than five usable ratios or fewer than two usable categories were available.
How to read the score
- 70–100: Screens Strong vs. Sector
- 40–69: Screens In Line with Sector
- 0–39: Screens Weak vs. Sector
- Insufficient: No overall score is produced.
These labels describe the mathematical result only. A strong screen is not a claim that a security is undervalued, suitable, safe, or likely to perform well.
Data source and score date
ValScreen uses licensed third-party financial and market data provided by Tiingo. The score date displayed for each stock identifies the data date associated with its calculation. Source data can be delayed, incomplete, revised, or inaccurate.
Important figures should be checked against primary materials such as company filings before they are used in a financial decision.
What the model deliberately does not do
The score does not evaluate personal goals, portfolio fit, taxes, risk tolerance, time horizon, management quality, competitive durability, every recent event, or future market prices. AI Insights can organize sourced business context, but those reports remain informational research—not individualized advice.