Methodology
How Arqon scores and values a company
Every company on Arqon carries two numbers: an overall score out of 100, which measures the business, and a valuation grade from A to F, which measures the price. Both are computed by fixed rules, written in advance and applied identically to every company. This page sets those rules out in full — the criteria, the thresholds, the weights, the data, and what the numbers cannot tell you.
Two numbers, two questions
- The overall score answers: is this a good business? It reads the financial statements only. The share price plays no part in it.
- The valuation grade answers: is the price below or above what the business justifies? It is computed separately and never mixed with the score.
The two are independent on purpose. A company can score 90 and be graded F — an excellent business at a demanding price — or score 35 and be graded A, which is more often a warning than a bargain.
The overall score: four pillars of 25 points
The score is the sum of four pillars, each marked out of 25. Each criterion earns points by thresholds: the first threshold the company clears sets its points.
Profitability (25 points)
How much the business earns on its sales and on the capital it uses. Latest annual report.
| Criterion | 5 pts | 4 pts | 3 pts | 2 pts | 1 pt |
|---|---|---|---|---|---|
| Return on equity | > 20% | > 15% | > 10% | > 5% | > 0% |
| Return on invested capital | > 15% | > 10% | > 7% | > 3% | > 0% |
| Net margin | > 20% | > 15% | > 10% | > 5% | > 0% |
| Free cash flow margin | > 15% | > 10% | > 5% | > 2% | > 0% |
| Gross margin | > 60% | > 40% | > 25% | > 15% | > 5% |
Below the last threshold, the criterion scores 0.
Growth (25 points)
Annual growth rates between the oldest and the most recent of the last five annual reports — a four-year span.
| Criterion | 5 pts | 4 pts | 3 pts | 2 pts | 1 pt |
|---|---|---|---|---|---|
| Earnings per share growth | > 30% | > 20% | > 10% | > 5% | > 2% |
| Revenue growth | > 30% | > 20% | > 10% | > 5% | > 2% |
| Free cash flow growth | > 30% | > 20% | > 10% | > 5% | > 2% |
Operating leverage, 10 points: when both earnings and revenue grow, 10 points if earnings grow more than 5 percentage points a year faster than revenue, 8 if faster at all, 6 if no more than 3 points slower, 4 if no more than 10 points slower, 2 otherwise. If only one of the two grows: 2 points. If neither: 0.
A growth rate is not defined when either end of the period is negative (a loss, or negative cash flow). That criterion is then left out, as described below.
Financial health (25 points)
Whether the balance sheet can carry the business through a bad year. Latest annual report.
| Criterion | 5 pts | 4 pts | 3 pts | 2 pts | 1 pt |
|---|---|---|---|---|---|
| Debt to equity | < 0.3 | < 0.6 | < 1.0 | < 2.0 | < 3.0 |
| Current ratio | > 2.0 | > 1.5 | > 1.2 | > 1.0 | > 0.8 |
| Interest coverage | > 10× | > 5× | > 2.5× | > 1.5× | > 1.0× |
| Cash to debt | > 0.3 | > 0.2 | > 0.1 | > 0.05 | > 0.01 |
| Debt to EBITDA | < 2× | < 3× | < 4× | < 5× | < 7× |
A company with no debt scores 5 on cash to debt and debt to EBITDA; one with no interest expense scores 5 on coverage. Negative shareholders’ equity scores 0 on debt to equity, and debt with negative EBITDA scores 0 on debt to EBITDA.
Quality (25 points)
Whether the reported numbers turn into cash, and whether shareholders keep their share of the business.
| Criterion | 5 pts | 4 pts | 3 pts | 2 pts | 1 pt |
|---|---|---|---|---|---|
| Piotroski F-score (0 to 9) | ≥ 8 | ≥ 7 | ≥ 5 | ≥ 3 | ≥ 1 |
| Operating cash flow ÷ net income | > 1.3 | > 1.0 | > 0.8 | > 0.5 | — |
| Change in share count over the period | fell > 5% | fell | rose < 5% | rose < 10% | — |
| Years of positive free cash flow, out of five | 5 | 4 | 3 | 2 | — |
| Change in free cash flow margin over two years | > +3 pts | > +1 pt | > −1 pt | > −3 pts | otherwise |
The cash-to-earnings ratio is only measured when net income is positive. Where a cell shows “—”, anything below the previous threshold scores 0.
When a number cannot be measured
A ratio that is not defined — a growth rate across a loss, a cash-to-earnings ratio on negative earnings — does not score 0. The criterion is left out and the pillar is rescaled over the criteria that could be measured, so that a missing number is not read as a bad one.
The valuation grade: six models, one score
No single valuation model is reliable on its own, so Arqon runs six and weights them. Each model produces a value per share, compared with the current price.
| Model | Weight | What it compares |
|---|---|---|
| Discounted cash flow | 35% | Revenue follows analysts’ consensus estimates; the company’s own margins, investment, tax rate and working capital turn it into free cash flow, discounted at 10% a year. The terminal value averages 2.5% perpetual growth and an exit EV/EBITDA multiple (between 5× and 30×). |
| Earnings and cash flow projection | 20% | Analysts’ earnings estimate for the furthest year available, valued at the company’s median historical P/E, alongside free cash flow valued at its median P/FCF — both multiples capped at 30 — and discounted back to today at 10%. |
| P/E against its own history | 15% | Today’s price-to-earnings ratio against the company’s median. |
| P/FCF against its own history | 15% | Today’s price-to-free-cash-flow ratio against the company’s median. |
| P/B against its own history | 10% | Today’s price-to-book ratio against the company’s median. |
| EV/EBITDA against its sector | 5% | Today’s multiple against a reference for the sector, from 7.5× for energy to 18× for technology. Not used for banks and real estate. |
Each model’s gap between value and price becomes a score from 0 to 100 along a smooth curve: 50 when the price equals the model’s value, about 65 at 10% upside, about 80 at 23% upside, and symmetrically below 50 when the price is higher. The six scores are averaged with the weights above; when a model cannot run, its weight is shared among the others.
| Grade | Valuation score | Roughly, for a single model |
|---|---|---|
| A | 80 – 100 | about 23% upside or more |
| B | 65 – 79 | about 10% to 23% upside |
| C | 50 – 64 | from no upside to about 10% |
| D | 35 – 49 | a value up to about 10% below the price |
| F | 0 – 34 | a value more than about 10% below the price |
The right-hand column is indicative: the grade comes from the weighted average of six scores, not from a single gap between one value and the price.
When there is no grade
A company without positive earnings, without a usable price history and without a positive cash flow projection cannot be valued by any of the six models. It gets no grade rather than a guessed one: a missing grade is more honest than a false one.
How often the numbers change
- Prices: every weekday, before the US market opens.
- Valuation grades: recomputed every Saturday, with that day’s prices. Between two runs, the grade does not follow the price.
- Overall scores: recomputed four times a year, after each earnings season — early March, mid-May, mid-August and mid-November.
Every figure on a company page carries the date it was computed.
The AI analysis
The deep analysis available to members is written by language models from the company’s financial data and its filings with the US Securities and Exchange Commission, including the risk factors of its annual report. It explains the numbers; it does not produce them. The score and the grade come only from the rules above, and the AI cannot change them.
Data sources
- Financial statements, prices, historical multiples and analysts’ estimates: Financial Modeling Prep.
- Annual reports and risk factors: SEC EDGAR.
- Coverage: large US-listed companies. Every company on the stock list is scored by the same rules.
What these numbers cannot tell you
- They look backwards. The score reads the last published annual reports. A business that is changing fast will be scored on what it was.
- The same rules for every sector. Thresholds do not adapt to the industry: a bank’s balance sheet and cash flows do not read like a software company’s, and its scores on those criteria mean less.
- Some criteria behave oddly at the extremes. A company with near-zero profits can show a very high ratio of cash flow to earnings, and a margin recovering from a bad year scores well on its two-year trend.
- A grade is not a forecast. It says how the price compares with the models today. It does not say when, or whether, the price will move.
Arqon is a research tool, not an investment adviser. Scores and grades are not recommendations to buy or sell any security.