Quality Methodology

1. What this is

The Quality row on every SignalEvo signal is a fundamentals screen we run against the company’s SEC filings. It’s shown as an informational badge — green pass, amber partial, red fail, grey for sectors we don’t evaluate.

The screen does not change the trading verdict and does not change the confidence number. Those numbers come from the technical pipeline, untouched. The Quality row is a separate lens you can choose to weight, ignore, or override entirely.

2. Why two lenses

SignalEvo’s trading signals are short-term technical setups with a 2–20 day horizon. They look at price structure, momentum, volume, and order flow. They answer the question: is now a good time to enter or exit?

The quality screen is a long-term durability check with a 10-year horizon. It looks at profitability, balance sheet strength, cash generation, growth, and competitive position from audited annual filings. It answers a different question: is this a business you’d be comfortable holding through a drawdown?

Both questions matter. We display them separately so you can decide which weight to give each. A high-confidence technical buy on a company that fails quality is a different trade from the same signal on a company that passes — but only you can decide what to do with that information.

3. The six criteria

Each criterion runs against the company’s most recent ten fiscal years of audited 10-K filings. Each returns one of three states: passed, failed, or unmeasurable (when the underlying data tag isn’t in the filer’s XBRL).

Profitability

We measure positive net income across the last ten fiscal years. The threshold is at least eight years positive out of ten. Filers using non-standard tagging fall back through several net-income concepts; if eight or more years still can’t be parsed, the criterion returns unmeasurable rather than a hard fail.

Return on Equity

We compute the average return on equity across the last five fiscal years. The threshold is 15% for general and cyclical companies, lowered to 10% for financials. Financials structurally run lower ROE because of their capital requirements, so 10% is the quality-bank bar. Years with negative or zero equity are skipped from the average — a ratio with negative equity is mathematically meaningless.

Leverage

We measure long-term debt to equity in the most recent fiscal year. Primary thresholds are 1.0 for non-financials and 2.5 for financials. If a company fails the primary check, we apply a net-debt rescue: subtracting cash and all marketable securities from long-term debt before dividing by equity. The rescue thresholds are 0.5 non-financial and 1.5 financial. A separate cash-rich-no-LTD branch passes filers with positive equity and cash but no reported long-term debt.

When the primary fails and cash data is genuinely missing from the filings, the leverage criterion is reported as unavailable rather than failing on absent data.

Free Cash Flow

We require at least 80% of parseable years to be positive across the last ten fiscal years, plus a non-negative ten-year compound annual growth rate. Years where operating cash flow or capital expenditures can’t be parsed are excluded from the denominator — if fewer than eight of ten years are parseable, the criterion returns unmeasurable. For financials we substitute positive operating income, since FCF is not a meaningful concept for a bank.

Revenue Growth

We require at least a 2% compound annual growth rate over ten fiscal years. The bar is intentionally low — slow-compounding defensives like Coca-Cola or Johnson & Johnson don’t need to grow fast to be high-quality businesses, but they do need to be growing.

Moat

A company passes the moat criterion if any of three branches hold:

• Pricing power — gross margin standard deviation across the last five years is at most 5 percentage points (indicating stable pricing through cycles).
• High margin — five-year average operating margin is at least 15%.
• Consistent compounder — five-year average ROE is at least 15%, with ROE standard deviation at most 5 percentage points, AND the leverage criterion also passes. The leverage guard prevents a buyback-heavy levered company from manufacturing high ROE through financial engineering.

4. The five verdicts

5. Sector adjustments

We classify each company by SIC code into one of four buckets and apply different thresholds where the sector structurally requires it.

• Financials (SIC 6000–6999, excluding REITs): ROE threshold lowered to 10%, leverage primary threshold raised to 2.5, leverage rescue raised to 1.5, FCF criterion swapped for positive operating income.
• REITs (SIC 6798): excluded entirely from the screen in v1.
• Cyclicals (mining, chemicals, primary metals, motor vehicles): annotated but use the general thresholds. Their cycle-driven earnings volatility shows up as criterion failures, which is the honest signal.
• General: everything else. Standard thresholds.

6. Data source

All inputs come from the SEC EDGAR XBRL company facts API. EDGAR is the public, audited filings repository every US public company is required to submit to. We use the structured XBRL data from 10-K filings — the same numbers in the same form the SEC sees.

We don’t use third-party data resellers, paid feeds, or estimated values. Each snapshot is cached for seven days to keep EDGAR request volume polite (the SEC asks for fewer than ten requests per second; we ceiling at nine).

When XBRL tagging differs between filers (a known SEC data quality issue), we fall through several alternative tag concepts before declaring a value unmeasurable. The criteria explanations above note where this matters.