The Gunning Fog Index
A readability score invented by Robert Gunning in 1952. It estimates the years of schooling needed to understand a passage on first reading. Newspapers target around 12. Investor letters from Buffett land near 14. Mega-cap 10-Ks routinely score 18-20. A Fog of 22 or higher is the signature of language calibrated to be technically accurate and substantively unreadable.
The formula is 0.4 · (avg words per sentence + percent of complex words), where a complex word is three or more syllables, excluding proper nouns and common verb endings. The math is trivial; the interpretation is everything.
Why this matters in a 10-K
Long sentences and Latinate vocabulary aren’t just stylistic. The forensic-accounting literature — particularly Loughran and McDonald’s 2014 paper on financial readability — has shown that material adverse disclosures cluster in passages with elevated Fog scores. Companies don’t fabricate disclosures; they bury them under syntax.
The score by itself is descriptive, not diagnostic. A high Fog can mean the company is in a regulated industry where careful language is required (pharma, defense, financial services). It can also mean management is intentionally elevating cognitive load on the reader. Reading the MD&A passage that produced the score is what tells you which.
Calibrating expectations
We use SP500 baseline thresholds, not the original Gunning thresholds. The reality is that mature operating companies file densely-written 10-Ks as a matter of course. We score:
- Below 16: plain English. Rare for a public company filing.
- 16-19: normal. The mid-range where most well-prepared 10-Ks land.
- 19-22: dense. Worth a closer read, but not a flag.
- 22 and above: obfuscatory. The MD&A is doing more work hiding things than explaining them.
Worked example
NVIDIA’s most recent Fog is 21.7 — dense but inside the normal mega-cap range. Apple’s is 19.6 — squarely in the normal band. ChargePoint is too short on disclosure to compute a stable score in our pipeline. The companies that consistently score above 22 in our coverage tend to be the ones in the middle of restructurings, regulatory investigations, or going-concern discussions.
What to take away
Treat the Fog Index as a navigation tool. A high score tells you which 10-Ks are worth reading carefully and which paragraphs to slow down on. It does not, by itself, tell you anything about the underlying business — but the prose patterns that elevate the score are the same prose patterns that historically precede write-downs.