Chevron's FY2025 10-K (filed February 24, 2026) scored 71.5 on our language signal — 1.4× the distress ceiling. Chevron is a $330 billion company, a 37-year dividend grower, and trades near an all-time high. So why did its filing language move like a distressed company's? Because the $53 billion Hess acquisition closed, and the FY2025 10-K is the first to absorb Hess's risk factors and the deal's financing boilerplate. This is a structural false positive — and a clean illustration of what the score measures and what it doesn't.
The score is built from directed, year-over-year change in filing language, weighted corpus-wide — not "did this company change a lot" but "did it change in ways that were unusual across the whole corpus this year." For Chevron's FY2025 10-K, the change was large and concentrated in one place: the risk-factors section roughly quintupled, from about 4,600 words to about 24,700 (+438%). The phrases driving the score:
| Phrase | Change | Score contribution |
|---|---|---|
| "default" | 2 → 8 occurrences (+300%) | 15.3 |
| "credit facility" | first appearance (×4) | 11.7 |
| "material adverse" | first appearance (×5) | 10.3 |
| "material adverse effect" | first appearance (×4) | 8.3 |
Read that list as a credit analyst would and it looks alarming: "default," "credit facility," "material adverse effect" all escalating at once. Read it as a deal lawyer would and it looks routine: this is exactly the vocabulary a company inherits when it consolidates a large, debt-financed acquisition.
In July 2025 Chevron closed its $53 billion acquisition of Hess Corporation, after an ICC panel rejected ExxonMobil's pre-emption claim over Hess's 30% stake in Guyana's Stabroek block. The FY2025 10-K is the first annual filing to consolidate the combined entity. That does several things to the risk-factors section at once:
A five-fold expansion of the risk-factors section, dominated by first-appearance credit and material-adverse language, is what that looks like quantitatively. The score is doing its job — the language genuinely moved, and it moved in a corpus-unusual way. But the cause is a balance-sheet expansion from a position of strength, not deterioration.
A score above the ceiling means the language is statistically unusual — it has moved more than the 95th percentile of healthy companies typically move in a year. It is not a prediction. Most flag events resolve without a distress outcome; the signal's edge is statistical (moderate-flagged companies reach a distress outcome about 1.2× the corpus base rate, and the signal recalls 75% of labeled crisis companies from pre-event filings), not a verdict on any single name. Chevron is the clearest kind of non-distress flag:
This is the merger pattern — the same structural language change we've seen at RTX (Raytheon–United Technologies) and others: a big corporate event rewrites the filing without any underlying distress. It's exactly the failure mode worth publishing, because it shows where the signal's mechanism (raw language divergence) and its intent (early distress) come apart.
It is worth being precise about what a score of 71.5 has meant historically. The table below shows companies whose scores landed in a similar range and what happened next — both true signals and structural false positives.
| Company | Score | Outcome | Lead / lag |
|---|---|---|---|
| Bed Bath & Beyond (BBBY) | 138.5 | Bankruptcy Apr 2023 | ~2 yrs before |
| Nikola (NKLA) | 85.3 | Bankruptcy Nov 2023 | ~3.7 yrs before |
| RTX Corporation | Above-ceiling score during Raytheon–United Technologies merger — structural language change, no distress event | ||
| Chevron (CVX) | 71.5 | Hess acquisition consolidated · stock at all-time high · structural false positive | Filed Feb 2026 |
Chevron belongs with RTX, not with BBBY or Nikola. The score captured genuine, large-scale language change — it just wasn't distress. That distinction is the whole point of publishing it.
The signal measures language divergence, not corporate health. Most of the time those track together — a company writing materially different risk factors is often a company under stress. But not always. A large acquisition rewrites the filing too. The honest version of this product flags Chevron, shows you why, and lets you see in thirty seconds that the cause is a merger, not a meltdown. The same score's other end — the long side, where the most-stable-language quintile earns 93.2 bps/month (t=9.16) under FF5+QMJ, with a 59.5 bps/mo FF3 long-short — is where the durable edge lives (on a survivors-only universe — correcting for delistings concentrates that edge in small-caps, and it inverts above ~$300M). The distress end is a screen, not an oracle, and Chevron is what a false positive looks like.
Not investment advice. This post presents publicly available data and the output of our scoring system. It does not recommend buying, selling, or holding any security.
Questions or corrections: hello@filingdrift.com