An empirical read of 57 first-wave filings on CARB’s voluntary docket: most name the hazard, almost none put a number on it.
California’s climate-related financial risk law, Senate Bill 261, set its first reporting deadline for January 1, 2026. For the first time, a large group of US companies had to describe, in public, how climate change threatens their business. The reports are now landing. This piece reads 57 of them and reports what the first wave actually did.
One finding sits above the rest. Almost every report names physical hazards as a risk. Almost none attaches a number to one. About 96 percent disclose no asset-level quantification of physical risk: no dollar figure, no share of assets, tied to a flood, a fire, or a heatwave. That share is stable even if every borderline coding call were reversed. The gap between naming a hazard and measuring it is the single clearest pattern in the data, and it is where the next reporting cycle will apply pressure.
How we read the first wave
The 57 reports come from a voluntary docket. After a court paused enforcement in late 2025, the California Air Resources Board (CARB), the state agency that implements the law, opened a public docket where companies could submit reports while the pause held. Our cohort is the set of substantive reports posted to that docket as of May 28, 2026.
This is a voluntary-docket cohort, not a census. The law reaches US-formed companies with more than $500 million in annual revenue that do business in California, a population far larger than this voluntary first wave. Companies that submit early are a self-selected group. So the percentages below describe what these 57 filers did. They do not estimate what the full covered population will do.
We reviewed 59 reports and retained 57. Two were dropped because the posted file was a landing page, not the report itself. For each retained report we coded seven things: the framework named, the depth of physical-risk treatment, the use of scenario analysis, third-party assurance, the hazards named, the sector, and the page length. Every coded value carries a quoted snippet from the report and a page locator. The full method sits at the end of this piece.
The cohort spans ten sectors. Technology, healthcare, and manufacturing are the largest blocks. Retail, financial services, energy, real estate, transportation, and professional services fill out the rest.
| Sector | Reports | Share |
|---|---|---|
| Technology | 11 | 19% |
| Healthcare | 10 | 18% |
| Financial services | 3 | 5% |
| Real estate & construction | 3 | 5% |
| Manufacturing & industrial | 10 | 18% |
| Energy, chemicals & utilities | 4 | 7% |
| Retail, consumer & hospitality | 9 | 16% |
| Transportation & logistics | 2 | 4% |
| Professional services | 3 | 5% |
| Other | 2 | 4% |
Sector distribution of the retained full-report cohort. n=57 · voluntary-docket cohort (CARB SB 261), not a census · as-of 2026-05-28
Report length varies widely. Many are short, standalone documents written only for this law. A long tail are broad sustainability or annual reports submitted as the climate filing, which run much longer for reasons unrelated to physical risk.
| Page count | Reports | Share |
|---|---|---|
| 1-8 | 14 | 27% |
| 9-15 | 17 | 33% |
| 16-30 | 8 | 16% |
| 31-60 | 5 | 10% |
| 61+ | 7 | 14% |
Page count of the physical-risk report (PDF-sourced reports only). 6 HTML-sourced reports excluded (no page count); pages-n=51 · n=57 · voluntary-docket cohort (CARB SB 261), not a census · as-of 2026-05-28
Most reports still cite a retired framework
SB 261 does not write its own disclosure template. It points reporters to the framework of the Task Force on Climate-related Financial Disclosures (TCFD), the four-part climate reporting structure most large companies already know, “or any successor thereto.” That pointer matters because the task force was wound down in 2023, and its work passed to the International Sustainability Standards Board (ISSB), whose IFRS S2 standard is the named successor.
So a first-wave filer had a choice of label: classic TCFD, the newer IFRS S2, or both. The cohort came down hard on the older name. About 77 percent cite TCFD only. About 21 percent use a hybrid label that names both TCFD and IFRS S2. Just 1 of 57 named IFRS S2 on its own.
| Framework | Reports | Share |
|---|---|---|
| TCFD-only | 44 | 77% |
| IFRS-S2 | 1 | 2% |
| hybrid | 12 | 21% |
Stated reporting framework, coded at company level from the report body. n=57 · voluntary-docket cohort (CARB SB 261), not a census · as-of 2026-05-28
In plain terms: a year after the framework was retired, the first wave still files under its name. The newer standard shows up mostly as reinforcement, not as a replacement. Whether one IFRS S2 report can stand in for an SB 261 report is a live question with its own answer, which we walk through separately in Can an IFRS S2 report satisfy California SB 261?. The cohort suggests few filers have tested that route yet. We ran the same read for IFRS S2 global filers separately; the pattern rhymes.
Physical risk is named everywhere, quantified almost nowhere
This is the center of the data. We sorted each report into one of three depth levels. A qualitative-only report describes physical risk in words, with no number attached to any hazard. A mixed report ties at least one number to a hazard, such as the share of sites in a flood zone or a numeric risk score. An asset-level report gives a site or portfolio figure, in dollars or share of asset value, modeled against a hazard.
About 74 percent of reports are qualitative only. About 23 percent are mixed. Only 2 of 57, about 4 percent, reach asset-level. Combine the first two groups and roughly 96 percent of the first wave disclose no asset-level number for physical risk.
| Physical-risk depth | Reports | Share |
|---|---|---|
| qualitative-only | 42 | 74% |
| mixed | 13 | 23% |
| asset-level-quantitative | 2 | 4% |
Depth of physical-risk treatment. ‘mixed’ = a disclosed number tied to a hazard; ‘asset-level’ = site/portfolio $ or % asset value at risk. GHG/energy metrics do not count. n=57 · voluntary-docket cohort (CARB SB 261), not a census · as-of 2026-05-28
The hazard counts make the gap vivid. Naming hazards is near universal. About 84 percent name extreme heat. About 79 percent name flood. Wildfire, hurricane, and drought follow at 63, 53, and 46 percent. Reporters know which perils matter to them and say so plainly.
| Hazard | Mentions | Share |
|---|---|---|
| wildfire | 36 | 63% |
| flood | 45 | 79% |
| extreme-heat | 48 | 84% |
| drought | 26 | 46% |
| sea-level-rise | 16 | 28% |
| storm/atmospheric-river | 21 | 37% |
| hurricane/cyclone | 30 | 53% |
| water-stress | 25 | 44% |
| landslide | 2 | 4% |
| cold/winter | 6 | 11% |
Count of reports naming each hazard as company-relevant in Strategy/Risk-Management (multi-select; 56 of 57 named ≥1 hazard). multi-select: sums to total mentions, not n · n=57 · voluntary-docket cohort (CARB SB 261), not a census · as-of 2026-05-28
What almost none of them do is size the exposure. The two reports that reach asset-level show what the others skip. One, a Japanese industrial manufacturer, disclosed a modeled site-level expected-damage figure. The other, a New Zealand freight operator, disclosed a property-level damage estimate across its sites. Both are foreign-headquartered filers reporting to home-market standards that ask for the number. Among the US-formed filers the law actually targets, asset-level physical-risk quantification is absent from our sample.
This is not a coding accident. SB 261 inherits its content rules from a framework that asks for the at-risk number but, through the statute, leaves the strict version optional rather than required. How that at-risk metric travels across the major climate frameworks, and why SB 261 treats it as a pointer rather than a hard test, is its own story, told in Assets at Risk: the at-risk metric across four frameworks. The empirical result here is simple. When the number is optional, the first wave mostly does not produce it.
Scenario analysis is mostly narrative
Scenario analysis asks a company to test itself against a future climate pathway, such as a warming level or an emissions track. We coded whether each report ran one and what it produced. Some name no scenario, or only a planned one. Others sit in a qualitative narrative, where a scenario is named and discussed but yields no number. A few are modeled, where a named scenario produces a numeric physical-risk result.
About 68 percent sit in the narrative middle. They name a pathway and describe directional effects, without a quantified output. About 16 percent reach a modeled result. About 16 percent run no scenario at all.
| Scenario treatment | Reports | Share |
|---|---|---|
| none | 9 | 16% |
| qualitative-narrative | 39 | 68% |
| modeled-quantified | 9 | 16% |
‘modeled-quantified’ = named scenario yields a disclosed numeric physical output; ‘qualitative-narrative’ = scenario named but output qualitative; ‘none’ = no scenario or planned-only. n=57 · voluntary-docket cohort (CARB SB 261), not a census · as-of 2026-05-28
In plain terms: most reporters have adopted the vocabulary of scenario analysis without the math. A named pathway with a paragraph of narrative reads as compliant on a quick skim, but it does not tell a reader how much is at stake under that pathway.
Assurance stops short of physical risk
Third-party assurance is an outside check that a disclosure is reliable. Limited assurance is a lighter review; reasonable assurance is the stronger, audit-grade level. We recorded whether each report carried completed assurance and what it covered.
About 67 percent carry none. About 28 percent carry limited assurance, and about 5 percent reasonable. The scope finding matters more than the split. Where assurance is present, it covers greenhouse gas or operational metrics. In this cohort, no report carried assurance scoped to its physical-risk analysis.
| Assurance | Reports | Share |
|---|---|---|
| none | 38 | 67% |
| limited | 16 | 28% |
| reasonable | 3 | 5% |
Completed third-party assurance only (forthcoming excluded). Where present, scope is GHG/operational metrics, not physical risk. assurance scope is GHG-only except where noted · n=57 · voluntary-docket cohort (CARB SB 261), not a census · as-of 2026-05-28
In plain terms: the part of the report a reader might most want checked, the physical-risk strategy, is the part no outside reviewer has signed off on. Assurance, where it exists, sits on the emissions tables next door.
What the first wave signals for the next cycle
SB 261 reports are due every two years, so the next cycle points to January 1, 2028. Enforcement is paused for now: a federal appeals court stayed the law in late 2025, and CARB’s advisory says further information, including an alternate reporting date as appropriate, will follow once the appeal resolves. As of June 16, 2026, no ruling has issued and the pause holds. Status and deadlines are tracked on our California climate disclosure hub.
The pattern in this cohort points to where the pressure lands next. A qualitative-only report satisfies a first, lenient cycle. It is weaker ground as peers, investors, and reviewers start comparing reports that are now public and searchable. The reports that named hazards but never sized them have set a low bar that a second cycle may raise.
Closing that gap is an analysis problem before it is a writing problem. A defensible physical-risk section rests on named sites, named hazards, scenarios, and time horizons, carried through to a number. Our worked examples show what that looks like end to end, for a manufacturer and for a retailer with a distributed footprint. They are the reference a first-wave filer can measure a second-cycle upgrade against.
See the number the first wave skipped
Our worked SB 261 disclosures show a defensible physical-risk section end to end, for a manufacturer and a distributed retailer: named sites, named hazards, scenarios, and a quantified loss figure.
How we coded the cohort
The cohort is the set of substantive reports on CARB’s SB 261 voluntary docket as of May 28, 2026: 59 reviewed, 57 retained, 2 dropped as non-reports. Each report was coded from its own posted file against a written rubric with mutually exclusive buckets. Every coded value carries a quoted snippet and a page locator from that report. The two judgment dimensions, depth and scenario, were re-checked against one consistent rule so the boundary is applied the same way across all 57. The headline that about 96 percent lack asset-level quantification combines the qualitative-only and mixed groups, and does not move if borderline calls are reversed, because only the 2 asset-level reports carry a site-level number.
All charts are aggregate. No company is named, and the two asset-level reports are described by sector and country only. Each chart is footnoted with its sample size, the as-of date, and the note that this is a voluntary-docket cohort, not a census.
Sources
- California Health and Safety Code § 38533 (SB 261, Stats. 2023, ch. 383; amended by SB 219, Stats. 2024; AB 154, Stats. 2025, ch. 609). Verbatim text: SB 261 standard text.
- California Air Resources Board, SB 261 voluntary reporting docket (opened December 1, 2025); SB 261 enforcement advisory (December 1, 2025).
- Task Force on Climate-related Financial Disclosures, Final Report: Recommendations (June 2017); IFRS S2 Climate-related Disclosures (IFRS Foundation, June 2023), the named successor framework.
- Chamber of Commerce v. California Air Resources Board (9th Cir.): enforcement stay granted November 18, 2025; oral argument heard January 9, 2026; decision pending as of June 16, 2026.
- Cohort dataset and coding rubric: Continuuiti first-wave SB 261 coding sheet, n=57, pile-freeze date May 28, 2026. Distributions plotted from the dataset; method note above.
Frequently asked questions
How many first-wave SB 261 reports did this analyze?
57 substantive reports from the California Air Resources Board (CARB) voluntary docket, as of May 28, 2026. This is a self-selected sample, not the full covered population.
What share of reports quantify physical risk at asset level?
Two of 57, about 4 percent. Roughly 96 percent disclose no asset-level number tied to a hazard.
Which framework do first-wave filers cite?
About 77 percent cite the Task Force on Climate-related Financial Disclosures (TCFD) framework only, 21 percent a TCFD/IFRS S2 hybrid, and one report IFRS S2 on its own.
Is SB 261 currently enforced?
No. A federal appeals court stayed enforcement in November 2025. As of June 2026 the pause holds and CARB has announced no alternate reporting date.
