16 September 2026 · Built from Verisk filings FY2016–Q2 2026, transcripts through 15 Sept 2026, and independent industry sources · Not a valuation and not a recommendation.
The unit is one insurer relationship. It pays about $0.30 of every $100 of premium, mostly as a prepaid subscription that renews each year at a higher price, and Verisk turns that into a 56% adjusted EBITDA margin because the costliest input, the data, is contributed by the customers themselves.
| Item | Summary |
|---|---|
| What the business is | The data utility of US property & casualty (P&C) insurance: statistical agent, standard policy forms and advisory loss costs (ISO), catastrophe models, property-claims estimating (Xactimate) and the industry claims database (ClaimSearch). |
| Industry | Data, analytics and workflow software for insurers; one reportable segment (Insurance), split into Underwriting ($2,180M) and Claims ($893M) in FY2025 (FY2025 10-K). |
| How it makes money | Insurers pay prepaid annual or multi-year subscriptions (83% of revenue) for products built largely on data the insurers themselves contribute; transactional fees (17%) sit on top (FY2025 10-K). |
| Unit of economics | One insurer relationship. Verisk collects about $0.30 for every $100 of premium its customers write (Barclays conference, Sept 2026) and converts it at a 56.2% adjusted EBITDA margin (FY2025, Q4 2025 call). |
| What protects it | A contributory database insurers must feed and cannot replicate alone; regulatory embedding of ISO forms and loss costs in all 50 states; Xactimate as the shared pricing language of adjusters and contractors. |
| What drives earnings | Annual price increases on renewals (just over 5 points of growth a year in 2023–25); cross-sell of new modules into the installed base; operating leverage on a mostly personnel cost base; share count reduction. |
| What to watch | Organic growth returning to the 6–8% target after 4.7% and 5.8% in Q1–Q2 2026; whether pricing holds as P&C premium growth slows; the AccuLynx appeal that could force a $2.35B acquisition. |
| Cycle exposure | Low for the subscription core; medium for the 17% transactional layer (weather, commercial-property quote volumes). |
An insurer's core problem is pricing a promise before it knows the cost. No single carrier sees enough losses to price reliably, and regulators require rates to be justified with data. Verisk's predecessor, the Insurance Services Office (ISO), was built as the industry's rating bureau to solve that: pool everyone's premium and loss statistics, and turn them into shared prices and policy language. Verisk still does exactly this, and has layered catastrophe models, claims tools and fraud databases on top of the same data.
The unit of economics is one insurer relationship. Verisk does not sell units of volume; it licenses a bundle of subscriptions to a carrier, renewed each year at a higher price, with usage fees on top. Every one of the top 100 US P&C insurers is a customer (FY2025 10-K). The whole bill is small relative to what it protects: about $0.30 per $100 of premium, a share that has risen roughly 20% since 2015 (Investor Day 2026). That small share is why price increases rarely become a procurement fight.
"So for every $100 a carrier gets in premium, they spend $0.30 on the Verisk data." CFO Elizabeth Mann, Barclays conference, Sept 2026
| Business | Size and role | Growth evidence |
|---|---|---|
| Forms, rules & loss costs (ISO) | Nearly $1B; policy language across 32 lines, advisory loss costs (Investor Day 2026) | Reimagine platform added ~250 bp to growth (Investor Day 2026) |
| Catastrophe & Risk Solutions | ~12% share of a $3B market; >45% of revenue outside the US; >98% client retention (Investor Day 2026) | 9.4% CAGR since 2021; target high single digits |
| Claims (Xactimate, ClaimSearch, casualty) | $893M in FY2025; ClaimSearch holds 1.9B claim records; Xactimate prices ~22,000 repair line items across 468 areas (FY2025 10-K) | +4.1% in FY2025; +6.1% OCC in Q2 2026 |
| Specialty (Whitespace) and Life | Each passed $100M (Investor Day 2026) | Life more than tripled in six years |
Subscription revenue is prepaid and grows through contract price steps; transactional revenue comes from per-report underwriting data pulls, claim-count overages, services and weather-driven claims estimates (Barclays conference, Sept 2026). In Q2 2026 the two layers moved in opposite directions: subscriptions +8% organic constant currency (OCC), transactional −4.2% (Q2 2026 call).
The portfolio has been narrowed to this insurance core, and the exits explain the margin. Verisk sold healthcare ($715M, 2016), 3E ($575M) and Financial Services ($498M) in 2022, Wood Mackenzie energy research ($3.07B, 2023) and Verisk Marketing Solutions ($80M, end-2025) (FY2016, FY2022, FY2025 10-Ks). Wood Mackenzie was bought for $2.89B in 2015 and sold at a $131M loss, and the auto-marketing unit was a business management called undifferentiated (Q4 2025 call). What remains is the part where Verisk owns the contributory data, which management says underpins over 90% of revenue (Investor Day 2026).
The industry exists because state insurance regulation is built on shared data. Under the McCarran-Ferguson Act states regulate insurance, and advisory organisations such as ISO, AAIS and NCCI are authorised to make filings on insurers' behalf (NAIC Product Filing Review Handbook 2024). Since the late 1980s these bodies file advisory loss costs rather than final rates, and each carrier adds its own expense and profit loading. The buyer base is large and fairly concentrated: US direct premiums written were $1.105 trillion in 2025, and the ten largest groups hold 47.6% of the market (NAIC 2025 industry analysis; NAIC Market Share Report).
Profits concentrate in firms that own contributory data or a two-sided workflow, not in general insurance software. RELX's LexisNexis Risk segment, which runs the rival C.L.U.E. claims-history database, earned a 37.4% adjusted operating margin on £3,485M revenue in 2025. CCC, the auto-claims network used by 27 of the top 30 US auto carriers, earned a 41% adjusted EBITDA margin. Guidewire, which sells core policy and claims systems, ran a 23% non-GAAP operating margin in its FY2026 (RELX FY2025 results; CCC FY2025 10-K; Guidewire FY2026 release). Verisk's 56.2% adjusted EBITDA margin sits at the top of that range; the measures differ, so the comparison is directional.
| Market | Verisk product | Named alternatives | Structure |
|---|---|---|---|
| P&C forms and loss costs | ISO | AAIS; NCCI and 11 independent bureaus in workers' comp | Near-standard, regulated |
| Catastrophe models | Catastrophe & Risk Solutions | Moody's RMS (insurance ARR $649M), KCC, Cotality, Aon, ARA | Oligopoly; 7 vendors accepted in Florida |
| Property claims estimating | Xactimate | Cotality Claims Estimate (ex-Symbility) | De facto standard (no share data) |
| Claims history and fraud | ClaimSearch | LexisNexis C.L.U.E. | Contributory duopoly |
Sources: NAIC PFR Handbook; Indiana Compensation Rating Bureau; Moody's 4Q25 release; Insurance Journal (Florida commission, June 2025); Cotality and RELX disclosures.
The contributory data barrier binds hardest. A new entrant would need carriers to hand over decades of loss history, and the carriers already send it to ISO because regulators require statistical reporting. RELX building a parallel claims database shows the model can be copied once, so the result is a duopoly in claims history rather than a monopoly; in loss costs and statistical agency there is no national equivalent of ISO outside workers' comp.
"I have 70% of the market that I don't see, and I need to know what's happening there to be able to price the best risk." A Verisk customer, as quoted by the CFO, Barclays conference, Sept 2026
Regulatory embedding binds in a quieter way. A carrier that writes its own policy form carries legal liability for every clause and faces a full regulatory review; starting from an ISO form, only the changes are reviewed (Barclays conference, Sept 2026). That makes the ISO form a default that is expensive to leave. It is not permanent: Texas barred insurers from using advisory loss costs in major personal lines until 2003, so the position rests on state law that can change.
Xactimate's position is workflow embedding on both sides of a claim. Adjusters and roofing or restoration contractors price the same job in the same line items, so a carrier switching estimating tools would break the common language with its contractor network. The clearest independent evidence is the FTC itself: in reviewing the AccuLynx deal it treated access to Xactimate pricing integration as a competitive bottleneck for contractor software rivals such as ServiceTitan (Delaware Court of Chancery opinion, August 2026; Axinn commentary). A 2025 court also found that insurance policies do not require any particular Xactimate setting, so the standard is practice, not contract (Belotti v. State Farm, M.D. Pa.).
| Verisk claim | Verdict | Basis |
|---|---|---|
| Contributory data that cannot be replicated | Partly supported | Regulatory status confirmed by NAIC; RELX shows a second claims network is possible |
| Every top-100 P&C insurer is a customer | Unverified | Only Verisk discloses it; structure makes it plausible (inferred) |
| Xactimate is the claims-estimating standard | Supported, indirectly | FTC and Delaware court treated its pricing data as a bottleneck; no share figure exists |
| Pricing power and low churn | Partly supported | Risk Assessment grew 3.9% in 2009 (Verisk FY2009 release); no independent churn data |
| Low cyclicality | Partly supported | True for subscriptions; transactional revenue fell 4.2% OCC in Q2 2026 |
What could weaken the position is less a competitor than a regulator or a customer. Antitrust authorities have twice checked Verisk's moves into adjacent property workflows: the FTC sued to block the $650M EagleView aerial-imagery deal in 2014, calling EagleView a virtual monopoly, and forced a Second Request on AccuLynx in 2025 (FTC press release, 2014; 8-K, 29 December 2025). And the largest carriers, which account for a large share of both premium and Verisk revenue, are the ones with the resources to build their own models with AI.
The winners in this industry combine three things: data customers must contribute, a regulatory or certification role, and software both sides of a transaction must use. Verisk is the clearest example of that type in US P&C, holding all three in underwriting and two of three in claims. Its constraint is the one that comes with the type: expansion into adjacent markets draws antitrust scrutiny.
2016–2021: growth excluding acquisitions, consolidated company including energy and financial services (10-Ks). 2022–2025: Insurance organic constant currency growth (Q4 calls). H1 2026: growth excluding acquisitions and disposals (Q2 2026 10-Q); OCC was 4.7% in Q1 and 5.8% in Q2. The 2023–25 target range was 6–8%.
Verisk grows mostly by charging existing customers more, not by finding new ones. Over 2023–25, pricing contributed "just north of 5 percentage points" a year against organic constant currency growth of 8.7%, 7.1% and 6.6% (Investor Day 2026; Q4 calls). The remaining two to three points came from selling new modules and data sets into the installed base and from transactional volume (inferred). Management has raised the pricing target to 350–450 bp for the next three years (Investor Day 2026).
Acquisitions are not what flatters growth here; disposals depress it. Reported revenue grew 7.4% in 2023, 7.5% in 2024 and 6.6% in 2025, close to organic rates, and in H1 2026 reported growth of 4.1% sat below 5.7% excluding M&A because the $68M marketing unit was sold (FY2025 10-K; Q2 2026 10-Q; Q4 2025 call). The pre-2022 record is not comparable because it included energy and financial-services businesses since sold.
| Period | Reported growth | Organic growth | What explains the gap |
|---|---|---|---|
| FY2023 | 7.4% | 8.7% OCC | 3E and Financial Services revenue in the 2022 base |
| FY2024 | 7.5% | 7.1% OCC | Small acquisitions; FX |
| FY2025 | 6.6% | 6.6% OCC | SuranceBay and Simplitium offset AER sale |
| H1 2026 | 4.1% | 5.7% ex-M&A (Q1 4.7%, Q2 5.8% OCC) | Sale of Verisk Marketing Solutions ($68M in 2025) |
Reported growth calculated from 10-K revenue (FY2022 $2,497.0M; FY2023 $2,681.4M; FY2024 $2,881.7M; FY2025 $3,072.7M). OCC from Q4 earnings calls; H1 2026 from Q2 2026 10-Q and calls.
Just over 5 points a year in 2023–25. Value-based increases are justified by content upgrades such as Core Lines Reimagine, and about 20–25% of revenue is on contracts partly linked to each carrier's premium growth with a two-year lag (Investor Day 2026; Barclays conference, Sept 2026). Management expects 2026 pricing to be "modestly coming down" from a historically strong level (Q4 2025 call).
XactAI, sold as a separate SKU, reached about 7,000 licensees, roughly ten times March 2026; XactXpert is used by 7 of the top 10 homeowners insurers (Q2 2026 call; Q4 2025 call). Management cannot yet quantify AI's contribution to growth (Q2 2026 call).
Catastrophe & Risk has compounded at 9.4% since 2021, is more than 45% international and is the benchmark model in over 70% of catastrophe bonds (Investor Day 2026). Q2 insurance-linked securities fees make it seasonal.
About 17% of revenue. A heavy hurricane quarter added slightly less than 1% to total revenue in Q4 2024, while low weather and soft commercial-property pricing cut transactional revenue 4.2% OCC in Q2 2026 (Q4 2025 call; Q2 2026 call).
Strong multi-year renewals in 2025 created tough H1 2026 comparisons, and a US federal contract paused in H1 2026 resumed in Q3 for one year (Q2 2026 call).
Margins are high because the expensive asset, the data, is supplied by customers at little cost, and the product is delivered digitally. Personnel is about 55% of operating expenses, and cost of revenue was 30% of revenue in FY2025 (FY2025 10-K). Cost grows with headcount, not with each extra renewal, so each point of price flows mostly to profit. Adjusted EBITDA margin rose from a 50–51% base in 2021 to 56.2% in FY2025, including 40 bp of FX benefit (Q4 2022 call; Q4 2025 call).
Cash conversion is strong and working capital is negative. Customers prepay, so deferred revenue funds the business, and capital spending is mainly capitalised software at about 8% of revenue ($244M in FY2025) (FY2025 10-K). Free cash flow of $1.19B in FY2025 was about 69% of adjusted EBITDA (inferred), but management called it "elevated by some one-time elements on interest and taxes" (Q4 2025 call).
2021 is the 50–51% normalized base management set for the margin programme (midpoint shown); 2022 is pro forma for disposals. Sources: Q4 2022–Q4 2025 calls; Q2 2026 call. FY2025 includes about 40 bp of FX benefit.
| $M unless stated | FY2018 | FY2023 | FY2025 | H1 2026 |
|---|---|---|---|---|
| Revenue | 2,395 | 2,681 | 3,073 | 1,589 |
| Organic growth | 6.5%* | 8.7% | 6.6% | 5.7%* |
| EBITDA margin, GAAP / adjusted | 47.8% / NF | 53.1% / 53.5% | 54.3% / 56.2% | 54.9% / Q2 57.5% |
| Free cash flow (OCF − capex) | 703 | 831 | 1,192 | 624 |
| Total debt | 2,723 | 2,867 | 4,737** | 4,474 |
| Diluted shares (M) | 168.3 | 147.3 | 140.1 | 133.0 |
Sources: 10-Ks FY2018–FY2025; Q2 2026 10-Q; adjusted margins and OCC from Q4 calls. FY2018 includes the energy (Wood Mackenzie) and Financial Services businesses sold in 2022–23, so it is not comparable with later years. *Growth excluding acquisitions, not constant currency. **Includes $1.5B of AccuLynx notes redeemed in January 2026. NF = not found in the filings.
Over FY2016–FY2025 Verisk generated $7.8B of free cash flow and spent it, ranked: $8.2B on buybacks, $3.1B on acquisitions, $1.4B on dividends (started 2019), with $2.3B of capex already deducted (cash flow statements, FY2016–FY2025 10-Ks). Returns exceeded free cash flow because divestitures brought in $5.0B, including the $3.07B Wood Mackenzie sale that funded a $2.5B accelerated repurchase in March 2023 (8-K, 7 March 2023). Diluted shares fell 18%, from 171.2M to 140.1M.
The behaviour is consistent: management treats its own shares as the default acquisition and borrows to buy them. After walking away from AccuLynx it launched a $1.5B accelerated repurchase in February 2026, funded with a term loan, revolver and cash, and added $200M more in Q2; leverage rose to 2.5x, the middle of its 2–3x range (8-K, 23 February 2026; Barclays conference, Sept 2026). The new Investor Day target is to return at least 75% of free cash flow to shareholders.
Verisk agreed in July 2025 to buy AccuLynx, roofing-contractor software, for $2.35B in cash. After the FTC issued a Second Request it terminated the deal on 26 December 2025 and redeemed the $1.5B of notes raised to fund it (8-Ks, 30 July and 29 December 2025). In August 2026 the Delaware Court of Chancery ordered specific performance, finding Verisk's own conduct in ending integration talks with ServiceTitan was the primary cause of the regulatory delay, and required Verisk to comply with the Second Request (Axios, 10 August 2026; Axinn commentary). Verisk appealed on 18 August; hearings start in late October (8-K, 18 August 2026; Barclays conference, Sept 2026).
This post-dates every reported figure above. If the deal is ultimately forced through, the CFO says it would be paid in cash and leverage returned to the 2–3x range "within a relatively short period" (Barclays conference, Sept 2026). Adding $2.35B to June 2026 debt of $4.47B would take gross leverage to roughly 3.9x trailing adjusted EBITDA before any AccuLynx profit (inferred), above the 3.75x base covenant but within the 4.25–4.50x acquisition step-up (FY2025 10-K).
End-market exposure is narrow but defensive. About 70% of revenue comes from US P&C primary insurers, the portfolio is roughly 60% commercial lines and 40% personal, auto is about 10%, and 82% of revenue is earned in the US (FY2025 10-K; Investor Day 2026). The top 50 customers are about 42% of revenue, and none is above 3%.
The evidence from past downturns supports low cyclicality in the core. In 2009, Verisk's Risk Assessment business, the ISO forms, loss costs and statistical services, still grew 3.9%, and LexisNexis Risk's insurance revenue grew 10% (Verisk FY2009 release; RELX 2009 annual report). Since its 2009 IPO, Verisk's organic growth has averaged 6.8% in soft insurance markets and 7.3% in hard ones (Q2 2026 call). Insurers do not stop filing rates or settling claims in a recession.
The business is now at a growth low point and a margin high point. Organic growth of 4.7% and 5.8% in Q1–Q2 2026 is the weakest since 2020 (3.4% excluding acquisitions, on the old consolidated basis) and well below the 8.7% of 2023, while adjusted EBITDA margin of 56.3% on a trailing basis is a record (Q1 and Q2 2026 calls; FY2020 10-K). Price is past its peak: after three historically strong years, management expects increases to moderate. The pressure is in the insurance cycle: 2025 was the best underwriting year in two decades (combined ratio 92.9%, NAIC), carriers are now competing on price, global commercial rates fell 6% in Q2 2026 with US property down 13% (Marsh), and H1 2026 insured catastrophe losses of $42B were far below the $66B trend (Swiss Re).
The downside case is a prolonged soft market, and the mechanism is specific. Transactional revenue falls first, as it did in Q2 2026, because commercial-property carriers pull fewer data reports per quote and scrutinise analytics spend (Q2 2026 call). The slower second-round effect is on price: about 20–25% of revenue is on contracts that take carriers' premium growth into account with a two-year lag, so the slowdown in premium growth to around 4% in 2025 (Triple-I/Milliman) would feed into 2027–28 renewals (inferred). The subscription base would keep growing, but at the low end of the target range.
The binding constraints are not capacity but permission. Verisk needs contributors to keep sending data, regulators to keep accepting advisory loss costs, and antitrust authorities to allow adjacent acquisitions. Privacy law is the live example: after GM stopped sharing driving data, Verisk discontinued its driving-behaviour reports in June 2024, and a consolidated class action over that data is proceeding in federal court in Georgia (The Record; MDL 3115, N.D. Ga., April 2026 ruling).
| Indicator | Why it matters | Where published |
|---|---|---|
| Subscription vs transactional OCC growth | Separates the durable core from the cyclical layer | Verisk quarterly calls |
| Pricing contribution vs 350–450 bp target | Main growth engine | Investor Day; Q4 calls |
| Commercial and property rate indices | Leads transactional volumes and the premium-linked contracts | Marsh Global Insurance Market Index; CIAB market survey (quarterly) |
| US premium growth and combined ratio | Carrier budgets and premium-linked pricing | NAIC mid-year and annual analyses; Triple-I/Milliman |
| Insured catastrophe losses | Claims estimating volumes | Swiss Re sigma; Aon (semi-annual) |
| AccuLynx appeal and FTC review | $2.35B of capital and leverage | Delaware Supreme Court; Verisk 8-Ks |
| Top-50 customer share of revenue | Early sign of large-carrier insourcing | Verisk 10-K |