The FCA’s 2026 regulatory priorities for insurance, published on 24 February 2026, contain one sentence that pet insurance pricing teams should have printed and taped to their monitors: “Pet and private medical insurance: both lines continue to be monitored by the FCA for potential future action, given price rises and consumer understanding issues.”
That is not the FCA at its most alarming. It is not an investigation, a thematic review, or a named firm review. But the FCA has a pattern: monitoring becomes a thematic review becomes enforcement. We have watched it happen in motor, home, and now travel insurance. The gap between “monitoring” and “formal action” is typically 12 to 18 months, which is exactly the time a pricing team needs to get its house in order.
Pet insurance is not a product where “we have always done it this way” is a defensible answer. The market has structural features that make poor consumer outcomes almost inevitable if pricing models are not actively maintained — and the FCA knows this.
The complaint data is already bad
The Financial Ombudsman Service published Q1 2025 data showing 600 complaints about pet-related policies in the quarter, up 26% from Q1 2024. Over five years, pet insurance complaints have grown by 146%. It is now the fourth most complained-about business line in UK general insurance.
The upheld rate is the number that matters. In Q1 2025, 52% of pet insurance complaints were upheld in favour of the customer. That is 19 percentage points higher than Q1 2024, and it is the highest upheld rate of any UKGI business line. For comparison, travel insurance — which is actively under FCA enforcement following the Which? super-complaint — ran at 37% in the same period.
A 52% upheld rate is not a communications problem. It is not a claims handling training problem. It is a systematic mismatch between what customers think they bought and what they actually bought — and in a significant portion of cases, the Ombudsman is finding in favour of the customer. That finding has a specific implication for pricing teams: if claims are being declined at a rate the Ombudsman considers unjustified, then your pricing model is calibrated to a claims experience that does not reflect what you are actually paying out. Your loss ratios are wrong in a direction that makes your product look better value than it is.
Of the Q1 2025 complaints, 70% were claims-related. Sales and advice complaints grew 73% year on year and now account for nearly 20% of all pet insurance cases — up from 14% in Q1 2024. The sales complaints are the more interesting signal: they suggest customers are being sold products on the basis of representations about coverage that the product does not deliver.
Vet inflation is structural, not cyclical
The ONS reported 9.1% inflation for vets and other pet services in 2024. Treatment costs have risen approximately 60% since 2015. The average vet consultation now costs £58.29, up 8% on the previous year.
The ABI reported £1.23 billion paid out in pet insurance claims in 2024 — the third consecutive year above £1 billion, up 4% on 2023. Total claims notified reached 1.8 million, an all-time high. Average claim cost is £685, up 3% year on year. Over ten years, the number of claims paid has more than doubled.
These numbers tell a clear story: vet inflation is running structurally above general CPI, and pet insurance claims experience is deteriorating on both frequency and severity simultaneously. Claims frequency is rising because more pets are insured, because veterinary diagnostics have improved (meaning more conditions are identified and treated), and because the population of insured pets is ageing. Severity is rising because treatment options are more expensive and because veterinary specialist referrals are more common.
The standard actuarial response to this — update your trend factors, refit the severity model, refresh the pricing — is necessary but not sufficient. The FCA’s concern is not that loss ratios are moving. Its concern is that premium increases are being driven by factors that are genuinely opaque to customers, that customers cannot meaningfully compare products between providers, and that certain segments — typically older pets and breeds with known health predispositions — are being priced in ways that produce systematic poor outcomes under Consumer Duty.
We think the FCA’s analytical target here is not the aggregate loss ratio. It is the distribution of value across customer segments. A product that is profitable overall but systematically extracts value from older-pet policyholders who are effectively locked in by pre-existing condition exclusions is a Consumer Duty problem even if the overall numbers look acceptable.
The lock-in problem
Pre-existing condition exclusions create a structural lock-in effect in pet insurance that has no real equivalent in motor or home. In motor, a customer who is unhappy with their renewal quote can switch insurer and take their no-claims bonus with them. In pet insurance, switching typically means any condition that has manifested since the original policy incepted is excluded by the new provider. This is not a hidden term — it is disclosed — but the practical effect is that customers with older or unwell pets are trapped.
The FCA is alert to this. The “consumer understanding” element of its monitoring concern is directly about this dynamic: customers do not understand, at inception, the full implications of lifetime policy coverage restrictions, and by the time they discover the consequences (at a renewal following a claim), they have lost the ability to exercise any meaningful choice.
For pricing teams, this creates a specific obligation under Consumer Duty that goes beyond actuarial adequacy. PRIN 2A.4.6 requires firms to monitor whether their products provide fair value and whether different groups of customers are experiencing poor outcomes. A customer who is effectively locked in to a product that is repricing aggressively at renewal, with no viable exit, is experiencing a poor outcome by definition — even if every individual renewal premium is technically “actuarially justified.”
The fair value assessment for a pet insurance product needs to address this directly. It is not sufficient to demonstrate that the pricing model is actuarially sound at a product level. You need to segment by tenure, pet age, and claim history, and you need to show what is happening to renewal pricing for the locked-in cohort versus the switchable cohort.
What the Value Measures data shows — and what it doesn’t
The FCA’s General Insurance Value Measures data for 2024 covers January to December 2024 and is published annually to drive transparency in the market. For the covered lines — motor, home, GAP, and several add-on products — the data provides claims frequency, acceptance rates, average payouts, and claims complaints rates at firm level.
Pet insurance is reported under the lifetime cover category. The 2024 data shows 4.58 million lifetime pet insurance policies generating £2.07 billion in premiums, with a claims frequency of approximately 42%. The data is useful for benchmarking, but it is less granular than the motor and home data — partly because pet insurance has historically been seen as a lower-priority line, and partly because the market is more fragmented.
The honest assessment is that we have less analytical infrastructure for pet insurance than for motor or home. There is no equivalent of the ABI Motor Premium Tracker for pet insurance. Breed-level claims data is proprietary to individual insurers and not pooled. The actuarial literature on pet insurance is thin compared to personal lines. Any pricing team working on this product needs to be honest with itself that the external benchmarks are sparse and that internal model validation carries more weight as a result.
This is precisely why the FCA’s monitoring concern is significant. The relative lack of transparency in pet insurance makes it harder for consumers to identify bad value, harder for the FCA to conduct like-for-like comparisons, and harder for pricing teams to demonstrate that their models are producing acceptable outcomes. The remedy is not to wait for better external data. It is to build the internal evidence base now.
What pricing teams need to do
We are not going to pretend this is a simple checklist. But there are three areas where we think pet insurance pricing teams are most exposed and where the work is tractable.
Fair value assessment with proper segmentation. The FCA’s thematic review TR24/2 found that most firms were producing “high-level summaries with little substance.” That finding was in the context of motor and home, but it applies equally to pet insurance and the bar will not be lower. A fair value assessment for pet insurance needs to segment by pet age, breed category, tenure, claim history, and distribution channel. It needs to show, with numbers, what the loss ratio looks like for a five-year-old Golden Retriever with a history of claims who renewed this year, compared to a first-year policy on a young cross-breed. If those numbers are significantly different and the premium trajectory does not reflect the difference proportionately, you have a fair value problem.
Our insurance-fairness library provides the statistical framework for this kind of disparity analysis. The ProxyDetection and FairnessAudit modules run segmented outcome monitoring and flag segments where outcomes differ materially from the product-level average.
Actual vs. expected drift monitoring for vet inflation. The structural feature of vet inflation is that it is not uniform. Orthopaedic procedures have inflated faster than dental. Specialist referral costs have outpaced general practice fees. Oncology treatments that were not routinely available five years ago are now standard of care and attracting claim volumes that were not in any historical experience dataset.
This means that a severity model fitted on data more than two years old is almost certainly mispriced on the tails. Your average claim cost may look acceptable in aggregate, but your 90th percentile claim cost has likely moved materially. If your model is not tracking actual versus expected severity by condition category and by treatment type, you do not know where your exposure sits.
The insurance-monitoring library handles A/E tracking with the statistical test infrastructure to distinguish genuine trend shifts from noise — including Wasserstein drift tests that are appropriate when the severity distribution is changing shape rather than just shifting in mean. Run these by condition category. The signal in orthopaedic claims will look different from the signal in soft tissue claims, and blending them into a single A/E ratio will obscure both.
Pre-existing condition modelling and age rating curves. The age rating curve in pet insurance carries a structural feature not present in motor or home: the hazard rate for most conditions is monotone increasing from roughly year three of a pet’s life and accelerates materially in the final years. A pricing model that uses a piecewise linear age adjustment is almost certainly underpricing old-age risk and overpricing young-pet risk.
This is not merely a profitability concern. Under Consumer Duty, systematic underpricing of old-age risk means that either (a) older-pet policyholders are subsidised by younger-pet policyholders, which is a cross-subsidy that needs to be disclosed and justified, or (b) older-pet policies are generating adverse loss ratios that are compensated by restricting claims, which is the pattern that produces 52% FOS upheld rates.
A more honest age rating curve is built by studying when claims first manifest by condition category, not by averaging over all causes. The compound problem in pet insurance is recurrence: a pet with orthopaedic problems at age four will almost certainly have orthopaedic claims at five, six, and seven. A GLM that treats each policy year independently misses this entirely and will systematically underprice the chronic-condition tail. Our insurance-recurrent library implements shared frailty models for within-policyholder claim recurrence — capturing the fact that a pet with a claim history is materially more likely to claim again than a naive frequency model would predict, and quantifying the breed-level frailty distribution that underlies this effect.
Document everything. The governance requirement under Consumer Duty is explicit: fair value assessments must be evidenced, must be reviewed by the board or equivalent governance body, and must be updated when material changes occur to the product or the market. “Material changes” in pet insurance in 2024 and 2025 included a step change in vet inflation, a shift in claims frequency from oncology treatments, and FOS decisions that reset the effective coverage interpretation on several common exclusion clauses.
Our insurance-governance library generates the documentation trail that turns analysis outputs into board-ready governance artefacts — decision logs, assumption registers, and the dated audit trail the FCA will ask for if it moves from monitoring to formal review.
The honest uncertainty
We should be direct about what we do not know. Pet insurance is less studied than motor or home in the actuarial literature. The FCA’s monitoring concern does not yet specify which practices it considers most problematic. The 52% FOS upheld rate is striking but the sample is small — 600 complaints in a quarter against a market of 4.6 million policies. Individual FOS decisions on exclusion clauses do not necessarily translate into systemic model risk.
What is not uncertain is the direction of travel. The FCA flagged pet insurance explicitly in its 2026 priorities. Complaint volumes are up 146% over five years. The upheld rate has hit 52%. The structural factors — vet inflation, lock-in, consumer understanding gaps — are not going away. Any pricing team that treats this as a “wait and see” situation is misreading the timeline.
The window to get ahead of this is probably 12 months. After that, firms will be responding to FCA requests rather than leading with their own evidence.