+3.76 percentage points versus 2025 YTD
2026 YTD to AugustEvidence and sourceLIVE EVIDENCE TRACKER · VERSION 1.1
Electric
Mobility
EV sales are moving faster than the insured parc.
Severity and model uncertainty are rising before frequency evidence is credible.
Evidence updated18 Sept 2026Lens assessment recorded18 Sept 2026
THE TRANSITION AT A GLANCE
Four signals that matter now.
New registrations show the direction of travel. The licensed parc shows the exposure insurers actually carry. These measures are deliberately kept distinct.
1.864 million cars
March 2026Evidence and sourceNew BEV share less licensed-car ZEV share
Latest comparable snapshotsEvidence and source7,443 new road-using ZEV LGVs
2026 Q1Evidence and source01 · ADOPTION AND INSURED EXPOSURE
Sales lead. The parc follows slowly.
The flow-to-stock gap is an early-warning indicator of portfolio change. It is not, by itself, a loss-cost answer.
New-car flow versus licensed-car stock
Latest available UK snapshots. Different periods are shown explicitly because the underlying publications have different cadences.
PRICING IMPLICATIONDo not price today’s insured parc from today’s sales mix.
New business can change faster than the back book, but neither measure replaces exposure-matched claims experience.
EVIDENCE BOUNDARYPublic data cannot yet support a credible UK comparison of EV and ICE claim frequency per exposure-adjusted mile.
Vehicle counts do not reveal annual mileage, road environment, journey purpose or claims propensity. The detailed research backlog remains in the governed admin workflow.
02 · LEADING INDICATORS
Watch where change arrives first.
Commercial fleets, London, new manufacturers and charging supply can reveal emerging exposure and repair pressures before they are visible across the national insured parc.
Electrification is uneven by vehicle class.
Share of new road-using registrations that are zero-emission. DfT VEH1153a.
New marques are entering faster than claims credibility can mature.
DEFINED CHINESE-BRAND SHARE~16.1%of new-car registrations
MG4.39%
BYD3.48%
Jaecoo3.14%
Omoda1.76%
Chery1.70%
Geely0.67%
Leapmotor0.67%
Other defined0.20%
Brand origin, ownership and assembly location are not treated as interchangeable. The published marque definition is repeatable.
LONDON LEADING INDICATORLatest available fleet comparison: 2024
Useful precisely because it is not representative.
London’s regulation, charging environment and high-intensity fleets make it unsuitable as a UK proxy—but useful for seeing emerging mechanisms early.
Battery-electric share accelerated fifteen-fold.
1.3%2020
3.1%2021
10.2%2022
14.1%2023
20%2024
PHV battery-electric share16.7×London multiple
London moved from 1.3% in 2020 to 20.0% in 2024.
Taxi electric-capable share28.9×London multiple
Battery-electric plus range-extended electric; this is not a pure zero-emission measure.
Definition: battery-electric plus range-extended electric; not a pure BEV measure.SUPPORTING CONTEXT · 1 JULY 2026121,171 public charging devices
Evidence Including 28,887 rapid or above and 174.9 devices per 100,000 people. Infrastructure can enable adoption and utilisation, but is not itself an insurance-risk measure.
03 · THE PRICING ACTUARY’S LENS
What this means for UK motor insurance.
The summary exposes the conclusion first. Open any test for the evidence, limitation and trigger that would justify changing the view.
01Claim frequencyHow could this affect how often claims occur?UncertainLow confidence+
Current assessment
Propulsion alone is unlikely to explain frequency. Utilisation, road mix, vehicle mass, driver assistance and commercial use need to be separated before drawing a pricing conclusion.
Current evidence
Vehicle adoption is measurable; comparable claims or collision exposure by propulsion and use remains incomplete.
What would change this view?
Initial assessment recorded · 18 Sept 2026Exposure-matched claims or collision evidence by propulsion, vehicle age, use and road environment.
02Claim severityHow could this affect average claim cost and its distribution?IncreasingMedium confidence+
Current assessment
Battery location, ADAS calibration, parts availability, repair capability and downtime can increase repair cost and total-loss propensity even where crash performance improves.
Current evidence
The mechanisms are credible, but a stable UK insurer-level severity distribution by propulsion is not yet public.
What would change this view?
Initial assessment recorded · 18 Sept 2026Credible repair-cost, total-loss and repair-duration evidence matched by vehicle segment and age.
03Exposure basisIs the exposure unit currently used still valid?IncreasingHigh confidence+
Current assessment
Vehicle-years and annual mileage may be insufficient where commercial EV adoption concentrates high-utilisation mileage in dense urban environments.
Current evidence
Commercial electrification is uneven and vehicle counts do not reveal mileage, road type, time of day or journey purpose.
What would change this view?
Initial assessment recorded · 18 Sept 2026Reliable mileage and road-environment measures available at rating or portfolio level.
04Tail and accumulation riskCould this limit or amplify large and correlated losses?MixedLow confidence+
Current assessment
Depot charging, common battery or software defects, flood exposure and concentrated fleets create plausible accumulation channels, but loss evidence is still sparse.
Current evidence
The dependency structure is visible; insured loss distributions and credible event frequencies are not.
What would change this view?
Initial assessment recorded · 18 Sept 2026A credible UK loss event, recall or portfolio study quantifying common-cause exposure.
05Price response and selectionHow price-sensitive is the market and how might selection change?MixedMedium confidence+
Current assessment
Insurance cost is part of EV total cost of ownership. Repair-cost uncertainty and sparse model-level experience can change take-up and attract adverse selection if price does not differentiate quickly enough.
Current evidence
New marques and models are entering faster than credible model-level claims experience can mature.
What would change this view?
Initial assessment recorded · 18 Sept 2026Observed quote conversion, retention and elasticity by propulsion and model maturity.
06Capital vulnerabilityCould under-informed capital be burned before experience emerges?IncreasingHigh confidence+
Current assessment
Yes. Fast growth, new manufacturers, changing repair economics and thin model-level credibility can make apparently attractive EV segments easy to misprice.
Current evidence
The flow-to-stock gap and new-entrant share show that portfolio composition can change ahead of mature claims evidence.
What would change this view?
Initial assessment recorded · 18 Sept 2026Stable loss-cost emergence by model, repair network and use class.
07Customer expectationDo customers want or expect change?MixedMedium confidence+
Current assessment
Customers expect the ownership economics of EVs to improve as batteries, charging and repair ecosystems mature—and will increasingly question unexplained insurance differences.
Current evidence
Adoption and charging supply are growing, while affordability and repair concerns remain visible barriers.
What would change this view?
Initial assessment recorded · 18 Sept 2026Reliable purchasing, switching and complaint evidence tied specifically to insurance cost and cover.
08Duty to actCould doing nothing permit foreseeable harm or poor value?IncreasingMedium confidence+
Current assessment
Where evidence shows material differences in repairability, usage or expected cost, insurers should test whether rating, disclosure, claims pathways and value assessments remain appropriate.
Current evidence
The evidence is not yet sufficient for broad propulsion-only conclusions, but it is sufficient to justify active monitoring and controlled testing.
What would change this view?
Initial assessment recorded · 18 Sept 2026Material portfolio evidence that is known, reproducible and not reflected in product or pricing decisions.
These assessments apply the eight-question framework. Read the framework
04 · EVIDENCE AND GOVERNANCE
Public data, visible provenance.
Only directly sourced or reproducibly derived measures appear above. Evidence can refresh through the governed process; the Lens never changes automatically.
Every public metric has passed the publication threshold.Value present · approved source · valid denominator · current period · clear interpretation
METHODOLOGYHow evidence is checked and published+
Detect a newer official or public release.
Confirm it supersedes the current snapshot.
Run documented derived measures.
Publish facts, dates and definitions.
Refer potential Lens changes to Johan.
SOURCE REGISTEROpen the full evidence provenance+
SourceWhat it supportsCadenceQuality
SourceUK new-car registrations and market dataSMMTSupportsFuel type, buyer type, marque and model registrationsCadenceMonthlyQualityIndustry primarySourceVehicle licensing statisticsDfT / DVLASupportsLicensed parc, new ZEV registrations, make/model stock and geographyCadenceQuarterlyQualityGovernmentSourceElectric vehicle charging infrastructure statisticsDepartment for TransportSupportsPublic devices, rapid devices and geographic densityCadenceQuarterlyQualityGovernmentSourceMOT testing data for Great BritainDVSASupportsAggregate test counts and failure categories; not a vehicle-level mileage denominatorCadenceQuarterlyQualityGovernment open dataSourceRoad safety open data (STATS19)Department for TransportSupportsCollision circumstances, severity, road type and propulsion fieldsCadenceProvisional and annualQualityGovernment open dataSourceNational Travel SurveyDepartment for TransportSupportsMileage, trip purpose, travel behaviour and fuel contextCadenceAnnualQualityGovernmentSourceTaxi and private-hire vehicle data tablesDepartment for TransportSupportsFuel mix, fleet age and London-versus-rest-of-England comparisonsCadenceAnnual / methodology under reviewQualityOfficial statistics in developmentSourceReproducible derived metricsThe LensSupportsDocumented transformations of public source tables; never treated as directly published factsCadenceWhen inputs changeQualityDerivedCHANGE HISTORYWhat changed—and what did not+
Evidence update
London leading indicator and calculation register added
- Evidence
- Published London taxi and PHV comparisons, six reproducible next-stage measures and a clearer boundary between observed signals and unavailable exposure denominators.
- Lens
- No material change
Method correction
Public MOT data limitation clarified
- Evidence
- Removed the implication that the current public MOT tables provide vehicle-level odometer histories by propulsion. Mileage-based comparisons remain gated pending suitable linked data.
- Lens
- No material change
Launch
Electric Mobility tracker launched
- Evidence
- Published the UK adoption, commercial, manufacturer, charging and exposure baseline. Lens interpretation recorded separately.
- Lens
- Initial assessment recorded
Evidence update
August registration evidence incorporated
- Evidence
- Updated new-car BEV share and the defined Chinese-brand registration measure; no separate Lens revision.
- Lens
- No material change