How are semi-autonomous features on vehicles affecting the UK motor insurance sector?

The purpose of Bank Overground is to share our internal analysis. Each bite-sized post summarises a piece of analysis that supported a policy or operational decision.
Published on 17 October 2019

New vehicles are fitted with increasingly advanced technology. We highlight three trends and consider whether new sources of data will be required to monitor the changing risks arising from autonomous technology on vehicles.

We might still be many years away from a fully autonomous vehicle, but preparations are underway. For example, the UK’s first autonomous vehicle test track launched recently. Against this backdrop, we look at three different data sets (Department for Transport, Society of Motor Manufacturers and Traders, and Solvency II) and examine whether they provide insights to enable monitoring and forecasting of the changing risk landscape.

In 2017, we projected a potential reduction in the size of the motor insurance market over the next two decades as autonomous vehicles take to the road. We also highlighted likely increased losses arising from product liability,
cyber-attacks, litigation and disruption caused by the technological failure of autonomous vehicles.

Data from the Department for Transport indicate a sharp rise in technology-attributed recalls (Chart A), suggesting the potential for these risks is increasing.

Chart A

Vehicle recalls relating to hardware and software failure are increasing

Chart A - How are semi-autonomous features on vehicles affecting the UK motor insurance sector?
  • Source: Department for Transport, Bank interpretation and calculations.

    The data for 2019 is estimated. The Department for Transport publishes a list of all the vehicles that have been recalled because of a safety issue. The list includes over 5,500 recall notices that have affected over 33 million cars, trucks, coaches, vans and motorcycles since 1992. For each vehicle that is recalled the concern, defect and the remedy are provided along with the year of the product recall, the number of affected vehicles and the make and model of the vehicle.  We have allocated each product recall into categories ('software', 'hardware/sensors' or 'other') based on the cause of the recall. In order to do this we manually categorised 20% of the entries and used these to calibrate various machine-learning techniques. Each machine-learning technique was calibrated using c.800 manually categorised recalls. Each technique was then tested against the remaining subset of c.200 manually categorised entries to assess its accuracy. The most accurate machine learning technique was a unigram model in conjunction with a random forest algorithm. The technique involved estimating the categorisation based on thousands of decision trees, each of which classified the data according to a set of rules based on single word counts in the text, and then denoting the ultimate categorisation as the most common classification amongst the trees. The machine-learning model that provided the greatest degree of accuracy in the test sample was then used to assign the categories 'software','hardware/sensors' or 'other' to each of the remaining c.4500 product recalls. We then reviewed each of the c.600 product recalls that were allocated to the categories 'software' or 'hardware/sensors' and checked these had been allocated correctly. Eleven of the c.600 were identified to have been inappropriately allocated and were reallocated manually. We were then able to see trends in the number of recalls and the number of vehicles recalled each year due to issues with their software or with their hardware/sensors. We have allocated all product recalls and not only those specifically in relation to the ADAS features included within Chart A. The data included recall notices up until 12 September 2019 and so we scaled up the 2019 results on a pro-rata basis to reflect a full year. 

    With thanks to the Bank of England’s Advanced Analytics Division for their support in employing the machine-learning techniques.

    The data used from the Department for Transport to create this chart is public sector information licensed under the Open Government Licence v3.0. The Open Government Licence can be found here.

Furthermore, the proportion of new vehicles with high-tech safety systems and semi-autonomous features is increasing (Chart B). For insurers this means, on average, fewer accidents, but repairs that are more expensive when they occur (Chart C) due to the cost of replacing or repairing the vehicle’s technology. 

Chart B

The proportion of new vehicles with semi-autonomous features is increasing

Chart B - How are semi-autonomous features on vehicles affecting the UK motor insurance sector?
  • Source: The Society of Motor Manufacturers and Traders, JATO Dynamics Ltd, the Department for Transport and Bank calculations.

    (a) 2016 = 0.3%, 2017 = 0.5% and 2018 = 0.4%

    The Society of Motor Manufacturers and Traders (SMMT) publishes information on the number of vehicles in the UK registered with semi-autonomous vehicle technology fitted as standard. This is based on JATO Dynamics Ltd analysis and SMMT new car registration data. 

    The Department for Transport publishes the number of cars registered for the first time in the UK by year within its Vehicle Licensing Statistics dataset VEH0150. We have combined blind spot monitoring (included in 2015) with overtaking sensor (included in 2016, 2017 and 2018). 

Chart C

Fewer insurance claims for vehicle damage, but claims costs are increasing

Chart C - How are semi-autonomous features on vehicles affecting the UK motor insurance sector?
  • Sources: Solvency 2 reporting templates S.21.01 and S.21.03, FSA form 32 and company annual accounts and Bank calculations.

    The frequency and average claim size data are based on Solvency 2 reporting templates for a sub-set of firms in the UK reporting line of business 'Other Motor Insurance'. FSA Form 32 and company annual accounts were used to supplement this information to estimate policy count back to 2007. The Solvency 2 line of business 'Other Motor Insurance' includes insurance obligations that cover all damage to or loss of land vehicles. It does not cover third party bodily injury claims or damage to third parties’ vehicles or properties. We assessed the data for 12 UK motor insurers and removed any firms where there appeared to be obvious errors in the data. The data remaining covered over half of the UK motor insurance market when measured by total claims incurred in 2018.

Understanding how these risks are changing the motor insurance market, and how products may need to adapt, is important in assessing the longer-term viability of insurers’ business models. This is a key part of our job as the UK’s prudential regulator.

Insurers need relevant, accurate and timely data to be able to price risks and assign liability. Regulators and insurers need to monitor any trends arising from automation in the insurance market.

We welcome views on what publicly available data sources or data sharing agreements will be required to continue to assess risks and to help understand the changing motor insurance landscape.

This post has been prepared with the help of Chris Wiltshire, Stefan Claus, Nick Silk and Sarah Munday. James Brookes and Tim Munday in our Advanced Analytics Division provided support in employing machine-learning techniques.

This analysis was seen by SRPC committee members in October 2019.

Share your thoughts with us at BankOverground@bankofengland.co.uk

This page was last updated 30 October 2019
Was this page useful?
Add your details...