Telematics: connecting the dots. Swiss Re telematics solutions for better risks and fairer premiums

Telematics: connecting the dots Swiss Re telematics solutions for better risks and fairer premiums

Contents Introduction 5 Telematics: A first step for insurers to get prepared for autonomous vehicles Telematics as a business opportunity in the changing motor insurance space 6 7 Better risk selection and fairer premiums 8 Closer ties with consumers via value-adding services and driving behavior feedback The big dilemma. Is there enough value for consumers to share their data? 10 10 The challenges of bringing telematics to consumers 10 Helping insurers prepare for the future motor insurance market 12 What s next? 14 About Swiss Re 14 Telematics in motor insurance Swiss Re 3

4 Swiss Re Telematics in motor insurance

Motor insurance is currently experiencing three major trends: growth in emerging markets, decreasing accident frequency and new business opportunities. Premiums for motor insurance are showing a moderate upward trend on a global scale. Despite the expected decline in the number of accidents due to improved car technology, the number of risks and the corresponding need for insurance will continue to increase. This development is largely due to growing populations and economies in emerging markets. Motor insurance continues to be the single largest line of non-life business globally, making up 42% of gross premiums written, or USD 700 bn. It is particularly important in emerging markets, where it accounted for 58% of gross written premiums in 2014, or roughly USD 200 bn. By contrast, in advanced economies the premium volume accounts for approximately USD 500 bn and a market share of 38%1. Baidu have announced highly automated vehicles for the mass market for 2020. Various other car manufacturers have built emergency auto-brake functionality into their cars. Cars are becoming increasingly connected, which opens a multitude of opportunities in the area of usage-based insurance (UBI). Products based on how often, where and how people drive not only offer new solutions for more customised insurance policies but also enable insurers to encourage better driving behaviours, leading to lower risks. New technologies, such as automated driver assistance systems (ADAS), are making driving much safer. For example, the new Mercedes Model E features a drive pilot capable of maintaining a safe distance to other cars at speeds of up to 210 km/h, including steering and broadside collision prevention. Tesla s Model S software will enable autopilot on US highways as of 2018, Volvo and Telematics offers direct and real-time access to data about the insured object. This enables insurers to base their risk assessment and pricing on the real risk policyholders are taking rather than on approximate tariff criteria about the driver, the car and how it is used. 1 Source: The future of motor insurance. Swiss Re, 2016 Telematics in motor insurance Swiss Re 5

Telematics: A first step for insurers to get prepared for autonomous vehicles Figure 1 Forecast of motor insurance premiums taking into account the impact of technology 1400.00 1200.00 1000.00 (USD bn/year) Effect of AV technology 800.00 600.00 400.00 2015 2020E 2025E 2030E 2035E Year Total Motor Premium (taking into account AV technology) Total Motor Premium without AV impact 14 largest motor markets: Brazil, Canada, China, Egypt, France, Germany, India, Indonesia, Italy, Japan, Mexico, Russia, UK, US Source: HERE-Swiss Re publication: The future of motor insurance how car connectivity and ADAS are impacting the market UBI demand is set to increase, driven by the innovative concept that the drivers pay for the amount of risk they take. This translates into continuous feedback, which in turn results in better driving behaviour. Telematics also enables the provision of additional services, such as stolen car recovery or efficient claims handling. These factors taken together improve both customer experience and the insurers results. In Italy, 4.4m passenger cars (16%) are equipped with a telematics box. The US has 3.3m, and the UK follows with 0.6m2. One could argue that even in the leading telematics markets such as Italy, the US and the UK, the telematics market share is so small that it is hardly worth the investment for insurers to bring such a product to the market. This may hold true when considering telematics as a short-term exercise, but we should not underestimate the role of telematics in the longer term. We believe telematics will be the basis for further digitalisation in insurance all the way to autonomous driving. To determine whether the driver or the system was driving in case of an accident or to give premium discount for utilizing ADAS, insurers will need to be able to interpret telematics data coming from an automated vehicle. This in turn will determine which insurance policies might be triggered, ie a motor third party liability (MTPL) or a product liability cover. Entering telematics is a clear strategic decision and the first step for insurers preparing for the new world of autonomous vehicles. Long before fully autonomous cars will actually hit the road, we expect the reduction in the number of accidents to accelerate due to ADAS. Fewer accidents are likely to lead to lower expected losses for insurers, which in turn should bring down overall third-party motor liability premiums. As sensors and car computing become more complex and replace 2 Source: Osservatorio Connected Insurance 2016 Bain & Company and ANIA 6 Swiss Re Telematics in motor insurance

Telematics as a business opportunity in the changing motor insurance space The data insurers receive from telematics devices opens a whole new world for risk assessment and insurance pricing. For data collection, four approaches are worth noting: Consumer devices, eg smartphones Third-party devices, eg 12V cigarette lighter plugs, blackbox Third-party devices with CAN-Bus access, eg OBD dongles Direct access to the car s OEM telemetry All of these approaches have their advantages and drawbacks, and the most suitable device in any given case depends on the business model the insurer chooses for its telematics product (eg the intention to link it to services such as ecall). With current hardware-linked solutions, such as a 12V or OBD dongle, making a business case can be challenging, especially in low-premium consumer segments. Going forward, mass production will result in decreasing hardware costs and ultimately market entry costs. Bundling interactive smartphone apps with telematics hardware is straightforward and gives the insurer direct access to the policyholder, which is important for live scoring, driving feedback and value-adding services. an increasing number of the driving decisions, part of the liability will shift from the driver to the product manufacturer. However, other motorrelated insurance risks, such as hail or theft, will still need to be covered by the driver s insurance. According to a recent study by HERE and Swiss Re, ADAS features could cut overall motor accident frequency by 4.3% by 2020. This will help to reduce the number of fatal road traffic accidents worldwide, which are currently at 1.2 million per year. Swiss Re s long-term forecast for the 14 largest motor insurance markets indicates that this segment will be worth an estimated USD 594bn by 2020. This implies that until then, over USD 20bn would be trimmed from annual premiums as a result of automated car technology and increasing road safety. Telematics in motor insurance Swiss Re 7

Better risk selection and fairer premiums Figure 2 Data flow in the Swiss Re IoT platform Contextual trip information - Weather conditions - Traffic conditions - Road conditions - Ambiance conditions GPS Positions Trip information (analytics by Swiss Re) - Situational description - Kinetic description - Event description Situational risk assessment (analytics by Swiss Re) - Big Data machine learning - Contextual patterns Crash incidence model Accelleration data Swiss Re IoT Platform Driver risk assessment (analytics by Swiss Re) - Big Data machine learning - Behavioural patterns Crash severity model Collection stage Preparation stage Evaluation stage Scoring stage Insurers are interested in approximating the average claim amount to be expected for a given policy. Today, actuaries segment portfolios to homogenous risk groups of similar properties and approximate the expected loss cost for this segment based on historical claims information and actuarial methods. This methodology bases the assessments of the policyholders on proxies for riskiness such as their age, type of car, ZIP code or their bonus malus rating. Although motor insurance has reached a remarkably sophisticated level of prediction quality, telematics technologies will help insurers predict claims even more accurately. Telematics offers direct and real-time access to data about the insured object. This enables insurers to base their risk assessment and pricing on the real risk policyholders are taking rather than on approximate tariff criteria about the driver, the car and how it is used. Two assessment types can be based on information obtained through telematics about the position, speed and acceleration (and derived features) of the car: Situational risk assessment addressing the context in which the driver uses the car. This assessment is position-based and requires additional context-related information. Driver risk assessment focusing on actual driving behaviour, ie on how the driver uses the car and behaves on the road. The assessment depends on acceleration data of the car but not (strictly) on its position. Contextual data is key in assessing the inherent risk of the real-time driving situation. Once the trip positions of an insured vehicle are established, the context of the vehicle plays a major role in identifying the actual risk taken. Today, contextual data for traffic densities and weather conditions can be obtained in very high resolutions. 8 Swiss Re Telematics in motor insurance

Figure 3 The classical bonus malus level has a significant drawback: it takes a long time to build up reliable claims experience, and data tends to be outdated when it matters most. Real-time information about changing risks provided by telematics technologies significantly reduces information asymmetry. Contains information about past 10 years Increasing driving experience Contains information about past 20 years Increasing claims experience Contains information about past 30 years amount of uncertainty about risk/claims average premium Decreasing driving abilities 20 30 40 50 60 +75 Contains information about past 40 years Contains information about past 50+ years age of policyholder A contextual risk assessment is crucial, since we all know that racing home through a blizzard is more dangerous than taking the car for a spin on a sunny day. Another example is that a high percentage of fatal driving accidents (in terms of the number of trips) happen at night. Anyone driving at night is thus, on average, exposed to a higher risk than during the day, a fact that a situational risk assessment can take into account. Weather conditions or the time of day are only a few simple examples of situational context information that can enrich the trip data acquired from telematics devices. Generally speaking, any additional information, such as general traffic density or the expected trip length between two consecutive traffic signal stops can be considered contextual features of a trip. To give a broader perspective, we understand contextual data to comprise features that relate to the environment (weather, sight, temperature, etc), traffic (traffic density, speed limit, etc) and topology (road type, urban region, etc). In contrast to contextual scoring, behavioural scoring refers to the actions the driver takes. Given that one third of all driving accidents can be attributed to distractions3, the evaluation and detection of driving manoeuvres and the possibility of relating them to such features as fatigue and measurable driving ability is an important factor for risk-adequate scoring. To obtain this data, trips need to be broken down into manoeuvre segments, such as cornering, lane change and manoeuvring at obstacles or intersections. Machine learning algorithms not only detect but also rate the manoeuvres with respect to their risk qualities. Details from this assessment can also help to infer driving abilities. In addition, specific manoeuvres, such as lane departures, may be directly linked to dangerous behaviours, such as using the phone while driving, and be scored accordingly. Although the behavioural scoring approach may not be able to answer the question of how competent a driver is, it allows to quantify the risk connected to his or her actions, enabling the insurer to price these risks into the tariff. 3 Source: Driver crash risk factors and prevalence evaluation using naturalistic driving data, T.A. Dingus et al., Proceedings of the National Academy of Science of the United States of America 113 (10), 2636-2641, doi: 10.1073/pnas.1513271113 Telematics in motor insurance Swiss Re 9

When combining a contextual with a behavioural approach in scoring, an insurer can build up substantial intelligence about crash prevalences within specified time periods. It is important to note that the measurement phase does not necessarily need to be as long as the whole insurance period. In the short term, targeting customers who are ready to embrace this innovation will result in positive risk selection, as the telematics-based insurance offer is more attractive to safer drivers. This will increase the profitability of the tariff. In the medium term, however, this benefit will shrink, although this can be offset by leveraging economies of scale, leading to cheaper hardware and app development costs. Traditional tariff models are suffering from the fact that the statistical methods in use will increase disproportionately in complexity for small improvements. By contrast, telematics-based scoring is more accurate and may even replace standard tariff criteria further down the road. Even if a telematics score only achieves the same result as the classic actuarial analysis, it will still do so in real time, thus reducing the number of pricing variables and the complexity of contract closure. Figure 4 Example of a possible risk measure: Modelling the amount of expected crashes per insurance period from a telematics analysis. relative probability of crash 1 2 3... # crashes per time period A contextual risk assessment is crucial, since we all know that racing home through a blizzard is more dangerous than taking the car for a spin on a sunny day. 10 Swiss Re Telematics in motor insurance

Closer ties with consumers via value-adding services and driving behavior feedback Cost reduction is not the only argument: based on telematics data, insurers can enrich their motor insurance offering with value-adding services and gamification. E-call and b-call functionalities are probably the best-known services that will increase drivers safety. The list of opportunities is long and includes stolen vehicle recovery, which is an attractive feature for markets with high theft rates, and find my car or geo-fencing features for parents of young and inexperienced drivers. Extended (and potentially automated) vehicle trip logs (for business trips that need to be expensed) or parking spot finders are also common features. Some insurers are building up entire ecosystems around telematics, featuring rewards such as a free coffee at a gas station in recognition for good driving behaviour. Driver feedback and coaching through insurance web portals and/or apps are also powerful tools for the insurer to positively influence a policyholder s driving habits. Additionally, gamification increases the insurer s interaction with its policyholders, potentially strengthening customer loyalty and offering valuable cross-selling opportunities, eg for travel or household insurance. Taking this a step further, real-time risk information may even create a more profound paradigm shift in the way insurance is sold. As telematics data enables insurers to rate the risk of a policyholder s individual trips, insurance contracts could increasingly deviate from the standard annual contracting period. Starting with mileage-based products, insurers could move from fully developed pay-how-you-drive or pay-as-you-drive models to solutions that insure individual trips. Such models will gain in importance, hand in hand with the growth of the sharing economy. Not only can telematics-based technologies open new ways of pricing and selling insurance, but they can also enable more efficient claims handling. Starting with the detection of the crash through telematics data, the whole process can be improved and accelerated for the benefit of the policyholder (more timely payment) as well as the insurer (decreasing costs through less involvement of claims handlers and lawyers). Further efficiency can be gained by fully digitalising the claims handling process. An additional benefit is that the insurer will be involved in the claims handling process from the start and can steer a policyholder to its preferred garage network a part of the value chain that is increasingly being owned by OEMs or other third-party providers. The big dilemma. Is there enough value for consumers to share their data? The bottom line is that end consumers do not only want cheaper, but also more convenient insurance solutions and processes. Insurance needs to be easy in all its phases: enrolling, announcing an accident, claiming a benefit. With data increasingly becoming a real currency not only for companies, but also the end users insurers have to protect data and the users privacy. The benefits for a policyholder to share more data with its insurer therefore need to be obvious and tangible, while data handling needs to be transparent and strictly protected. The challenges of bringing telematics to consumers For many insurers, especially small and mediumsized players, setting up telematics capabilities can be a resource-intensive, complex and costly process. Challenges include compiling vast amounts of data to develop a proper scoring, finding the right telematics technology provider or sensor manufacturer that offers a competitive solution and reliable data, as well as building up know-how and expertise. Without a doubt, the development cost of telematics insurance policies is higher than for standard products. But so are the opportunity costs of missing out. Telematics in motor insurance Swiss Re 11

Helping insurers prepare for the future motor insurance market In collaboration with selected strategic partners, Swiss Re has built up comprehensive telematics capabilities and is developing additional expertise and technology to provide clients in selected markets with an end-to-end telematics solution in return for reinsurance. What s next? Motor is among the first lines of business to be affected by digitalisation. This is an opportunity for insurers to build up capabilities in preparation for the way underwriting will change in other areas, such as property and life and health. Swiss Re can provide primary insurers with any form of support their individual product and tariff approach demands. Swiss Re has acquired a leading telematics solution provider, akquinet SLS, to provide telematics infrastructure. We have also established a specialised scoring & analytics team to develop a machine learning-based scoring platform, including actuarial services to support clients in interpreting and developing their own driver scoring. Furthermore, we seek to engage in dialogue with technology companies, motor OEMs, governments and regulators in order to contribute to the discussion on liability and autonomous driving, and to develop innovative insurance solutions. For example, in the area of fleet telematics, we are working on combining a telematics solution with retrofitted driver assistance systems to improve safety, reduce the risk of traffic accidents and reduce operating costs. Swiss Re s end-to-end telematics offering consists of the following three pillars Telematics infrastructure (in selected markets only) Complete and accurate solution. Any device (eg a black box, 12V plug, OBD II or a smartphone) can be connected to the telematics platform linked to a fully customizable front-end (app and web) for driving feedback. Analytics & Scoring (in selected markets only) Dynamic, machine learning-based scoring platform being developed by data scientists and actuaries dedicated to motor telematics. The platform feeds back scores of each policyholder through an insurer web-portal. This allows insurers to access the raw data, do analytics on the portfolio or individually adapt the scoring. Reinsurance Swiss Re offers its telematics solution (end-to-end or scoring only) in return for reinsurance. Swiss Re shares the risk and helps insurers to start engaging in telematics, to mitigate actuarial uncertainties arising from the lack of claims data and to spread costs. At Swiss Re, we believe that telematics solutions should be modular and give insurers transparency and flexibility: besides making use of the end-to-end solution, from device, app, post-processing and scoring, insurers can also internalise any steps of the process. 12 Swiss Re Telematics in motor insurance

Telematics in motor insurance Swiss Re 13

About The Swiss Re Group is a leading wholesale provider of reinsurance, insurance and other insurancebased forms of risk transfer. Dealing direct and working through brokers, its global client base consists of insurance companies, mid-to-largesized corporations and public sector clients. From standard products to tailor- made coverage across all lines of business, Swiss Re deploys its capital strength, expertise and innovation power to enable the risk-taking upon which enterprise and progress in society depend. Founded in Zurich, Switzerland, in 1863, Swiss Re serves clients through a network of around 70 offices globally and is rated AA- by Standard & Poor s, Aa3 by Moody s and A+ by A.M. Best. Registered shares in the Swiss Re Group holding company, Swiss Re Ltd, are listed in accordance with the International Reporting Standard on the SIX Swiss Exchange and trade under the symbol SREN. For more information about Swiss Re Group, please visit: www.swissre. com or follow us on Twitter @SwissRe. Authors Andrea Keller Business Development Manager, Swiss Re Automotive Solutions andrea_keller@swissre.com Fabian Transchel Scoring & Analytics Expert, akquinet SLS fabian.transchel@akquinet.de Swiss Re Automotive Solutions The Automotive Solutions unit works with clients and partners to create leading telematics and autonomous vehicle insurance solutions globally, in exchange for reinsurance. akquinet SLS Based in Bremen (Germany) and a fully owned subsidiary of Swiss Re, akquinet SLS is a technology company founded in 1999 offering customized telematics solutions to primary insurers and fleet operators. 14 Swiss Re Telematics in motor insurance

2016 Swiss Re. All rights reserved. The entire content of this publication is subject to copyright with all rights reserved. The information may be used for private or internal purposes, provided that any copyright or other proprietary notices are not removed. Electronic reuse of the data published in this white paper is prohibited. Reproduction in whole or in part or use for any public purpose is permitted only with the prior written approval and if the source reference Telematics: connecting the dots is indicated. Courtesy copies are appreciated. Although all the information used in this study was taken from reliable sources, Swiss Re does not accept any responsibility for the accuracy or comprehensiveness of the information given or forward-looking statements made. The information provided and forward-looking statements made are is for informational purposes only and in no way constitutes or should be taken to reflect Swiss Re s position, in particular in relation to any ongoing or future dispute. In no event shall Swiss Re be liable for any financial or consequential loss or damage arising in connection with the use of this information and readers are cautioned not to place undue reliance on forward-looking statements. Swiss Re undertakes no obligation to publicly revise or update any forward-looking statements, whether as a result of new information, future events or otherwise. The information and opinions contained in this document are provided as at the date of the document and are subject to change without notice. This document does not constitute or form part of, and should not be construed as, an advice and the reader is requested to obtain its own advisors. Furthermore the document does not constitute or form part of, and should not be construed as an offer for sale or subscription of, or solicitation of any offer or invitation to subscribe for, underwrite or otherwise acquire or dispose of any securities and it does not constitute an invitation or inducement to engage in investment activity. Order no: 1507215_16_EN