Big Data Analytics and Insurance Paul MacDonnell @pmacdonnell 2ND Annual Global Insurance Distribution & Bankassurance Conference May 13, 2015
ABOUT THE CENTER FOR DATA INNOVATION The Center for Data Innovation is a nonprofit think tank studying the intersection of data, technology, and policy. The Center s mission is to educate policymakers about the opportunities to use data-driven technologies to promote economic growth and societal progress. To learn more, contact the Center: Web: datainnnovation.org Twitter: @datainnovation Email: info@datainnovation.org
WHAT IS DATA-DRIVEN INNOVATION? New technologies have made it easier and cheaper to collect, store, analyze, use, and disseminate data. Data-driven innovation combines data with software and hardware technology and allows us to: Better understand our world, make better decisions about it, and implement these decisions. This allows us to address challenges in areas such as: Business, education, health, public administration, environmental protection, transport, energy, and security.
WHAT IS DATA-DRIVEN INNOVATION? Mobile retailing Performance tracking Health monitors Online courseware Customer data Monitor the environmental Social networks More complete and up-to-date data about more areas of activity can yield new insights. Big-data derived intelligence about an entire patient, user, or customer dataset will help design better solutions. Smart objects connected to the Internet of Things and new business and service delivery models can provide new kinds of services and make possible what used to be luxury. Genomics Navigation Smart manufacturing Monitor crop-yields Monitor human-rights abuses By 2020, there may be 50 billion connected devices, contributing up to $11 trillion in value per year globally by 2025. McKinsey Global Institute estimates that innovation using big data could save Europe s public sector more than 100 billion annually.
DATA ANALYTICS AND INSURANCE - Transforming risk analysis. - Insurance distribution, algorithms and Internet platforms. - The challenge of anti-discrimination and data protection rules. - Public policy recommendations.
TRANSFORMING RISK ANALYSIS - Underwriters use a narrow range of sources to determine risk. - Personal insurance: e.g. income, age, health, occupation. - General insurance: e.g. value, weather, previous experience (using underwriter s own experience). - New sources of data and analysis reveal new correlations, e.g. - Driver safety and on-time payment of bills. - Treatment and health outcomes. - Health & safety practices and rate of workers injury claims
TRANSFORMING RISK ANALYSIS - New sources of data - Open data: e.g., Copernicus, European Science Cloud, statistics. - Innovative Medicines Initiative - Data exhaust: e.g. the Internet, social media. - Internet of Things: telematics, health implants - Personal fitness trackers - Genomics. - New analysis tools that, e.g. - blend mortality, health, diet & lifestyle trends - monitor and track the accumulation of risk on an account and guide the underwriter. - Identify key behaviours, elements that affect overall risk and price of cover
Insurance distribution process: insurer / distributor perspective R&D Market research. Product design. Product introduction Risk assessment / underwriting. Advice. Personalised quote. Negotiation. Creating the contract Contract signing. Policy issued. Premium payment. Post sale Policy admin Claims management. Risk management.
Insurance distribution process: Customer perspective Review market Identify suitable products. Compare benefits Compare pricing. Product introduction Narrow choice down to 2 or 3. Negotiation Creating the contract Contract signing Policy issued Premium payment Post sale Make claim Renew
Structured Data Unstructured Mobile phone/gps Social networks Credit history Blogs Travel Property Census sensors CRM HR Records Twitter Online forums Sales records Financial results Fitness apps
- Traditional channels still dominate: brokers, banks and affinity groups. - Aggregator/price comparison sites are now important (personal motor and property insurance + cover of small business). - Data enabled touchpoints with consumers will grow: telematics, geolocation, personal devices. - More data will make more people insurable.
- Big data allows customer demands to be better understood and products and pricing to be personalised. - UK motor market is leading the way. - Intermediaries are still important. - Insurance can be more viable / affordable.
- In no region do intermediaries have more than 50% of life or general market. - Globally, intermediaries have 60%, direct is 25% and rest includes affinity and bancassurance. - In EU e-commerce is 14% of all sectors and less than 5% for insurance. - Internet use has risen five-fold since 2000.
- In 2012 McKinsey found that 73% of individuals shopping for auto insurance in the U.S. used the Internet for information gathering, up from 55% in 2008. - Life Insurance Marketing and Research Association (LIMRA) found that 61% of US consumers researched life and annuity products online in 2012, compared with 38% in 2006.
- Web aggregators are becoming standard tools in U.S., UK., Netherlands, Germany, France, Spain, Italy, Australia, Hong Kong, Ireland, South Korea and Canada. - Swiss Re found that consumers trust the Internet more than any other source of advice (2012)
- Information and online tools are more available. - Websites, expert and consumer blogs, chat rooms. - Need calculators use income to help determine risk, consumers can calculate their own risks and choose cover on that basis in 2013 South African insurer Liberty introduced its web-based Risk Revealer tool that uses a personalised avatar created from the responses to an online questionnaire. - Price comparison sites will begin to accommodate more complex products. -
Technology provides insurers and consumers with opportunities to - use social media - To create peer-to-peer insurance, for example the UK s jfloat, or Germany s Friendsurance which connects people willing to share possible losses below the deductible in standard insurance policies. - To buy/sell insurance, e.g. Malayan (Philippines) and Aegon s Kroodle (Netherlands) on Facebook. - 60% of insurance companies reported plans to develop social media distribution capability (2013). - use Mobile devices - To report and photograph damage caused by accidents and weather (Aviva iphone app). - To track behaviour thus expanding cover to the previously uninsurable. - develop usage-based insurance - Through the use of telematics (number of monitored drivers worldwide will rise from 1.85 million in 2010 to 89 million by 2017). -
The opportunity - Europe s insurers have 8.5 trillion of assets under management and 35% of global insurance premiums. - Emerging markets today account for 40% of global gross domestic product but only 17% of global insurance premiums. - Uninsurability can be addressed through better data and Internet-based distribution.
THE CHALLENGE OF ANTI-DISCRIMINATION AND DATA PROTECTION RULES - General Data Protection Regulation will impose limits on use of personal data. - Right to be forgotten - Consent needed for data reuse. - Proposed Anti-discrimination directive - Exempts insurers on basis of price differentials that are actuarially justified. - Experience of Equal Treatment Gender Directive suggests that this may be insufficient to protect use of age-related data.
THE CHALLENGE OF ANTI-DISCRIMINATION AND DATA PROTECTION RULES - Proposed Anti-discrimination directive - Draft text mis-construes the role of actuarial data in insurance - Emerging view in the EU: personal data as a tradable commodity or currency - Statements by Commissioners Vestager and Oettinger.
THE CHALLENGE OF ANTI-DISCRIMINATION AND DATA PROTECTION RULES - Potential implications of data-driven innovation for insurance are profound. - It is unlikely that policymakers and regulators will remain on the side-lines in the face of significant innovation. - Potential for abuse e.g. redlining, exists - Insurers need a dialogue with policymakers and regulators - Explore the possible use of industry / sectoral codes of practice to prevent surprises.
THANK YOU! PMACDONNELL@DATAINNOVATION.ORG TWITTER: @PMACDONNELL