Big Data - Transforming Risk and Insurance. Driving Change

Transcription

1 Big - Transforming Risk and Insurance George Attard Head of Aon Benfield Analytics, International Catastrophe Risk Management Market Challenges Lack of risk awareness Low disposable income High concentrations of risk (Asian mega cities) Anti selection => high premiums => inhibits sales Local Insurers in Asian markets face a number of challenges in actively promoting and extending cover for Nat Cat perils Difficulty in generating economically viable risk pools with sufficient diversification of risk and adequate premiums Challenge of modelling and pricing potentially high severity risks with little valid experience Driving Change Solid economic growth Low insurance penetration Low reinsurance premiums relative to population size Increasing wealth, urbanisation driving exposure increase Growing middle class driving personal lines growth Increasing frequency and severity of natural catastrophes increasing demand and review of regulation, models and mitigation strategies Evolving regulatory and rating agency environment Geographic and cultural diversity 2 Positioning for Success resolution Underwriting and pricing Structuring and placement of appropriate risk transfer solutions Catastrophe modelling Catastrophe Risk insight and understanding impact of data and modelling limitations 7th Model development to enhance understanding of risk and facilitate product and market development Product Innovation underwriting expertise and local insight

2 : the raw ingredient of decision making [Big] Three Vs Volume Variety Velocity Value Veracity BIG = hard to manage on existing technology B = Behavioral : Comparison of Insurance and Internet Big Models Insurance Sparse, expensive data Policy, claim, client Expensive to collect & maintain Static Structured Inhomogeneous within type Potentially high value density Low to medium volume 3 Internet Massive, free data Internet of Things Free: user contributed, digital exhaust Streaming Unstructured - variety Homogeneous within type 7th Low value density Very high volume Successes and potential Credit Direct Insurance growth Catastrophe modeling Memes, curiosities and some value Google Flu, translate Amazon Your recommendations Click on me!! 4

3 : the raw ingredient of decision making data What puts the Big in Big? Traditional data Observation Outcomes Features Characteristics 1 2 p Big 5 Observations = insureds Observed quantities Losses Age Sex Marital status Vehicle use Accident history Etc. 7th n n > p Observations = sentences Observed quantities Word frequencies 6

4 What puts the Big in Big? Extension I: Lots more observations Observation 1 2. n n+1 1,000,000 n Outcomes Features Characteristics 1 2 p Big n >> p What puts the Big in Big? Extension II: Lots more parameters Observation Sentences, results radically improve with billions of test sentences Global satellite images Tick-level financial data Insurance examples More years of experience Some experience vs none! Computing challenge Same modeling approaches Outcomes Features Characteristics 1 2 p p 7th 7 Big n p > n 8

5 What puts the Big in Big? Extension II: Lots more parameters Detailed credit history, in use since mid-1990s Observation n Outcomes Features Characteristics 1 2 p p p > n The insurance market Big Minute by minute driving log, auto telematics Home telematics Genome information Hyperspectral image Computing challenge New modeling challenge 7th 9 Source: Global Insurance Market Opportunities, Insurance Risk Study, 10 th Edition, Aon Benfield 10

6 How does more data impact risk and insurance? Unknown, ignorance, no insurance Partial knowledge, uncertainty, risk, insurance Complete knowledge, certainty, managed, retained How does more data impact risk and insurance? Extension I. More observations, bigger n Insurance gain from decreased ignorance Incomplete understanding creates opportunities for insurance markets Ignorance and certain knowledge generally rule out insurance 11 More observations can create markets Risk measured by risk owners Measurement begets management Risk more quantifiable for insurers 7th More data is a GOOD THING Emerging Risk Property catastrophe Cyber Business interruption Terrorism Giga liability Brand 12

7 How does more data impact Catastrophe risk and insurance? Extension I. More observations, bigger n Insurance gain from decreased ignorance More data is a GOOD THING Emerging Risk Demand: Aon 2015 global risk manager s top risks Damage to reputation/ brand Property damage Capital availability/ credit risk Merger/ acquisition/ restructuring Economic slowdown/ slow recovery Commodity price risk Directors & Officers personal liability Regulatory/ legislative changes Cash flow/ liquidity risk Failure of disaster recovery plan Increasing competition Technology failure/ system failure Corporate social responsibility/ sustainability attract or retain top talent Distribution or supply chain failure Injury to workers innovate/ meet customer needs Political risk/ uncertainties Crime/ theft/ fraud/ employee dishonesty Better weather prediction models to leverage risk transfer solutions Early warning for thunderstorms, typhoons allow risk mitigation strategies outside of insurance Development still required for: Typhoon impacts Severe multiyear drought events Impact of climate change 13 Business interruption Corporate governance/ compliance burden Loss of intellectual property/ data Third-party liability Exchange rate fluctuation implement or communicate strategy 7th Environmental risk Inadequate succession planning Lack of technology to support business needs Workforce shortage Product recall Accelerated change in market & geopolitics Aging workforce and related health issues Computer crime/ hacking/ viruses Weather/ natural disasters Counter party credit risk Globalization/ emerging markets Interest rate fluctuation Outsourcing Unethical behavior Natural resource scarcity Terrorism/ sabotage Asset value volatility Understaffing Pandemic risk/ health crises Climate change Social media Absenteeism Joint venture failure Share price volatility Pension scheme funding Sovereign debt Kidnap and ransom/ extortion Harassment/ discrimination 14

8 Demand: Aon 2015 global risk manager s top risks Damage to reputation/ brand Economic slowdown/ slow recovery Regulatory/ legislative changes Increasing competition attract or retain top talent innovate/ meet customer needs Business interruption Third-party Party liability Liability Computer crime/ hacking/ viruses Property damage Damage Capital availability/ credit risk Merger/ acquisition/ restructuring Interest rate fluctuation Commodity price risk Directors & Officers personal liability Environmental risk Outsourcing Social media Absenteeism Cash flow/ liquidity risk Failure of disaster recovery plan Inadequate succession planning Unethical behavior Joint venture failure Technology failure/ system failure Corporate social responsibility/ sustainability Lack of technology to support business needs Natural resource scarcity Share price volatility Economic slowdown/ slow recovery Regulatory/ legislative changes Inadequate succession planning Lack of technology to support business needs Insurable & Generally Insured Third-party liability Property damage Insurable & Not Enough Business interruption Computer crime/ hacking/ viruses Distribution or supply chain failure Injury to workers Workforce shortage Terrorism/ sabotage Pension scheme funding Political risk/ uncertainties Crime/ theft/ fraud/ employee dishonesty Product recall Asset value volatility Sovereign debt Insurance Design Challenges Damage to reputation/ brand Failure of disaster recovery plan Corporate governance/ compliance burden Loss of intellectual property/ data Accelerated change in market & geopolitics Understaffing Kidnap and ransom/ / extortion Demand: where are the new product opportunities? General Business Risk Increasing competition attract or retain top talent innovate/ meet customer needs Corporate goverburden implement or communicate Merger/ acquisition/ restructuring Workforce shortage Outsourcing Unethical behavior Understaffing Joint venture failure Weather/ natural disasters Directors & liability Injury to workers Crime/ theft/ fraud/ employee dishonesty Kidnap and ransom/ extortion Technology failure/ system failure Distribution or supply chain failure Political risk/ uncertainties Product recall Terrorism/ sabotage Absenteeism responsibility/ sustainability Loss of intellectual property/ data Social media Social or Global Risk Environmental risk Accelerated change in market & geopolitics issues Globalization/ emerging markets Natural resource scarcity Pandemic risk/ health crises Climate change 15 Exchange rate fluctuation implement or communicate strategy Aging workforce and related health issues Pandemic risk/ health crises Harassment/ discrimination Financial Risk Commodity price risk Cash flow/ liquidity risk Exchange rate fluctuation Capital availability/ credit risk Counter party credit risk Interest rate fluctuation Asset value volatility Share price volatility Weather/ Natural natural Disasters disasters Counter party credit risk Globalization/ emerging markets Climate change Pension scheme funding Sovereign debt 7th discrimination 16

9 Demand: where are the new Catastrophe product opportunities? Traditional cat insurance area Parametric or modelled loss solutions for Disaster Relief Agencies Rapid payment of funds the critical issue Public sector (infrastructure, municipal buildings, public services) Industrial/ Commercial Property Financed housing / Insured for fire Non-financed/ low-income housing Community/Nationwide Disaster Relief A) Parametric or ILS solutions Rapid payment to provide necessary budget liquidity following disaster Global reinsurance market provides the solution for state B) Indemnity Insurance Solutions provided by commercial market Field of traditional commercial insurance in Asia Market PPP Cat Pool/Scheme Compulsion creates the basis for a viable risk pool with affordable premiums Private sector provides policy distribution, loss assessment and some risk assumption Public sector provides cheap risk capital through government guarantees Global reinsurance market provides risk capacity Demand: where are the new Catastrophe product opportunities? Innovation Developing innovative ways to reach customers ie mobile phone applications, eg micro insurance Insurable & Not Enough Catastrophe bonds for governments protecting Government finances Disaster relief Infrastructure Event based coverage 7th Index insurance based on satellite images 17 Insurance Design Challenges Government disaster mitigation policies Climate Change 18

10 Demand: Why Buy? Risk assessment and and risk management risk management Lower loss cost PML assessment, construction Risk control: WC, auto, liability, med mal, Building engineering Monitoring: surety, credit, political risk Insurance Non-market motivations Non-market motivations Rating agencies & regulation price cat risk Regulation spurs demand for mortgage credit Claims Claims handling Handling.word Lower cost of LAE Lower ultimate cost of indemnity Insurance or reinsurance Terrorism How does more data impact risk and insurance? Extension II. More parameters, bigger p Risk Risk financing Financing word Lower cost of capital Reinsurance CYBER 19 More parameters may destroy markets Granular underwriting Less risk sharing Affordability and availability issues Risk signalling can create a more resilient society 7th Insurance loss from greater certainty More data is a BAD THING Existing Risk Genomics in health insurance Flood insurance 20

11 Customisation: personalise service and disaggregate the risk pool Customised Service How does more data impact risk and insurance? Remains to be seen Indeterminate net growth effect Veracity Anti-Fraud Personalised Coverage Customised Pricing 21 Net growth impact on risk-transfer insurance indeterminate Different data models apply in different markets 7th Disruption is certain More data is a??? Disruption 22

12 Ten trends will define the next ten years 1. The Big technology connected world revolution will continue to grow 2. We will collect more detailed individual underwriting data 5. We will collect more loss experience 3. We will continue to improve risk management 6. We will demand more effective risk management The insurance market 4. Existing risk lines will continue to shrink relative to GDP 7. New opportunities will arise from emerging risks 8. Capital will continue to be plentiful New players will drive creative destruction 10. Tomorrow s winners will provide holistic risk management services 7th Source: Global Insurance Market Opportunities, Insurance Risk Study, 10 th Edition, Aon Benfield 24