2011 Fair Isaac Corporation.

Transcription

1 Fair Isaac Corporation.

2 Use of Credit Scores in the Property & Casualty Insurance Industry Boost Operating Efficiency and Underwriting Profit with Predictive Analytics Lamont D. Boyd, CPCU, AIM Insurance Industry Director, Scores and Analytics FICO (Fair Isaac Corporation) 2 This presentation is provided for the recipient only and cannot be reproduced or shared without Fair Isaac Corporation's express consent Fair Isaac Corporation.

3 Agenda» FICO Analytics in Insurance» FICO Insurance Scores» Background» Industry Usage» Discussion Fair Isaac Corporation Fair Isaac Corporation. Confidential.

4 FICO is the leader in Predictive Analytics and Decision Management. Thousands of businesses worldwide, including nine of the top ten Fortune 500, rely on FICO to Make Every Decision Count Fair Isaac Corporation. Confidential Fair Isaac Corporation.

5 FICO Overview: Committed to the Insurance Industry We have provided services and technologies to companies in multiple industry sectors property & casualty (personal and commercial lines), life & health, and brokerage. Plus over 350 other insurers worldwide, including: Aetna Amerigroup Corporation Amica Mutual Insurance Co. Atlantic Mutual Insurance American Family Insurance Aviva Chubb Co-Op Network Farmers Insurance Group Fireman's Fund Insurance Co. Great American Insurance Guardian Life Insurance Kemper National Insurance Liberty Mutual Insurance Liberty Northwest Insurance Mercury Insurance Missouri Employers Mutual Mutual Of Enumclaw Insurance Nationwide Mutual Insurance Norwich Union OH Bureau Of Workers' Comp One Beacon Insurance Peerless Insurance Company Prudential Insurance Safeco Travelers Unitrin VHI Westfield Insurance Zurich Insurance Company Fair Isaac Corporation.

6 FICO s Solutions for Insurance Power profitable decisions across core insurance processes while increasing precision in the numerous customer decisions» Build a profitable book-of-business» Bring right products to market, faster» Faster, more efficient claims processing and management» Acquire and grow profitable customers» Increase customer loyalty and retention» Fight fraud from beginning to end Two-thirds of the Top US P&C Insurers Fair Isaac Corporation.

7 FICO Insurance Analytic Solutions Insurance Industry Property & Casualty Life Health Rules Mgmt, UW Profit, Claims P&C Personal P&C Commercial Private Payors Government Auto CBIS Scores Custom Risk Models Fraud Detection UW Decision Rules Mgmt. Homeowners CBIS Scores Custom Risk Models UW Decision Rules Mgmt. Workers Comp Fraud Detection Loss Reserving Subrogation Custom Risk Models Rules Mgmt. Property & Liability Rules Mgmt. Rules Mgmt. Fraud Detection Federal Medicare Fraud Detection State Medicaid Fraud Detection Fair Isaac Corporation.

8 D a t a D a t a FICO s view: Analytics in operation External Data DATA MANAGEMENT ANALYTICS (Models, Scores, Strategies) STRATEGY DEPLOYMENT Internal Data Prodn./ Ops. Env. Data Modeling Analytic Data Mart Cleansing Exp. Design Predictive Modeling Exploratory Analysis/ Data Mining Decision Modeling and Strategy Design Decision Engine Rules Mgmt. Case Mgmt. Prodn./ Ops. Env. Reactions Prospects & Customers Actions Fair Isaac Corporation.

9 Agenda» FICO Analytics in Insurance» FICO Credit-Based Insurance Scores» Background» Industry Usage» Discussion Fair Isaac Corporation Fair Isaac Corporation. Confidential.

10 FICO Credit-Based Insurance Scores» CBIS introduced in the US in 1993, Canada in 1995» Current generation, commercially-available FICO CBIS» Equifax US and Equifax Canada» Experian US» TransUnion US» Custom FICO CBIS Models (FICO Model Builder)» Developed for individual clients, based on underwriting strategies, past performance, and future needs» US industry estimates» ~90-95% of personal lines automobile and homeowner insurance underwriting and pricing decisions are based in part on CBIS 10

11 FICO Credit-Based Insurance Scores» Global FICO Scores for Insurance» Available in countries» Robust consumer credit data collection and reporting practices» Allow the insurance industry s use of consumer credit information» FICO Scores Partner in Mexico» Circulo de Credito» Juan Manuel Ruiz Palmieri Director Comercial» Ricardo Moreno Contreras Gerente de Productos de Riesgo 11

12 FICO Credit-Based Insurance Scores» FICO CBIS Benefits» Greater predictability of future loss» More accurate risk decisions for greater profitability» Facilitate efficient, consistent and objective decisionmaking for swifter processing» New and renewal underwriting, pricing, fraud detection» Account and book management 12

13 FICO Credit-Based Insurance Scores» Consumer benefits» Relationship between how people manage their credit and how they manage their risk is significant» Conscientious credit managers are conscientious risk managers IMPORTANT TO NOTE: Majority of consumers manage their credit well and benefit from the use of insurance scoring» Up to 75% of US auto and home insurance applicants pay lower premiums due to insurance scores 13

14 Insurance Scoring Background» Statistical correlation proven repeatedly» FICO studies» FICO CBIS client studies» Tillinghast Towers-Perrin study» University of Texas study» Epic Actuaries study» Texas Department of Insurance study» FICO & Círculo de Crédito study in México 14

15 Insurance Scoring Background» US Federal Trade Commission FTC» FTC Auto Study July, 2007» CBIS scores are objective tools for more accurate risk evaluation» Use benefits most consumers» CBIS scores cannot be used to identify demographic groups so no effect as proxy for race or ethnicity» CBIS scores not correlated to income levels but to an individual s management of credit obligations» Restricting CBIS scores would result in higher rates for better risks without regard to race and ethnicity» FTC Homeowner study 2013/2014 ETA» Precisely the same findings are anticipated 15

16 Insurance Scoring Background Study Tillinghast Towers-Perrin» Study Title: CBIS and Loss Ratio Relativities.» Study Date: Dec 1996» Sponsors:» 8 Auto, Homeowners Companies in US market.» Findings» The relationship between the CBIS and the loss ratio relativities is highly statistically significant, or» Is very unlikely CBIS and loss relativities are not correlated Fair Isaac Corporation.

17 Insurance Scoring Background Study EPIC Actuaries» Study Title: The Relationship of CBIS to Private Passenger Automobile Insurance Loss Propensity.» Study Date: Jun 2003» Sponsors: Alliance of American Insurers The American Insurance Association The National Association of Mutual Insurance Companies The National Association of Independent Insurers» Findings» CBIS correlated with propensity for loss, primarily with claim frequency, rather than with average claim severities.» CBIS do overlap with other risk characteristics, but finally CBIS increase the accuracy of the risk assessment process.» CBIS are among three primary risk factors for six automobile coverage's studied. - BI Liability - PD Liability - Pers.Inj.Prot - Med Pay - Comprehensive - Collision 9% 33% -19% Fair Isaac Corporation.

18 Insurance Scoring Background Study EPIC Actuaries» Study Title: The Relationship of CBIS to Private Passenger Automobile Insurance Loss Propensity.» Study Date: Jun 2003» Sponsors: Alliance of American Insurers The American Insurance Association The National Association of Mutual Insurance Companies The National Association of Independent Insurers» Findings» CBIS correlated with propensity for loss, primarily with claim frequency, rather than with average claim severities.» CBIS do overlap with other risk characteristics, but finally CBIS increase the accuracy of the risk assessment process.» CBIS are among three primary risk factors for six automobile coverage's studied. - BI Liability - PD Liability - Pers.Inj.Prot - Med Pay - Comprehensive - Collision 9% 33% -19% Fair Isaac Corporation.

19 Insurance Scoring Background Study EPIC Actuaries» Study Title: The Relationship of CBIS to Private Passenger Automobile Insurance Loss Propensity.» Study Date: Jun 2003» Sponsors: Alliance of American Insurers The American Insurance Association» Findings» CBIS correlated with propensity for loss, primarily with claim frequency, rather than with average claim severities.» CBIS do overlap with other risk characteristics, but finally CBIS increase the accuracy of the risk assessment process.» CBIS are among three primary risk factors for six automobile coverage's studied. - BI Liability - PD Liability - Pers. Inj. Prot. - Med Pay - Comprehensive - Collision The National Association of Mutual Insurance Companies The National Association of Independent Insurers Fair Isaac Corporation.

20 % LRR Insurance Scoring Background Study Círculo de Crédito» Study Title: CBIS and Loss Ratio Relativities in México.» Study Date: Jun 2013» Sponsors:» Auto Insurance Companies in Mexican market.» Findings» The conclusion is that it is very high that CBIS and loss ratio relativities, considering the average claim severities, are correlated. 150% 140% 130% 120% 110% 100% 90% 80% 70% LOSS RATIO RELATIVITY - RANK ORDER LOW MEDIUM HIGH 790 Score Fair Isaac Corporation.

21 FICO CBIS Scorecard Development» Using depersonalized data, over 100 credit characteristics are analyzed against insurance policy premium and loss records» Adjusted for relationship of overlapping characteristics» Statistical attribute weighting» Insurance scoring models» most predictive characteristics» No single factor drives the FICO CBIS score 21

22 FICO Analytics: Multivariate Modeling for Underwriting» Determine value of predictive information available» Calculate statistically accurate weights for most predictive information» Produce statistical model(s) which allow for greater policy processing efficiency and predict future performance Outstanding Debt Outstanding Debt Payment History Oldest Account VS. Payment History Oldest Account 22

23 FICO Model Development: Initial data analysis Fine Binning Coarse Binning» Common data issues» Incomplete and/or inaccurate data» Outlier anomalies can skew statistics» Highly correlated items blur the incremental value contributed by each predictive variable» FICO s approach improves results» Binning preserves patterns while reducing noise» No information bin effectively handles outliers» Tools automatically assess correlation impact & information value contributed for each variable 23

24 Loss Ratio Relativity FICO CBIS Models Rank-Order Example low Automobile Insurance high Score Range 24

25 FICO CBIS Five General Areas of Predictive Information Credit History Length 15% Outstanding Debt 30% Pursuit of New Credit 10% Credit Mix 5% Payment History 40% 25

26 Information NOT Considered by FICO CBIS Models» Race, color, national origin» Religion» Gender» Marital status» Age» Income, occupation or employment history» Location of residence» Any interest rate being charged» Child/family support obligations or rental agreements» Certain types of inquiries» Whether or not a consumer is participating in credit counseling of any kind» Any information that is not proven to be predictive of future performance» Any information not found in the credit report 26

27 FICO CBIS Usage Lifecycle: Customer Acquisition» New Business Underwriting, Pricing, and POS Fraud Strategy» Earliest CBIS application» Consistent and objective new business decisions» Quickly accepting applicants with greatest profit potential» Dedicating resources to appropriate risks» Minimizing exposure to riskier applicants» Highlighting potential fraud at point-of-sale 27

28 Loss Ratio New Business Underwriting, Pricing, and POS Fraud Strategy Max Surcharge Surcharged Pricing Discounted Pricing Levels Population by deciles Average

29 FICO CBIS Usage Lifecycle: Customer Relationship Management» Renewal underwriting/pricing strategy» Secondary CBIS application» Updating CBIS scores at renewal» Identifying renewal policyholders for swift processing» Highlighting renewal policyholders requiring greater attention» Renewal tier-placement and pricing relative to exposure» Consistent and objective renewal underwriting decisions 29

30 Loss Ratio Renewal Underwriting and Pricing Strategy N/R Refer to Underwriting Discounted Pricing Levels Population by deciles Average

31 Loss Ratio Profitable Customer Retention Strategy No Cross-Selling or Contact Regular Cross-selling and Contact Population by deciles Average

32 FICO CBIS Usage Lifecycle: Book Management» Book management» Managing books of business at variety of levels across the enterprise with a focus on production and profitability goals» Underwriting territory management for greater book understanding» Production source management for more effectively managed and strengthened relationships 32

33 Loss Ratio Profitable Book Management Strategy Monitor closely for action Reward production source Population by deciles Average

34 Key Takeaways» FICO Analytics are uniquely positioned for the insurance industry» 400+ clients over 30+ years» FICO Credit-Based Insurance Scores» Predictive scoring and analytics market leader for past two decades» Commercially-available or custom models» Enhanced decision-making at every point in lifecycle» Proven predictive solutions to meet client strategies and needs» Focused on regulatory compliance, operational efficiency, and greater profitability Fair Isaac Corporation.

35 Agenda» FICO Analytics in Insurance» FICO Credit-Based Insurance Scores» Background» Industry Usage» Discussion Fair Isaac Corporation Fair Isaac Corporation. Confidential.

36 Thank You In USA Lamont D. Boyd, CPCU, AIM Insurance Market Director, Scores and Analytics In Mexico Juan Manuel Ruiz Palmieri Sales Director (55) Ricardo Moreno Contreras Product Risk Manager (55) To learn more visit: FICO.com > Industry > Insurance 36 This presentation is provided for the recipient only and cannot be reproduced or shared without Fair Isaac Corporation's express consent Fair Isaac Corporation.