How Can YOU Use it? Artificial Intelligence for Actuaries. SOA Annual Meeting, Gaurav Gupta. Session 058PD

Transcription

1 Artificial Intelligence for Actuaries How Can YOU Use it? SOA Annual Meeting, 2018 Session 058PD Gaurav Gupta Founder & CEO

2 Audience Poll What is my level of AI understanding? Novice Advanced beginner Competent Proficient Expert

3 1 What is AI? Audience Poll What is the level of AI adoption in my organization? Talking but have not done anything Dipping our toe in the water Multiple pockets of usage Ubiquitous 6 Unknown 3

4 Audience Poll Usage of AI in Actuarial functions Thinking about it Actively looking for use cases We have a few use cases We use it in every possible way 5 None of the above

5 Powering Life Insurance with AI techniques SVM Clustering Dimensionality reduction Pricing Underwriting Risk management Regression GLM Monte Carlo GAM Simulation Supervised learning Random forest Gradient boosting machine Unsupervised learning AI Recommendation system Decision trees Deep learning Bayesian NLP RNN Reinforcement learning Distribution Antiselection Cross-sell Retention Product development Life Insurance Premium persistency Emerging experience Mortality study Assumption setting Reserving Reinsurance How to build the bridge? 5

6 Where is AI being used in Life Insurance? PRICING UNDERWRITING RETENTION PRODUCT DEVELOPMENT RESERVING CROSS-SELL PERSISTENCY LAPSE EXPERIENCE STUDY MORTALITY STUDY ANTI-SELECTION RISK MANAGEMENT ASSUMPTION SETTING 6

7 Where is AI being used in Life Insurance? PERSISTENCY MORTALITY STUDY ANTI-SELECTION EXPERIENCE STUDY ASSUMPTION SETTING CROSS-SELL PRICING RESERVING LAPSE RISK MANAGEMENT PRODUCT DEVELOPMENT RETENTION UNDERWRITING AI + LIFE INSURANCE * Data collected from Google scholar search, number of articles 7

8 What if you could Monitor mortality constantly Add and analyze unlimited data, new variables quickly Predict policyholder behavior (shock lapse, anti-selection, ) Develop new mortality assumption in half an hour Price at an individual level Develop customized products on-the-fly 8

9 Why AI has not created more benefits for actuaries? Data Regulatory Other Death is a rare event, and has a long duration low credibility Inconsistency of data dimensions, variable definitions etc. across systems, over time Blackbox is a no-no Regulation Hurdles for AI to fit existing paradigm Talent to understand both AI and life insurance is scarce Difficulties in changing business processes and systems Lengthy validation 9

10 AI can help you do things easier, faster, and with improved insights Example Areas How AI Achieves it Experience studies Automate reporting More powerful data handling - 1) more data sources; 2) easier data manipulation; 3) faster Easier to develop and monitor KPIs (e.g. deviations, trends) Large selection of algorithms, software and platforms Seamless integration of data, technology, methodology and business metrics Case Study: Experience Study Uncover hidden insights about mortality risk Integrate underwriting, transaction, claim, external data in real-time Discover key drivers for mortality/lapse deviation Detect early alerts for better/worse mortality trends Generate reports in minutes instead of weeks

11 AI can help you do things easier, faster, and with improved insights Uncover hidden insights about mortality risk Actuals to Expected (A/E) ~500,000 exposures Detect and discover segments of mortality deviation verifiable, consistent and credible BAU without AI <10 Variables With AI ~100 Variables > 120% Segment A/E Credibility Seg 1 130% 100% Seg 2 120% 85% Seg 3 125% 60% ~100% In a few hours < 80% Segment A/E Credibility Seg 4 75% 80% Seg 5 50% 70% Seg 6 65% 65% 11 11

12 AI can help you do things better Example Areas Assumption setting Projection of claims, lapses, surrenders, withdrawals Protective Value studies Segmentation Reinsurance structuring and pricing Stress and scenario testing How AI Achieves it Supplement actuarial credibility with AI validation assessment metrics Automatically detect and utilize correlations and interactions Tools to capture more reliable, sustainable relationships to minimize overfitting Case Study: Assumption Setting Develop better assumptions that satisfy inherent dimensional relationships Set assumptions at more homogenous level, e.g. by distribution channel Provide more accurate forecasting for future events (death, lapse, surrender, etc.) Estimate marginal impacts of underwriting requirements 12

13 Issue Age AI can help you do things better Develop better assumptions that satisfy inherent dimensional relationships Duration AI can Borrow information from adjacent cells along multiple dimensions Enforce complex relationships Select, ultimate periods Omega rate Monotonic constraints The cell does not have enough credibility to make any adjustment With borrowed information from adjacent cells, we can be more confident to make adjustment * VBT 2015 Female, NS table 13

14 AI can help you do new things Example Areas Completely fluid-less underwriting New underwriting, pricing process powered by image, text, voice, cognitive AI New capability to provide more granular (closer to individual level) pricing Personalized product How AI Achieves it Capability to handle large volumes of data Create more powerful features via combinations of image/text/voice/cognitive AI Larger selection of algorithms Future Use Case: On-demand Insurance Create personalized life products variable duration, structured riders, unique risks, etc. Determine individualized pricing for the personalized product Determine, gather and process all sources of data on the applicant, coverage and scope (e.g. IoT, images and voices, social media, circumstance specific risk factors, etc.) Estimate mortality risk for that circumstance (e.g. traveling to foreign countries, participating in risky events, covering for predefined period, etc.) 14

15 How to get started Pick someplace to start Small enough problem to finish in 3-5 months Delivers business value Spend time defining the business problem Example: Do I want to predict mortality or mimic the underwriter decision in my predictive underwriting model? Do I want to optimize response to direct marketing campaigns or response of prospects with the best risk profile? Other? Start with the data you have The best source of data is your own, e.g. your experience data. You can get more out of it than you think More data does not equal better results harder to find a needle if the haystack is bigger Assemble the right team Actuaries + Domain Experts + Data Scientists 15

16 Thank You! If you have more questions on AI, please come see us at Booth 519 in the Exhibit Hall, or contact me at 16