2020 In Focus Virtual Seminar - October 20-22

DAY 1 – Pandemics and the P&C Industry

11:00 AM - 12:30 PM, ET

The Road Less Traveled

The COVID-19 pandemic has impacted all aspects of everyone's lives in America, including driving behaviors. Like all companies, insurance carriers must find ways to keep up with these changes, understand their impact, and plan for ways to evolve their service to their customers in response. Panelists during this session will discuss how the spread of COVID-19 and corresponding government reactions have affected driving behaviors as measured via telematics data. Additionally, panelists will discuss the actuarial implications of these changed behaviors, including ratemaking implications.

Learning Objectives:

• Understand how telematics information can be used to gain real time insights in uncertain and unprecedented situations
• Understand the shifts in driving behaviors and risk on the road as well as the geographic variations in auto insurance exposure response to the spread of COVID-19
• Explain the long-term impact that pandemics can have on auto insurer's claims experience, including the driving factors behind those impacts and the auto insurance ratemaking implications that actuaries should consider

Speaker(s):
Eric Krafcheck, Consulting Actuary, Milliman
Sam Madden, Chief Scientist and Founder, Cambridge Mobile Telematics
Rob Nendorf, Director Data Science, Arity

1:45 PM - 3:15 PM, ET

Insights on WC, Company and Industry Estimates from Pandemics

The COVID-19 pandemic has impacted property/casualty insurance in multiple ways. Most P&C companies have had to provide estimates of the impact, in particular, for Workers’ Compensation insurance. We will have an open discussion on the top-of-mind issues such as the effects of the unprecedented changes in the workforce, expansion of presumptive benefits for first responders and health care workers, consideration of the 2020 experience in the rate filing process, and financial reporting questions. While there are many unknowns regarding COVID-19 and its ultimate impact on the US population, we will also discuss possible approaches to estimate the potential WC system cost impacts under various hypothetical scenarios that may be used to gain insight into the potential implications of a state enacting legislation related to the compensability of COVID-19-related claims.

Learning Objectives:

• Describe the imperative issues affecting WC insurance from pandemics; in particular, COVID-19
• Discuss considerations of approaches to estimate the impact on the WC system from the COVID-19 pandemic
• Explain the potential implications to companies and the industry from the evolving compensability of COVID-19-related claims

Speaker(s):
Sean Cooper, Practice Leader & Senior Actuary, NCCI
Derek Jones, Principal and Consulting Actuary, Milliman

4:30 PM - 6:00 PM, ET

Pandemics Risk transfer – Dealing Data, Demand and Price

Opportunities to provide new forms of protection need to address a real problem. While the pandemic has led to a resurgence in risk-transfer discussion, problems with price and data could put pressure on demand for protection. Learn how our industry can make pandemic risk transfer effective.

Learning Objectives:

• Evaluating reporting sources for pandemic risk transfer and other actuarial efforts
• Understand the factors that can cause disparate data reporting from authoritative and reliable data sources
• Refine assumptions more effectively in actuarial analysis involving pandemic data

Speaker(s):
Tom Johansmeyer, Co-Head, PCS, ISO/Verisk

DAY 2 – Implementing Data Analytics for Underserved Topics

11:00 AM - 12:30 PM, ET

Using Analytics for Claims Investigation During and After the Pandemic

Tough economic times lead to increased levels of fraudulent claims. This is especially true amid COVID-19. Identifying fraudulent claims post-COVID-19 will be more challenging, and companies are turning to analytics for help. Identifying claim fraud using predictive analytics represents a unique challenge. 1. Predictive analytics requires a target variable that can be analyzed. However, much historical fraud has not been identified, so the definition of a complete target variable is not possible. 2. Modelers assume the past will bear some resemblance to the future. Types of fraud change quickly, however, and make the analysis of a historical database less valuable for identifying future fraud. This is especially true given how COVID-19 has changed our lives. 3. Historical identification of fraud is inconsistent. This inconsistency can lead to erroneous model results due to data that is not homogenous. This presentation will demonstrate how analytics can be used in several ways to help identify fraud: 1. More consistent referral of suspicious claims to claim investigative units 2. Better identification of suspicious claims, even when techniques used to defraud insurers are changing 3. Incorporating claim adjuster insight into the analytics process to improve the identification of fraud

Learning Objectives:

• Understand how COVID-19 has significantly impacted the claims process
• Identify data sources to apply to claims analytics
• Identify model techniques that can be applied to different claims applicaitons

Speaker(s):
Roosevelt Mosley, Principal & Consulting Actuary, Pinnacle Actuarial Resources

1:45 PM - 3:15 PM, ET

Explainable Machine Learning: Practical Challenges and Lessons Learned

Data-driven and AI-driven decision-making cultures are no longer a fantasy. Machine learning strongly augments business functions across various industries, and Insurance is no exception. Insurance companies take advantage of ML to solve business challenges in fraud detection, claims handling, underwriting, loss prevention/mitigation and customer experience. While there is clearly great potential, there are also tough challenges on the road to realizing value: model risk management, operationalization, uncertainty quantification, ethical concerns, and interpretability. This talk illustrates why, when, and what kind of model-agnostic interpretability tools are important and useful in practice. The methods are presented using real world data and model examples.

Learning Objectives:

• At the conclusion of this session, participants will be able to recognize and use post-hoc model-agnostic interpretability methods like feature importance, surrogate models, partial dependence plots, individual conditional expectation plots and shapley additive explanations.
• Participants also will be able to leverage fairness-aware data mining techniques for bias detection, measurement and it's correction when it is necessary.

Speaker(s):
Yelena Kropivnitskaya, Manager, Data Analytics, Wawanesa
Erik Christianson, Manager, Data Analytics, Wawanesa

4:30 PM - 6:00 PM, ET

Data Architectures and their Potential Impacts on Analytics and Models

Improper assumptions about data can result in models that don’t deliver the intended outcome, or models that aren’t implementable. Understanding your data architecture – or how the data you use for analytics is captured, managed within the system, and transformed – is critical to properly using that data for analytics and predictive models. This interactive session will introduce key data architectural concepts and provide strategies you can use to make the most of your data.

Learning Objectives:

• Understand data architecture
• Identify data architecture impacts to the implementation of predictive models

Speaker(s):
Meghan Goldfarb, Technology Director, State Farm
Todd Lehmann, Quincy Mutual Fire Insurance Co.

DAY 3 – Expanding the Actuarial Toolbox

11:00 AM - 12:30 PM, ET

Business Savvy Actuaries - From the Back Office to Trusted Advisor

Actuaries have a deep understanding of insurance, through their technical analysis, knowledge of industry and institutional data, and interactions with a variety of business partners. Simultaneously, actuaries are challenged by stereotypes of being too technical and out of touch with the business, risking that business partners will seek advice elsewhere. The business savvy actuary must work effectively across silos to help solve problems holistically. This session will focus on how actuaries can evolve to meet the changing demands of stakeholders and solve increasingly complex business issues, by focusing on both the art and science behind: Effective Teams – How can we better organize to avoid a silo mentality, drive cross-functional collaboration, and enhance feedback loops? Future Leaders – How can we develop leadership qualities to empower individuals in their careers and foster a culture of innovation? Well-rounded Individuals – How do we better train individuals to embrace ambiguity and amplify their enduring human skills?

Learning Objectives:

• Identify ways to reimagine organization design to enhance cross-functional collaboration and enable high performing teams
• Recognize the critical leadership capabilities and personal factors that drive innovation and empower individuals
• Discover the importance of enduring human skills, such as design thinking and intellectual curiosity, and key mindset shifts to becoming more adaptable individuals

Speaker(s):
Taralyn Slusarski, Manager, Deloitte Consulting
Jennifer Beers, Lead Casualty Pricing Actuary, Ironshore
Christopher Fanslau, Senior VP, Chubb
Eric Blancke, Assistant VP, CNA

1:45 PM - 3:15 PM, ET

Underwriting Insight from Reports Using AI and Machine Learning

Commercial lines insurance underwriters need to rapidly review and synthesize information from a variety of sources at quote and renewal. These sources include loss control and underwriting reports which provide important qualitative information about the risk. We will discuss different ways to support underwriters from using tools from artificial intelligence (AI) to improve process efficiency to decision support with predictive modeling. This will help the audience to identify opportunities and implement solutions to support their underwriters. We will use polling and Q&A to create an active discussion on both use-cases and tools. The audience will also receive a hand-out that summarizes the key concepts covered in the presentation.

Learning Objectives:

• Awareness of underwriting challenges and information sources
• Knowledge of opportunities to support underwriters
• Understanding of artificial intelligence and machine learning technology applications to address the opportunities

Speaker(s):
Vadim Filimonov, Willis Towers Watson
Liam McGrath, Data Scientist, Willis Towers Watson
Jeff Hay, Senior Director, Willis Towers Watson

4:30 PM - 6:00 PM, ET

Data Visualization for the P&C Actuary

The goal of the session is to demonstrate to the audience (P&C actuaries) the benefits of effective data visualization, and how to apply proper visualization principles and techniques in their day-to-day work. This session will include visualization examples to supplement an actuary's own analyses, and visualization(s) and principles for the purpose of communication to others.

Learning Objectives:

• Gain an appreciation of the importance of data visualization techniques
• Learn how to apply these data visualization techniques to everyday actuarial work

Speaker(s):
Jennifer Levine, Allianz Group SE
John Deacon, Executive Director and Actuary, NCCI
Jeff White, Senior Consultant, Gross Consulting