Scaling Laws, Tabular Data and Actuarial Ratemaking Models

You did not meet the minimum attendance requirements for this event.

Thank you

This live web event has ended. Thank you for attending.

Description

PLEASE NOTE: REGISTRATION WILL CLOSE 10 AM ET THE DAY OF THE WEBINAR. DUE TO HIGH VOLUME OF DEMAND, WE ARE ONLY ABLE TO GUARANTEE A SEAT TO THOSE WHO REGISTER 48 HOURS AHEAD OF THE WEBINAR TIME.

This webinar investigates whether the neural scaling laws driving Large Language Model advancements apply to actuarial ratemaking. Utilizing a 4.5-million-record motor insurance dataset, we demonstrate that power-law performance improvements do exist for tabular data, though they are highly dependent on model architecture.

Speakers will present empirical evidence showing that while standard Transformers exhibit weak parameter scaling, "TabM" architectures and Transformers augmented with self-supervised learning achieve superior results in large-data regimes. The presentation covers the identification of distinct data regimes—from GLM-efficient small datasets to deep learning-dominated large datasets—and introduces novel architectural enhancements like multi-CLS pooling and TokenMoE.

Learning Objectives:

  1. Quantify the relationship between data volume, compute resources, and model performance in actuarial ratemaking using empirical power-law scaling relationships.
  2. Identify the optimal modeling strategy for distinct data regimes, distinguishing when traditional GLMs suffice and when advanced architectures like TabM or TokenMoE are required to maximize predictive accuracy.
  3. Apply specific architectural enhancements - such as multi-CLS pooling and swap-style self-supervision - that enable Transformer models to overcome weak parameter scaling on noisy tabular datasets.
Registration Information and Fees

PLEASE NOTE: REGISTRATION WILL CLOSE 10 AM ET THE DAY OF THE WEBINAR

Registration Fees (in U.S. Dollars)
Individual$60
Multiple Connections**
(Unlimited internet connections for individuals working
for the same company. Please note that audio for this
presentation will be streamed via the web)
$820

*Multiple Connection Registrations should contact Mallory Peebles directly at [email protected]. The registering party for the Multiple Connection Registration will be responsible for distributing all event details to attending individuals within their company.

Speaker Opinions
The opinions expressed by speakers at this event are their own and do not necessarily reflect the opinions of the CAS.

No Broadcast Policy
Your registration entitles you to access CAS’ webinar programs for which you have registered. By registering to attend the webinars, you agree not to sell, transfer, or otherwise share your webinar access link and/or code unless such sharing is authorized by CAS. You further agree not to record or broadcast CAS’ webinar programs without advance written authorization by CAS.

Cancellations/Refunds
Registrations fees will be refunded for cancellations received by September 3, 2026 less a $25 processing fee.

CAS Continuing Education Policy
The CAS Continuing Education Policy applies to all ACAS and FCAS members who provide Actuarial Services. Actuarial Services are defined in the CAS Code of Professional Conduct as “professional services provided to a Principal by an individual acting in the capacity of an actuary. Such services include the rendering of advice, recommendations, findings or opinions based upon actuarial considerations.” Members who are or could be subject to the continuing education requirements of a national actuarial organization can meet the requirements of the CAS Continuing Education Policy by satisfying the continuing education requirements established by a national actuarial organization recognized by the Policy. For further information regarding the CAS Continuing Education Policy please visit the CAS web site.

CAS Webinars may qualify for up to 1.8* CE Credits for CAS members.
Participants should claim credit commensurate with the extent of their participation in the activity. CAS members following the USQS earn 1 CE Credit per 50 minutes of educational session time, not to include breaks and/or lunch. *The amount of CE credit that can be earned for participating in this activity must be assessed by the individual attendee. It also may be different for individuals who are subject to the requirements of organizations other than the Casualty Actuarial Society.

Contributors

  • Ron Richman

    Ron Richman is the Founder of insureAI, a startup focussing on building machine and deep learning tools to improve actuarial processes. Before this, he was Chief Actuary at Old Mutual Insure, South Africa’s second largest P&C group, where he was responsible for oversight of all P&C actuarial activities across Old Mutual. Prior to this, Ron was Chief Actuary and Chief Risk Officer for AIG’s African entities. Ron is a Fellow of the Institute and Faculty of Actuaries (IFoA) and the Actuarial Society of South Africa (ASSA), holds practicing certificates in Short Term Insurance and Life Insurance from ASSA, a Masters of Philosophy in Actuarial Science, with distinction, from the University of Cape Town and a PhD in applying deep learning in actuarial science at the University of the Witwatersrand.

  • Morgan Bugbee

    Morgan Bugbee is Director of Data Science at Quanata, a State Farm company focused on advancing context-based insurance through risk prediction, telematics, and AI-assisted modeling. He brings over two decades of actuarial and insurance analytics experience across pricing, underwriting, risk strategy, and predictive modeling, with prior leadership roles at Farmers Insurance and Cover Financial.

    Morgan is a Fellow of the Casualty Actuarial Society and a Certified Specialist in Predictive Analytics. He is a longtime contributor to CAS research initiatives and currently serves as Vice President for Research & Practice Advancement for the CAS. He holds degrees in mathematics and music performance from Northwestern University.

    Morgan’s work sits at the intersection of actuarial science, data science, and insurance innovation, focusing on translating advanced analytics into practical business and risk-management decisions.

September 10, 2026
Thu 12:00 PM EDT

Duration 1H 30M

This live web event has ended.

For Content or Purchase Questions
9am - 5pm ET**