General Insurance Spring Conference 2022 Webinar Series

The presentation covers the opportunities, the obstacles, and the optimal way forward for the widespread adoption of AI/ ML across the insurance market, including:

• Ways in which AI/ ML can transform the entire insurance value chain, covering all core activities from marketing & underwriting to claims management
• Current status of AI/ ML adoption across the insurance market - both the systemic trends favouring its adoption, along with the issues that are hindering it (with a particular focus on the need for AI/ ML models to be demonstrably trustworthy)
• Best practice approaches to capture AI/ ML’s full potential. This includes both the organisational and operational capabilities firms require to achieve the full benefits of utilising AI/ ML, while mitigating the associated risks associated with AI/ ML models

The presentation is suitable for all GI Spring Conference attendees and is based on 8 years of academic research (predominantly at Carnegie Mellon University), hands-on experience of building, deploying and governing AI/ ML models (across multiple banks and insurers) and a deep understanding of the evolving regulatory landscape.

Speakers: David Marock, Shameek Kundu

Chair: Ajay Shah, General Insurance Lifelong Learning (GILL) Sub-Committee

WTW worked collaboratively with a large multi-national insurer last year to transform their reserving process. This case study based session would help any insurers looking to do the same with tangible examples of what worked well and what less so. The process is now heavily automated with robust governance and controls throughout, whilst still maintaining flexibility. This is a very new way for the team to be working and we believe a lot of insurers will be moving to this model in the future. It is a real life example of how embedding automation throughout actuarial processes enables the team to work more efficiently, get better insights and add more value across the business. The automated solution removes manual tasks that don't require expert judgement and provides more analytics / insights to facilitate reserving decisions. It also provides better audit, governance and an improved working environment. We utilised lots of systems and tools as part of this project: Power BI, a new Data Warehouse, reserving software, Automation software. This set of tools has completely removed the reliance on Excel - something a lot of insurers are struggling with. We will share what we learnt from using these tools as part of a reserving transformation project.

Speakers: Holly Layton, Ian Thomas

Chair: Richard Williams, General Insurance Lifelong Learning (GILL) Sub-Committee

This session will dive into how Zurich UK has integrated actuaries and data science into its Claims function and the new insights these skills are driving in managing operational performance, driving decision making and providing insights into emerging claims trends for both internal stakeholders and Zurich’s customers. It will be an interactive session challenging the traditionally held views of the role of the Claims function and exploring the additional value added by an actuarial mindset over data scientists. Examples will include forecasting claims volumes and severity to drive resource modelling and insights into emerging experience; claims automation and model oversight; understanding customer and employee satisfaction; inflation modelling and large loss modelling. The session will be presented by Liz Ryan, who following c. 15 years working in risk and capital modelling roles joined the Zurich UK Claims Leadership Team as Head of Performance and Insights. She will share the opportunities and challenges of bringing actuarial skills into a new area and the business value the team have driven over the past 4 years.

Speaker: Liz Ryan

Chair: Rachel Haldane, General Insurance Lifelong Learning (GILL) Sub-Committee

A presentation describing the impact of mortality shape on the reserves of PPOs. We will show how different mortality shapes can lead to very different levels of reserves, even for a given life expectancy. Using sample model points we will show how the initial reserve can vary with shape and how that variation expands as the policy matures. And hence we will show why it is extremely important for holders of PPO risk to have a much better understanding of mortality shape. Plus, an update on the quantitative trends of PPOs as at 2020 year-end, including PPO propensity, possible behavioural trends as a result of Ogden discount rate changes and other factors, as well as an insight into the potential impact of COVID-19 on PPOs.

Speakers: Niel Daniels, Chris Francis

Chair: Cat Drummond, General Insurance Lifelong Learning (GILL) Sub-Committee

A high level analysis of the practical considerations that are preventing the UK GI market from adopting machine learning (ML) techniques in reserving. This builds on the GI ML in Reserving Working Party’s previous survey of the UK market. It highlights the causes behind the lack of adoption of ML techniques in the UK market. Particular focus is given to: The time and resource constraints facing reserving actuaries……and the “Catch-22s” that develop as a result of these constraints.

Speaker: Isabelle Williams

Chair: Amit Parmar, General Insurance Lifelong Learning (GILL) Sub-Committee

With the advent of telematics, pricing and underwriting for certain car insurance companies has evolved into a more complex, data-rich task; a greater focus can be given to driver behaviour compared to other characteristics.

When using behaviour as a rating factor, it is key to present context to decisions that are made. This talk explores the usage of optimal self-driving car algorithms in providing reasonable benchmarks against which such behaviour can be judged.

Speaker: Karthik Tumuluru

Chair: Richard Williams, General Insurance Lifelong Learning (GILL) Sub-Committee

The Capital Research Group, sits under the GI Board, and is a forum to discuss current issues and key challenges in GI Capital Modelling. This session with cover some of the recent discussion points and conclusions, covering topics such as: Dependencies, Capital Allocation, External models, Inflation, Reserve Risk, One-year Risk and Climate Change. We will share the views of the group and will discuss any related thoughts the audience may have.

Speakers: Christian Bird, Jonathan Bilbul, Ajay Chhabra, Mirjam Spies, James Toller (IFoA GI Capital Research Group)

Chair: Matt Facey, General Insurance Lifelong Learning (GILL) Sub-Committee

Inflation risk has been on the watch list of emerging risks for many insurers and regulators over the recent years and the past few months saw real evidence of inflation spike in many countries around the world. Our presentation will cover how insurers can respond to rising inflation risk through: Understanding underlying drivers of inflation risk at portfolio level Establish an enterprise wide process to model claims inflation Assess the full impact of inflation risk on insurer’s financial statements and operations Mitigate inflation risk effectively through multiple dimensions.

Speakers: Martin Cairns

Chair: Arihant Kothari, General Insurance Lifelong Learning (GILL) Sub-Committee

Actuarial work is widely valued by insurers, but existing approaches to our work can hinder us from tackling the questions insurers really want answers to. In this session, we will focus on how analytics can transform actuarial work and empower actuaries to answer questions such as:

• How is the business performing at the level below the current reserving segmentation?
• Which business should we write more/less of?
• What are the key judgements made in setting our reserves, and how can we get earlier business input on these judgements?

We will share lessons learnt from each stage of the analytics journey: from aligning business culture, to proof of concept, through to embedding the solution. We will consider each stage from the perspectives of the business and the analytics team.

From aligning business culture to proof of concept:

• Business: What’s in it for the business? What are the criteria for success?
• Analytics: Starting simple and identifying limitations as the motivation (or not) for increasing complexity.

Embedding the solution:

• Analytics: Behavioural and process differences between working on the proof of concept and the end solution.
• Business: Identifying changes in behaviour and training required for the solution to be used effectively.

Speakers: Charlie Stone , Ajay Shah, Tom Durkin

Chair: Arihant Kothari, General Insurance Lifelong Learning (GILL) Sub-Committee

In this session Dorian Hicks (Partner at Mazars), James Orr (Senior Actuary at the PRA) and Meirion Board (responsible for ESG and financial risks at QBE Insurance) will cover:

• High level background on climate change and the impacts physical, transitional and litigation risks could have on insurers.
• Regulatory landscape including PRA Supervisory Statement and Dear CEO letter, TCFD disclosures and Climate Biennial Exploratory Scenario (“CBES”).
• Practical examples on putting the regulatory requirements into practice.

Speakers: Dorian Hicks, James Orr, Meirion Board

We will present a framework for bridging the gap between the insights of climate change experts and the specific strategic and risk management needs of insurers. Our experience is that climate change modelling experts sometimes think very differently from insurers or actuaries and there is much to be learned from their approach. This session will show you, as an insurance actuary, how to get the best out of the wide range of climate change modelling expertise in the scientific community at large. This will help you provide valuable input to discussions on strategy, asset and liability modelling and risk management in your own firm.

Speakers: Russ Bowdrey, Jessica Clark, Lara Palmer

Chair: Ajay Shah, General Insurance Lifelong Learning (GILL) Sub-Committee

Our talk will take the audience on a journey of a large-scale actuarial transformation project which has successfully delivered an advanced actuarial modelling platform for the business, based on a new target actuarial operating model and a modern cloud technology stack. The platform orchestrates an automated end-to-end process, covering remote data queries and ingestion, data transformation, modelling (including best estimate and uncertainty reserving, AoC’s and AvE’s, IFRS17, and Solvency II, with links to the capital model), diagnostics, and reporting. This has brought significant and long-lasting benefits to the actuarial function and wider finance function within the organisation. We’ll highlight the benefits that were achieved in each key stage of the project, covering aspects such as the requirement setting, model building, UAT, and implementation. In addition, we’ll describe the forward thinking actuarial operating model, including the impact on broader issues such as cultural change and team composition, and the positive impact that this has had on the resilience, speed, and consistency of information from the actuarial function. Our goal is to show that large scale transformation projects can be undertaken successfully in organisations, effecting substantial and beneficial change, and to embolden senior actuaries to embark on the process fearlessly.

Speakers: Adrian Ericsson, Cormac Bradley

Chair: Rachel Haldane, General Insurance Lifelong Learning (GILL) Sub-Committee

Dealing with outwards reinsurance is one of the most challenging aspects of IFRS 17. In this session, we will explore some of the issues that insurers are struggling with and provide practical suggestions for how to overcome them.

Speakers: Simon Sheaf, Bhavin Bhatt, Adrian Gilder

Chair: Matt Facey, General Insurance Lifelong Learning (GILL) Sub-Committee

The talk will be from the perspective of two actuaries who have incorporated explicit inflation elements into their reserving. One operates in the inflation indexed world of Israeli insurance. The other has worked explicitly with inflation in both in the simple link ratio world and also with econometric stochastic models. The will speak of how they are currently dealing with the recent Inflation spikes and the nuanced issues that come up when doing this. They will also discuss what additionally can be done, based on mixture of actual experience and research findings of the Sensing and Responding to Change and Nonsense working party. Many of the non Israeli ideas can be found in this recent linked-in article.

They will bring out key ancillary issues of economic, social and business driven changes which complicate making inflation adjustments.

Speakers: Chris Smerald, Yaakov Moser

Chair: Cat Drummond, General Insurance Lifelong Learning (GILL) Sub-Committee