Key issues impacting the Lloyd's market: insight from BW's roundtable

Our General Insurance (GI) team recently held another roundtable event, bringing together experts in both first line capital modelling and second line validation to discuss their views on a range of topics affecting the Lloyd’s market. 

The topics for discussion included recent loss trends and their impact on profitability, climate change, casualty catastrophes and the value of using economic scenario generators (ESG).

Profitability

Has loss experience in the GI market been relatively benign, or are we becoming accustomed to an increasingly volatile world?

Underwriting profitability has become a hot topic. Whilst a welcome phenomenon, its impact on prospective business planning is significant. For instance, do the most recent underwriting years inform those to come, or how should an insurer justify its latest loss ratio assumptions? 

We appreciated hearing about how teams leveraged cooperation between actuarial, finance, and other business areas in their approaches to this year’s syndicate business forecast (SBF) submission, and the differences underpinning those who are under pressure to maintain capital from external providers.

What is fascinating is the unpredictable nature of insurance profitability. At the start of September 2024, the year could have been considered somewhat benign due to hard market rates, which had seemingly provided a sufficient buffer against loss events such as the CrowdStrike-related IT outages, the Baltimore Bridge collapse, Hurricane Beryl and Debby, and other severe thunderstorm losses in the first half of 2024. Move forward six weeks and the impact from Tropical Storm Helene and Hurricane Milton may have changed the discussion.

Major insurance and reinsurance industry loss events since 2011
 

Source: Reinsurance News
 

How do we balance Lloyd’s focus on loss ratio adequacy with a parameterisation approach suitable for full model use?

The efficacy of the Lloyd’s Capital Returns (LCR) loss ratio testing approach was considered, specifically for catastrophe and casualty-exposed classes. Questions around credibility arose within any given five-year period as well as uncertainty from loss development on longer-tailed classes. Further notes were made around segmentation, specifically where inconsistencies exist between internal and external stakeholders. An overarching theme noted during the discussion was the conflicting needs surrounding model use relative to regulatory requirements.

Insurance firms generally function through diversification - a poorly performing class or division is offset elsewhere by a well-performing class. Given this holistic view of the function of an insurance company, loss ratio testing at a very granular level may lead to inaccurate conclusions about the health of the whole underwriting function. It is expected that some classes will underperform and others overperform to plan. Attendees preferred testing at a higher level, such as divisional or even at an overall syndicate level. It is felt that the market is generally comfortable with testing that includes granular class-level fails with justification, given a pass at a higher level of aggregation.

Climate change 

Even with support from boards, underwriters, and pricing actuaries for a climate change allowance, how do we incorporate this into our models?

Recent updates to external natural catastrophe models materially increased losses in the tail of capital models, reflecting increased uncertainty driven by climate change with sometimes conflicting views on realism. For example:

• The catastrophe models feed directly into pricing and into risk appetite limits. These downstream impacts of changing models are some of the more explicit impacts of climate change parameterisation impacting multiple areas of the business.
• Modelled loss curves are changing shape, rather than just scaling up and down. The changes are driven by shifts in both frequency and severity. Balancing expectations around the mean and the 1 in 200 can lead to unexpected movements across the curve. 

Regulatory requirements are sometimes felt incompatible with model use, especially where best estimate and sufficient prudence come head-to-head. Despite the prudence believed to exist in the internal models, it is largely believed that the potential financial impact of climate change continues to be underrepresented. Further, it is unclear whether market allowances for prudence are adequately considering the change in shape of the risk, when considering its potential future impact on both the frequency and severity.

Considering model use in pricing, we discussed a potential first-mover penalty i.e. there is no commercial advantage in fully recognising the potential cost of climate change on certain market perils as it could price the insurer out of the market. Commercial and financial forces seem to be acting in equilibrium, such that when no one wants to move first, no one will move at all. Whether there is a potential for this to lead to a systemic pricing risk, could be theorised as having already materialised. 

Casualty catastrophes

If insurers price in the true theoretical risks, would consumers still be able to afford this? Or will parts of the industry become increasingly uninsurable?

Whether a risk is insurable or otherwise depends on many factors. Even a very volatile risk with the right limits, controls, and buy-in from all business areas may still be written at an acceptable price. Of course, once these volatile types of risks are written, it becomes imperative to monitor aggregations.

Commercial concerns reigned supreme for our attendees. Set the ‘true’ rates, and insurers may price themselves out the market. Balancing commercial pressures against the data and models is paramount. Social and judicial inflation, endemic in US casualty classes, are often hard to fully capture. Slow claims development can sometimes allow the can to be kicked down the road in the name of commercial concerns. 

Economic scenario generators (ESG)

Whilst it is largely accepted that economic scenario generators represent current best practice, do the vanilla asset portfolios underpinning the Lloyd’s syndicate market diminish their value?

It was interesting to hear from those who have recently made the move from internal bespoke modelling of economic indicators towards the use of an economic scenario generator. They noted that the additional scrutiny and prudence required began to outweigh the benefit. Others noted that they are considering the plausibility of moving away from the scenario generators, towards their own bespoke derivation. We also heard from those who have integrated the parameterisation of economic indicators with their respective investment teams and investment managers, each highlighting that such integration improved the quality of investment risk modelling.

The group discussed the limited number of providers of economic scenario generators in use in the Lloyd’s market. The discussion considered whether Lloyd’s should consider centralising the validation of the main providers, alleviating this work to some extent for the market participants. A common counterpoint was that each syndicate would need to consider its own risk profile. Whilst a wholly centralised exercise would not be plausible, key characteristics which are consistent across the market could be considered for a centralised review. 

We would like to thank all of those who participated in the discussion for sharing invaluable insights across the market, and we look forward our next roundtable.

If you are interested in taking part in future capital modelling and validation roundtable events, please contact Julien to express your interest.