By Charl Cronje, Partner, LCP
As markets soften rapidly in 2026, the importance of technical pricing will increase. Firms need to monitor closely how far they allow rates to fall relative to technical adequacy, and there comes a point where the technical rates need to “bite” and prompt conversations about limiting premium volumes on a particular class of business.
The best firms have detailed “retreat plans” for each class, and for the overall portfolio mix. As one underwriter recently told me “anyone can reduce volume in a soft market, but it’s all about knowing where to reduce that volume.”
It’s especially important in a soft market to capture data on underwriting decisions versus technical prices. This data needs to differentiate between instances where the underwriter accepts the technical price but cannot realistically achieve that rate in the market, versus cases where the underwriter has a different view on the technical price itself.
Over time, this dataset can reveal whether model weaknesses, inconsistent underwriting, market pressures or strategic choices are driving pricing outcomes. This is potentially even more valuable than the underlying policy and claims data for refining pricing models over time. Only a minority of firms are doing an effective job in this area, so it is a potential angle for generating competitive advantage in difficult market conditions.
Pricing team structure and operation
Leading firms have surprisingly different operating models for actuarial pricing. Some pricing functions stand entirely apart from underwriting and provide independent challenge on rate adequacy. Others have actuaries embedded with underwriting teams, providing day-to-day technical support. This approach loses much of the power for independent challenge but firms argue that it helps underwriters make better pricing decisions over time.
Other firms have pricing actuaries working closely with underwriters but reporting into a central actuarial pricing function. This can provide a better combination of independent challenge and business-as-usual assistance. Importantly, it can be helpful for each pricing actuary to have visibility of multiple classes of business. This enables them to spot differences in underwriting approaches and to highlight risks to the central pricing function.
Pricing versus portfolio management
Firms now typically have a portfolio management team (PMT) separate from the pricing and underwriting functions. PMTs tend to include a mix of individuals from underwriting, actuarial and data analytics backgrounds.
The remit of PMTs varies a lot from firm to firm. A common approach is for PMTs to do a lot of work on “thematic” issues like inflation or silent cyber risk, as well as deep dives on problematic classes of business. This is in addition to the core function of helping management decide how to allocate capital or premium volume “budget” between classes.
In other cases, PMTs play different roles, like “translating” between the actuarial pricing function and underwriters, developing new pricing models or helping with new product strategy.
A pitfall for firms to watch out for is creating the sense that the “exciting” pricing work is all going to the PMT, leaving only repetitive daily pricing work to the actuarial pricing team. This can affect performance and retention – rather, the actuarial pricing team and PMT should work closely together to ensure that both benefit from the insights generated by the other.
Capturing AI opportunities
Underwriters are already using large language models (LLMs) for general research purposes. However, there is growing pressure for insurers to deliver AI-related innovations or cost savings in all areas of work.
A key current focus for many firms is using LLMs to automate the ingestion of submission data into the underwriting system. The smallest firms are some way from achieving this, because of lower business volumes (meaning that there is less money to be saved) and lack of resources. The largest firms are tending to pursue this on an “enterprise” scale, which ultimately may be very efficient but may take a couple of years to implement. This may create an opportunity for well-resourced mid-tier firms: they can test focused use cases more quickly than firms pursuing enterprise-wide transformation, while still having sufficient scale for the savings to matter.
An obvious challenge to LLM-assisted data ingestion is “won’t the LLM get things wrong or hallucinate data?” These concerns are valid but, when firms have backtested these solutions against past data, it turns out that humans were making plenty of data ingestion mistakes in the past!
We are still a way off from deep integration of AI into underwriting decision-making itself. The potential gains are clear, but specialty underwriting involves such a high degree of judgement that there will be a high bar for models to add real value.
One way forward is a model that observes underwriter behaviour and identifies decisions that are inconsistent with past practice. This could lead to better decisions over time. Some firms already have simple solutions to help with this, such as automatically showing the underwriter 3-5 examples of similar risks that they have underwritten recently, before they make their decision on the risk at hand.
Where next?
We’ve only scratched the surface in this article. There are many other pressing issues for pricing teams to consider in the coming years, including:
Balancing the benefits of pricing models that mimic underwriter expert judgement with the slightly different benefits of more data-heavy models that provide fully independent challenge.
Balancing the desire for pricing models to provide close to 100% coverage of the portfolio with the reality that models are much weaker and less predictive in some classes of business than in others.
Understanding why some underwriting teams engage much better with the pricing models than others.
Making it easier for underwriters to engage with the technical price, and doing this at the right point in the underwriting workflow.
London Market underwriting is a business that is massively driven by relationships and expert judgement. This makes it a particularly exciting area to apply the actuarial skillset to help manage risk, unlock efficiencies and ultimately maintain competitive advantage. Pricing actuaries will need to continue to innovate in order to achieve this.
|
