Insurance Capital Markets Research (ICMR), an independent analytics and research firm founded by former leaders of Lloyd’s internal research and performance team, has released new findings showing how relative benchmarking, enhanced by predictive AI, can transform the way underwriting performance is assessed in the Lloyd’s market.
ICMR explains that the Lloyd’s market, with its complex structure and sensitivity to market cycles, has long been difficult to evaluate using traditional absolute metrics.
Performance figures that appear strong in soft conditions may mask weaknesses, while the same methods can unfairly penalise well-run syndicates during tougher phases of the cycle.
This, ICMR notes, makes it challenging to distinguish genuine outperformers from those simply moving with wider market trends.
In contrast, ICMR’s research shows that relative benchmarking offers a more reliable measure. By comparing syndicates directly against their peers, ICMR demonstrates that relative performance is both more stable and more revealing of true underwriting quality.
Their analysis of historical data indicates that syndicates tend to remain in the same performance quintile over time, highlighting that relative standing reflects lasting characteristics rather than short-term fluctuations.
ICMR emphasises that this approach has become viable only in recent years, thanks to greater access to detailed performance data and advances in probabilistic programming.
Drawing on its proprietary dataset of gross underwriting results, ICMR has developed an AI-powered Bayesian model, inspired by established ranking frameworks such as the Plackett-Luce model.
This model predicts the relative loss ratio performance of Lloyd’s syndicates, offering forward-looking insights into profitability and stability across classes of business.
The firm explains that these predictions can be visualised in ways that highlight which syndicate classes combine strong outlooks with consistency and which are more exposed to weaker or unstable performance. When layered with broader market forecasts, the result is a complete picture of potential outcomes that supports more strategic underwriting and portfolio decisions.
According to ICMR, clients adopting this approach see tangible benefits. Portfolios designed using relative benchmarking and predictive AI exhibit higher expected returns and reduced volatility compared with relevant benchmarks.
In one recent collaboration, documented in a joint white paper, ICMR’s analysis enabled a syndicate’s management team to strengthen portfolio quality year after year, producing measurable improvements in results.
ICMR concludes that relying solely on absolute measures is no longer adequate for the complexities of the Lloyd’s market.
By focusing on relative benchmarking and applying predictive AI, syndicates can more reliably identify true outperformers and build portfolios that deliver superior long-term performance.
About the Author
