Soft market insights from BW’s Risk & Capital Roundtable

As rates soften across multiple classes, capital and risk teams are becoming increasingly involved in management decision-making and often asked to assess and quantify future risks. In our December roundtable, senior professionals from leading insurers and Lloyd’s syndicates discussed what the current cycle means for pricing discipline, capital efficiency, modelling choices, and validation practices, and how teams can respond with confidence. This article distils the most actionable takeaways.

1. The softening market is back, and it’s not uniform

Rate adequacy is a common theme. While some classes have been softening for a while, others remain relatively profitable for now. The distinction by class matters: several attendees questioned whether property business ever truly hardened at all for certain insurers, with some already seeing soft conditions and deterioration concerns.

Despite rate pressure, many books remain profitable thanks to higher baseline rates (structural risk premiums) built up since 2019–2023. As margins shrink, senior management attention is shifting from income growth to return on risk-adjusted capital (RORAC) by class, and, crucially, to capital efficiency. Insurers worry about a 'double impact' of less profit and higher capital requirements hitting balance sheets.

Participants who had experienced past cycles noted that this one feels quicker and different from the cycles of the 2010s. In the 2016–2017 soft market, many classes were projected to lose money on a business plan basis, pushing a focus on break-even and income rather than RORAC. But if softening accelerates, the implication is that RORAC may again be crowded out by near-term income priorities unless leadership holds the line on discipline.

2. Regulation and market oversight: stability vs. cyclicality

Solvency II has arguably reduced market exits that previously tightened capacity during soft phases, contributing to today’s persistent competition; managing general agents (MGAs) are not disappearing, capital remains abundant and new syndicates continue to launch, which are all factors that can extend soft conditions.

At Lloyd’s, the Decile 10 initiative (launched around 2018) improved performance by forcing remediation or withdrawal of the lowest-performing lines of business. As capacity expands again, the Decile 10 exercise stands as a reminder: profitability can fall fast when discipline slips, and internal and external oversight may need to re-engage as rates flatten. Attendees also observed Lloyd’s increased scrutiny of business plans and underwriting facilities this year, aimed at preventing growth for growth’s sake and guarding against the looser controls that are typical of soft markets.

3. Capital modelling and validation: hold your nerve, sharpen your evidence

A key perspective: a softening market should not automatically imply higher risk. If internal data across cycles is sufficient, capital teams can test whether prior market phases materially altered loss ratios and volatility, to avoid automatic parameter increases. However, if average profits are expected to decrease, then downside risk will increase even if volatility stays flat as the overall distribution shifts towards lower profit. Loads for parameter uncertainty and events not in data (ENIDs) already sit on top of purely data-driven coefficients of variation (CoVs); adding more loads “because the market is softening” could be an overreaction that then needs to be reduced in future. Some cited the pandemic as a cautionary tale, with material capital loadings added quickly and unwound later.

However, several attendees cautioned: for casualty-driven capital, deteriorations take time to emerge in data, and near misses never appear in parameterisation datasets. That means scenario testing and back-testing must be enhanced with longer histories and external benchmarks to challenge optimistic ultimate loss ratios (ULRs), rate-change assumptions, and plan loss ratios, especially in classes prone to drift outside appetite.

Risk appetite governance generally becomes easier to enforce in a softening market. As margins shrink, buffers disappear and management becomes more supportive of risk teams maintaining limits. Clear board-approved appetites provide a strong defence against broker/underwriter pressure. These limits and appetites should also be reviewed in the context of the market – limits may have been increased in the hard market due to pressure to take advantage of commercial opportunities, and so they may need to be adjusted to reflect the new market reality. Regulators, too, increase scrutiny of business plan realism and appetite evidence during soft phases.

4. Dependencies and model architecture: make cycle effects explainable

Adjusting dependency structures for cycle effects is challenging, especially with copulas, where parameter adjustments may be hard to explain to boards. Teams using driver-based structures felt more confident, noting they can explicitly include an underwriting-cycle driver with historical evidence at class level. Being able to frame the fact that correlations have changed due to a documented risk driver is often more intuitive to senior stakeholders.

5. Validation testing in a softening market: evolve scenarios, separate signal from noise

When carrying out internal model validation in a softening market, risk teams should expect reverse stress testing (RST) scenarios to become more severe to reflect pricing pressure and thinner margins. Communicating these changes via return periods and scenario narratives can help leadership grasp volatility shifts by class. In the detailed analysis of change, it is important to carefully distinguish:

• Movements from volume/data changes (e.g. smaller adverse claims having bigger proportional effects).
• Parameterisation changes (e.g. discipline slippage, broader coverage, lower deductibles, higher limits, riskier accounts to maintain premium, and reserve uncertainty from delayed strengthening).

Reserve risk CoV reviews should be triaged: not every class needs a deep dive every cycle, but the more volatile lines (liability, financial lines, specialty) merit more frequent scrutiny. Attendees also discussed volume-adjustment formulas and market-level curve refits; practices vary and transparency around the chosen methodology and any specific expert judgements and limitations remain important.

6. Emerging risk landscape: non-peak perils, vendor model limitations, ESG gaps

Recent analysis shows non-peak perils now dominate aggregate losses over peak perils on a five-year running average. At mid-2025, non-peak perils stood around $76bn vs. $37bn for peak perils (inflation-adjusted), with non-peak steadily rising over the past decade while peak perils remained relatively flat. This trend challenges historical parameter settings for classes long considered 'secondary'. Validation should stress-test whether current distributions and scenarios remain adequate.

Teams warned of over-reliance on vendor models for emerging perils. Sophisticated projections can hide weak assumptions, especially where historical claims data is thin or dynamics differ from model logic. Examples included flood models misreading basement dynamics and wildfire models anchored in outdated land/building data. Many capital teams layer additional loadings (e.g. climate) on external model outputs, and these loadings can come under pressure in soft markets. Documenting the rationale and uncertainty clearly is essential.

On economic scenario generators (ESGs), several participants reported persistent back-testing failures and insufficient responsiveness to recent geopolitical shocks and inflation regimes. Methodology changes and widening credit spreads have produced volatility in capital numbers for some, forcing stabilising self-loadings. This underscores the need for robust governance around ESG integration and transparency on how dependency structures link to ESG outputs.

What BW recommends now

Based on the discussion, here are our recommendations for pragmatic steps that capital and risk leaders can take immediately:

1. Resegment your portfolio by cycle sensitivity. Identify classes most at risk of discipline drift and RORAC compression. Align monitoring, RST severity, and governance accordingly.
2. Strengthen evidence, not loadings. Extend back-testing windows, add external benchmarks where credible, and document parameter choices. Only adjust capital where data and scenarios clearly warrant change.
3. Make dependencies explainable. If possible, incorporate cycle-aware risk drivers so changes in correlations are easier to communicate and defend at board.
4. Challenge external vendors for their model assumptions. Record uncertainty, justify loadings transparently, and avoid single-model dependence on under-researched perils.

Soft markets compress margins, expose weak assumptions, and test governance. The good news: with enhanced evidence, clearer appetites, explainable dependencies, and disciplined use of technology, capital and risk teams can steer through the cycle without overreacting, or even under-reserving. BW’s Capital & Risk team partners with insurers and syndicates to refine parameterisation, strengthen validation toolkits, and build cycle-aware models and governance that stand up to board and regulatory scrutiny.

Want to explore what this means for your company?

Get in touch with BW’s General Insurance team to discuss bespoke diagnostics, scenario libraries, and cycle-aware dependency frameworks tailored to your classes and appetite.