This blog series is a practical guide on solvent exit planning for practitioners and those charged with governance. Part 4 explores assumptions used in solvent exit planning, in the lead-up to and during the exit, as well as models used.
In the period leading up to a potential solvent exit, the assumptions used in the projection should reflect the conditions of the scenario that ultimately drives the exit trigger(s). These assumptions may differ materially from business as usual modelling and should be considered and calibrated to capture the pressures contributing to the deterioration in business performance.
Lapse behaviour may change ahead of a solvent exit. For example, where the scenario includes reputational strain, market rumours, or broader economic stress, lapse assumptions may need to reflect increased policyholder sensitivity to perceived financial weakness. Distributor behaviour may also diverge from normal patterns, with advisers or brokers moving business away from the firm if they believe the company is facing challenges. These dynamics described can amplify anti selection.
Claim experience may worsen in the run up to the exit and may be a direct driver of the exit trigger. Potential deterioration could include increases in morbidity, more volatile claim incidence, or behaviourally driven increases in claims. Anti selective claims may emerge alongside adverse lapse movements, strengthening the interaction between these two assumptions.
For general insurance business, similar considerations would need to be made around claims frequency and severity, changes in reporting patters and policyholder behaviour.
Expenses may escalate before the exit, particularly if the scenario reflects operational strain, inflationary pressure, or reductions in scale that diminish cost efficiency. Contracted supplier and third party costs such as administration providers, reinsurers, or service partners may increase if counterparties perceive the firm to be at a heightened risk. These rising expenses can contribute directly to liquidity pressure and may be one of the factors leading to the exit trigger.
During the projection period itself, the assumptions may continue to evolve in a manner consistent with the stresses experienced before the exit. The same drivers seen in the lead up may persist or intensify, and the modelling framework should reflect the fact that behavioural, operational, and financial pressures often escalate after the formal announcement.
Lapse behaviour may become more extreme once the exit has been announced. Policyholders may react adversely to heightened uncertainty or a perceived weakening of the firm’s future support capability. Advisers and brokers may actively migrate policies away from the insurer, accelerating the outflow of healthier risks and deepening anti selection. Assumptions may therefore require higher or more prolonged lapse stresses than those used in the pre exit period.
The firm may choose not to offer renewals as part of the exit strategy. However, in some cases renewal may represent a contractual obligation (perhaps for some long-term business), which would constrain the extent to which lapses can be assumed to increase. This element should also be considered when setting the lapse assumption for the solvent exit analysis.
Claims experience may worsen further during the projection period, particularly if policyholders are concerned about the firm and its ability to pay future claims. This may manifest as anti-selection, resulting in increased incidence, and being operationally constrained may result in longer claim durations. The interaction between anti selective lapses and deteriorating morbidity may reinforce the need for more severe claim assumptions post announcement.
Per policy expenses are likely to rise as the book contracts and economies of scale are lost. Overhead absorption may weaken materially as lapses increase and the volume of in force business declines. Operational inflation and service provider repricing may further elevate costs, particularly if third party administrators or reinsurers adjust fees due to the changing risk profile. Additional project, governance, and oversight costs associated with managing the exit process will also add to the expense burden throughout the projection, for example increased legal costs and/or additional short-term contractor needs.
On the other hand, some expenses may reduce through the period such as reduction in staff salaries and the exiting of supplier contracts. The phasing of reducing costs would need to be mapped out through the run-off period.
Liquidity requirements often increase during the run off period to meet claim payments and operating expenses, which may necessitate a more conservative investment strategy. Reduced scale and increased volatility may limit investment flexibility, potentially lowering assumed investment returns and increasing capital strain.
On the other hand, investing in lower risk assets would reduce investment related SCR, potentially releasing capital. Therefore, there will be an overall trade-off between lower expected returns and a slight boost in liquidity through the release of SCR capital.
Reinsurance arrangements may become less efficient as volumes fall. Minimum premium commitments or expense loadings may become disproportionately large relative to the shrinking portfolio, and the firm may lose cost effectiveness in its treaties. There may also be a risk of termination of reinsurance treaties. These changes may need to be reflected explicitly in the projection assumptions.
Once the exit has commenced, management actions typically become less effective and may already be exhausted. Actions (such as pricing levers, underwriting adjustments, and retention activities) may no longer be realistic, and assumptions should reflect reduced ability to mitigate adverse experience.
Additionally, regulatory expectations may tighten during the projection, increasing reporting, governance, or supervisory requirements. These interventions may raise operational costs or restrict flexibility and may be considered in the modelling where material.
Overall, a thorough analysis of which actions can be taken and at which point (pre or post exit) of the solvent exit. For example, some actions may be taken as an attempt to recover the business, and then post-exit decision can be fully taken – such as a change in investment strategy.
Pros
Using existing models for solvent exit run-off analysis is typically cheaper and less time-consuming than developing a new modelling framework. These models already have established infrastructure, data feeds and calculation logic, meaning the analysis does not start from scratch and can be progressed more quickly within constrained timelines.
Existing models may already contain calibrated stress and scenario frameworks that can be adapted for solvent exit purposes. Therefore, existing models can be an efficient way to get to the point of non-viability. Their behaviour, sensitivities, and limitations are generally well understood by actuarial teams, reducing uncertainty around how results will respond under adverse conditions. This familiarity can also improve the speed and quality of iteration when refining assumptions or scenarios.
From a governance perspective, existing models often have established credibility with the Board, regulators, auditors, and assurance providers. They benefit from proven controls, documentation, and audit trails, which can reduce the level of challenge compared to introducing a new model. As a result, they are typically easier to progress through internal governance and approval processes.
Existing models are also commonly integrated into wider the risk management process. This supports consistency between solvent exit analysis and other internal metrics, making outputs easier to interpret for senior stakeholders. Boards and committees may be more comfortable engaging with results from a familiar framework, improving the effectiveness of oversight and decision-making.
Cons
Existing models may not be sufficiently tailored to the specific requirements of a solvent-exit run-off. They are often designed for business-as-usual use and may embed assumptions or structures that do not adequately reflect the dynamics of a solvent exit scenario, for example a stable lapse or expense assumption where in reality this would likely be non-linear in a solvent exit. As a result, the model outputs may not fully capture the severity or interaction of stresses that drive the firm towards exit.
There is also a risk that the solvent exit scenario is constrained by what the existing model can realistically deliver, rather than what is economically plausible. There may therefore be a risk of mis understating key risks or obscuring the true boundary for viability.
Finally, performance and run-time constraints can be a material drawback. Existing capital or reporting models may be slow or resource-intensive when projecting long run-off periods or running multiple scenarios and may not have the capability to project the full run-off. Additional spreadsheets may also need to be created to deal with solvent exit scenario specifics such as many increased one-off costs such as redundancy (which existing models may not currently have). This can limit the ability to iterate efficiently which may increase time spent and costs associated with preparing the solvent exit run-off projection.
Pros:
Creating a new model specifically for solvent-exit run-off analysis allows the modelling framework to be built specifically to the regulation. The model and outputs can be tailored to the solvent-exit trigger and assessing run-off viability, rather than adapting a business-as-usual framework for this regulation. This may be a simpler approach than adapting an existing model.
A new model can be designed to reflect the current shape of the business and the latest understanding of risks, products, and operating structures. This avoids reliance on legacy assumptions or outdated design features that may persist within existing models and ensures greater alignment with the firm’s present risk profile.
Purpose-built models also allow for explicit recognition that many business-as-usual management actions cease to be available once a solvent exit is initiated. Assumptions can be structured to reflect reduced flexibility in pricing, underwriting, retention activity, and cost control, avoiding the risk that inappropriate business-as-usual mitigations remain embedded in the projections.
Finally, developing a new model reduces the risk of inappropriate reliance on a framework outside its original design intent. A new model can provide clearer specificity for solvent–exit analysis and discussion with senior stakeholders.
Cons:
Developing a new model for solvent-exit run-off analysis can be significantly more time-consuming and costly than adapting an existing model and furthermore, the build phase may take longer than expected.
A new model is also likely to face a longer and more intensive governance process. Heightened scrutiny may be applied to the review and sign-off of the model due to this model not being seen by management before. This is compounded by the fact that newly developed models may have less established credibility, making stakeholders more cautious about reliance on the results.
There is an increased risk of modelling errors in the early stages of development. The model will be less well tested than existing models, increasing the likelihood of issues that require rework. This can further extend timelines and increase delivery risk.
Resource constraints may also be a limiting factor. The organisation may not have sufficient in-house expertise to build, validate and maintain a new model, particularly if the insurer is relatively small. In such cases, creating a new model may not be proportionate to the scale or complexity of the business and may divert resources from other critical activities. Proportionality is a point which is very important in implementing the regulation.
In addition, a new model may produce outputs that are not fully aligned with existing business-as-usual capital, ORSA, or reporting metrics. This can create additional work to reconcile differences and explain inconsistencies to senior stakeholders. The documentation burden is also materially higher, with full documentation, controls and maintenance processes needing to be developed from scratch.
