Changing internal models in a controlled environment

Capital models have advanced significantly since the early 2000s, but the increased complexity can slow decision-making and raise operational risks. This article explores when a model may no longer be fit for purpose, the role of off-the-shelf and external models, and how to transition safely using strong governance, clear analysis of change and external reviewers.

Evolution of capital models

Over the past two decades, technological advances have resulted in a drastic change in the modelling capabilities of insurers. Capital models have played a key part in the insurance industry since the early 2000s with the introduction of the Individual Capital Adequacy Standards (ICAS) framework, which steered the industry towards more risk-based practices. Then came the move to Solvency II, which officially went live in 2016. This led to a major regulatory overhaul across Europe that raised the bar for how insurers assess their financial health. Since then, global market shifts and political tensions have made modelling even more complex. Today, insurers face tougher rules and higher expectations, leading to more advanced models but also more oversight than ever before.

Is your capital model still fit for purpose?

Insurance models are getting more complicated and it is starting to show. Every time regulations change, new products are launched or markets shift, insurers update their models to keep up. But with each update, the models grow more complex. That means more data, more moving parts and more chances for things to go wrong. The added complexity also results in more time being spent to parameterise the models, making it harder for insurers to remain nimble with their business strategy. In many cases, these models are struggling to keep pace with the fast-changing insurance world and may no longer be fit for purpose. The following sections cover the areas that should be considered when assessing whether your capital model is still fit for purpose.

Benefits of off-the-shelf and external models

"More insurance companies are moving towards off-the-shelf models due to their ease of implementation and accessibility."

These models are straightforward for teams to operate and their transparency makes them easier for reviewers, regulators and Boards to understand. That clarity helps speed up approvals and consequently, the decision-making process. Simplified models also work well with modern technology platforms like cloud services. They run faster, are more flexible and give insurers a competitive edge by helping them fine-tune how they assess risk and set prices.

Interestingly, simpler doesn’t always mean less accurate. In many cases, these models perform just as well or even better, depending on the situation. These models use fewer expert judgements which reduces the risk of spurious accuracy. They also give key stakeholders better insight into which parameters are driving the results of the model. 

As market conditions shift and regulations tighten, insurers are expanding into new areas of business. External models, like Economic Scenario Generators (ESGs) and catastrophe models, provide valuable outputs on key areas of the business that insurers may not have direct access to such as forecasts of key economic measures, and monitoring global weather patterns (including catastrophic events such as floods and earthquakes) respectively. By using outputs from external models, insurers can respond and make decisions that are  informed and forward-looking.

However, it is important to ensure that external models are still independently validated and deemed appropriate for use. A model that aims to be suitable for the average market participant may not reflect the specific features of an individual insurer without adjustments and these will need to be justified. Furthermore, external models may not be updated quickly enough to reflect changing market conditions (e.g. inflation volatility seen in recent years), external data or the latest research as they may value consistency over time to provide a stable view to the market. Any external model update cycles will need to be managed to fit in with the insurers’ timelines and any limitations of the exercise should be clearly communicated to key stakeholders within the business. This will enable the external models to go through the appropriate governance while still reacting to both internal and external changes and remaining up to date.

Considerations when changing to a new model

The first step is to ask whether the new model is truly a better fit. It should solve real problems with the current setup and offer clear advantages. Future-proofing is a key consideration as this provides the insurer with flexibility to adapt to the fast-changing environment that we are currently experiencing. Other operational / strategic factors that should be factored in are model run times, in-house expertise available and future model uses.

Once those are clear, the selection process begins, generally through a full Request for Proposal (RfP). This involves inviting multiple vendors to submit proposals detailing how their solution meets business needs, including timelines, costs and implementation plans. While cost is always a factor, long-term value, strategic alignment and future flexibility are just as important. A cost-benefit analysis helps confirm that the expected improvements justify the time, effort and investment required to move to the new model.

Once models have been shortlisted, then comes technical validation. In this step, the capital modelling teams stress-test the model, run simulations and check the outputs to make sure they are reliable and robust. Finally, the selected new model needs buy-in from key stakeholders, particularly the Board and regulators. Their approvals ensure that the new model supports the financial stability of the company and continues to protect policyholders. 

Model transitions in a controlled environment

Model changes should be approached carefully, as adjustments are inevitable during the development phase. The crucial steps to follow are:

1. Firstly, it is critical that the Model Change Policy is regularly reviewed to ensure it remains appropriate and does not act as a barrier to future model changes. A balance needs to be struck between proportionality and flexibility.
2. One effective method is to conduct parallel runs, where the new model operates alongside the existing one using the same inputs. This allows for reconciliation of outputs throughout the process, helping to ensure accuracy and consistency. It is also important to make sure the model’s methodology, data and assumptions are aligned with current practices and that everything is clearly documented.
3. Additionally, conducting a comprehensive analysis of change is critical. This would include breaking down the movements between the old and new model into high-level categories such as methodology, data and parameterisation to inform key stakeholders of the material movements. It supports smoother adoption by helping users understand the rationale behind the changes and ensures that any deviations from the existing model are well justified.
4. Bringing in external reviewers, such as consultants or contractors, adds an extra layer of scrutiny. Their independent perspective helps validate the model and ensures that it is consistent with industry standards. There is also the added benefit of providing assurance to senior management and the Board.

As actuaries and risk managers are always trying to keep up with a complex environment and provide valuable insights, the importance of robust, adaptable and transparent models cannot be underestimated. Whether refining an existing internal model or transitioning to an off-the-shelf solution, the ultimate goal should be to ensure that the model remains fit for purpose, delivering appropriate and timely insights that support sound decision-making and regulatory compliance. 

This also requires faith in external models provided by a third-party, leveraging validation for a proper look-through approach.

"Future-proofing, operational flexibility and rigorous validation processes are essential to safeguard against emerging risks and evolving market conditions."

By adopting a controlled, well-documented approach to model changes and engaging stakeholders throughout the process, insurers can maintain confidence in their modelling framework while positioning themselves for long-term resilience and success.

Please feel free to get in touch with our experts if you would like to discuss this topic further.