FIS: Future modelling

FIS's Martin Sarjeant, head of insurance risk solutions management and strategy, and Scot Glasford, senior actuarial product manager, US, at FIS, discuss the future of risk modelling — and why models are so important for insurers.

What technology trends do you expect to shape the insurance industry in the near future?

Martin SarjeantThe pandemic has shown that insurance companies cannot rely solely on a brick-and-mortar office building for employees to carry out insurance operations. Insurers must look to public cloud and SaaS solutions to be part of their operational resiliency programmes.

Deploying cloud solutions enable the employees of insurance companies to operate from both home and the office, diversify operational risks and help the companies retain employees by providing the flexibility over working location. Cloud solutions also scale to enable collaborative working with resources around the globe. Also, using cloud-based technology free insurers of the burden of hosting physical on premise servers freeing up IT resources to also work remotely.

Another outcome of the pandemic tested the ability of insurers to respond to crisis. Insurers are looking for a more streamlined approach to getting the C-level information quickly. This can be achieved by insurers investing in the development of proxy modelling techniques to supplement full models to deliver important information without the need for full detailed model runs. This may be especially important in today's world with the speed in which information can be shared.

Why are models so important for insurers?

Models are central to the whole insurance company; without them it would be like driving a car with a blindfold through Manhattan.
You would not be able to price contracts, you would not be able to establish what reserves should be held and not establish if you are even solvent. Even then, a model is just a representation of the company.

Models are only as good as the effort the vendor and the insurer has put into them and only as good as the assumptions put into them. Clearly some models out there are "one size fits all" with little flexibility. These may be okay for many types of investigations but equally may miss the real risks particularly in certain stress scenarios. A good model should reflect the product features, regulation, business strategy, assumptions based on experience, market factors and judgement management actions and policyholder reactions in sufficient detail.

Models are used to run the business, from pricing and capital allocation and efficiency and for regulatory reporting. The Covid pandemic was a great example of how models and modelling teams were central to how insurers navigated through the pandemic. How the models were used and updated to provide near real time view of current risks and allowed insurers to quickly reprice contracts.

What are the key insurer challenges you believe models will help with and how does FIS's solution stand out from others?

Scot GlasfordFIS's approach to actuarial modelling does not solely focus on the actuarial calculation engine. Our approach is to think through the entire workflow of actuarial departments.

We acknowledge the actuarial job function is multi-faceted and our software suite reflects the needs of the actuarial job function. Our integrated set of software tools that help actuaries manage data, manage assumptions and coordinate modelling runs, which separates us from our competition.

The concept of actuarial modernization has been discussed in the industry as a response to the increased complexity of statutory and GAAP reporting requirements. Insurers know they must update their modelling processes to do more with the same head count and keep pace with their competitors. FIS's integrated set of software tools allow insurers the ability to build and automate actuarial processes with software from a single vendor. FIS's solution can help firms realise the idea of actuarial modernisation.

When actuarial processes become more streamlined and automated the actuarial job function will change as well. Our vision for the actuarial job function is that actuaries will be focused more on the analysis and explanation of results and not be bogged down by the management of data and building of models. We see that actuaries will shift from being number crunchers to story tellers who are able to provide important narratives about the risks facing their business.

Did you make any upgrades or improvements in the last 12 months?

Regarding our USrisk libraries and example models, we are continuing to grow with the evolving stat and GAAP regulatory environments. Recent improvements for statutory reporting include VM21 functionality to support New York State requirements. We have also begun our development of VM22 which is principals-based reserving (PBR) for non-variable annuities.

While VM22 will not be effective for a few years, we want to provide our users with the ability to test the new regulation ahead of its effective date and provide an opportunity for users to participate in VM22 field testing when its conducted.

For GAAP reporting, we have continued to follow the evolution of Long-Duration Targeted Improvements (LDTI). As LDTI is still a standard being implemented and some areas of LDTI practice is still changing. We recently have provided our users the ability to calculate market risk benefit reserves for fixed indexed annuities. We also have added a robust reporting solution for LDTI disclosures.

We have had two major modelling releases in the last 12 months as well as several minor releases which have added more model run process improvements for increased scalability, flexibility and automation and performance. We also further extended the suites API capabilities as our solution becomes more embedded into the overall organisation.