With IFRS 17 implementation in January 2021, insurers' finance and actuarial systems will be tested to their limits. Companies are under pressure to find faster, better and cheaper ways to measure, model and report, driving renewed interest in actuarial software across all markets. Mark Brown and Juriaan Borst discuss how modelling products, innovative technologies and techniques are evolving to meet the rapidly changing needs of insurers.
What is the difference between financial and capital modelling?
Brown: Financial modelling is about projecting how a business will evolve over time for the purpose of producing statutory reports and capital assessments; whether the focus is on seriatim liability modelling or stochastic ALM modelling.
The trend for regulatory change in recent years has created a number of challenges for companies in balancing accuracy and precision, whilst maintaining a manageable set of models. Scaling up the computing power has always been one solution, but it has at the same time been necessary to move away from coding a pure simulation to introducing approximations in the model, whether through simpler code, compressed data or smart valuation formulae.
Borst: Capital modelling is an extension of financial modelling, where capital modelling allows for assessments of return on capital (which financial modelling doesn't). It allows for understanding what the capital requirements would be under stressed conditions, under different asset strategies, business mixes, hedging and reinsurance strategies.
Different methodologies and tools are typically required to undertake the projections required for capital modelling. For example, projecting the balance sheet of a company and the policies towards the future is something that can be done for a few thousand scenarios on a per policy basis. At some point, however, you will be unable to close the calculations fast enough, for example when doing daily solvency monitoring.
Therefore, specialised capital calculations, different techniques, models and concepts are necessary. And that is exactly what we can do with Willis Towers Watson's RiskAgility Economic Capital Aggregator.
Both specialisms have been around for some time. What are clients now looking for in their models?
Borst: Our clients are looking for the continuous refinement of results, expanding functionality and usability, in addition to allowing better models to reflect reality as closely as possible. Speeding up the models to allow results to be produced in near real time and shortening the reporting cycle is another priority. Capital models are also being increasingly used beyond their original function of a better calculation of the required capital, with clients using them to support business decisions, M&A activities or ALM-types of studies. In today's world, it is of key importance that economic capital measures are used at all levels of risk management and decision making, which can only be properly achieved with a robust economic capital model in place.
Furthermore, the implementation of Solvency II exposed considerable shortcomings in the models and applied pressure on them to report much more in a shorter timeframe, driving greater interest in the quality and performance of liability models. With IFRS 17 fast approaching, we are seeing a second wave of pressure building which will have an even more profound impact on the liability modelling function.
Brown: My experience is similar. Companies have models that were built many years ago, which were once fit for purpose under a previous reporting regime. They are now trying to keep the best of what they have got, but at the same time make it easy to maintain. They are also looking for additional reporting speed and reduced costs when doing the calculations, especially for financial modelling which tends to involve large calculations, without losing the accuracy or the consistency with what they have done in the past.
There are some markets where capital models have yet to break in, where companies continue to use traditional reporting methods and remain heavily focused on financial modelling. Typically, these markets have features that avoid the need for complex dynamic and stochastic routines. These are in contrast to more financially demanding companies and markets that have a greater focus on economic capital modelling and the investment opportunities and risks beyond mortality and morbidity.
How are new technologies and techniques helping to meet these demands?
Brown: Finding ways to do things faster, better and cheaper has come to be a defining goal for many insurers' systems, processes and financial models. After a significant period of resistance, cloud computing has been accepted remarkably quickly, giving companies the ability to make massive scale calculations on demand, while saving costs during off-peak periods. Companies are now exploring new technologies, such as AVX vectorisation, as a means to not only complete any additional calculations required but also to drive down the costs of doing those calculations. Companies are also looking at alternative techniques for monitoring risk, instead of simply throwing more computers at the problem.
Borst: Capital modelling has primarily been about using advanced statistical and mathematical techniques to run a lot of projections of the future balance sheets. So, until recently it's been less about using new technology and more about expanding the use of these techniques. But what we are now seeing is an increasing focus on the next step, which is to make greater use of new technologies, like AVX, and the distribution of calculations in the cloud to make the models run faster.
In response to increasing demands from insurers to embed, for example, more sophisticated pricing approaches as competition continues to ramp up globally, Willis Towers Watson has been actively researching and developing the application of Artificial Intelligence and the latest predictive modelling techniques, such as Gradient Boosting Machines and neural networks, across its software portfolio. As a result, organisations are already making progress in using machine learning platforms to enable a quantum leap in data utilisation in order to quickly optimise expertise and technology in solving a variety of business problems faster and better than their competitors.
How do you see your products evolving to meet the changing needs?
Brown: Over the last couple of years, Willis Towers Watson has augmented its life product offering, Solutions for Life, with a range of cloud-based managed environments and hybrid-computing options. This has opened up our actuarial users to a new world of capability, unfettered by the shared IT capacity.
Going forward, we are updating and optimising the calculation engine to control the data implications of large-scale cloud usage and to take advantage of vectorisation technologies, such as AVX and GP-GPU, with a focus on driving down the cost to our customers of performing necessary actuarial work.
Borst: Significant investment and functionality development goes into our capital modelling tools in order for them to be integrated with our Unify process-automation platform. This added functionality allows us to make these calculations and analysis part of a fully automated and easily repeatable process, running 24/7. So, complex calculations are possible on a very frequent, robust and automated basis. This also means that these tools, which the clients already have, deliver even more added value as a result of being used on a more regular basis.
Brown: Unify is also hugely powerful in the financial modelling world. Not only does Unify allow a user to perform more advanced inter-dependent modelling, the platform also helps an insurer tighten up the governance process, reduce the time taken to produce their statutory reporting, and integrate with accounting systems for the purpose of adopting IFRS 17.
More value can also be generated from a company's IT investment by allowing finance professionals to work smarter, not harder, and take advantage of 24/7 processing and calculations. Instead of being limited by inefficient or fragmented processes, lots of manual intervention, tight timelines, and increasing business and regulatory demands, actuarial teams can be given the opportunity to provide quick, value-added analysis.
What practical advice would you give to a client starting a new project?
Brown: It is important to stay focused on the goals and projects, rather than be distracted by over-polishing the pieces that are familiar. This is true whether an insurer is working on an IFRS 17 integration project, an actuarial model replacement or implementing a capital model. It is also key to engage with experienced and competent subject matter experts ahead of a project, so you know what you are trying to achieve and can stay focused on a cost-effective completion of the project.
We've all heard the statistics about nine out of ten large projects overrunning. Historically, actuaries have tended to spend a long time initially building what they are comfortable with and hence making the project difficult to complete. For this reason, massive overruns in actuarial software projects are incredibly common, with many ultimately falling far short of delivering the required benefits. Good planning and prototyping are key to better control.
Borst: And this is where robust and proven modelling solutions go hand-in-hand with implementation experience. As an example, we have completed cost-effective implementations of EC-related risk in extremely short timeframes where we have rolled out working models within six weeks.
Brown: The future of life financial modelling will consist of new IT technologies, new innovations from suppliers, and careful management of insurers' model portfolios driving down costs and the risks associated with their life modelling. We've seen cloud computing become acceptable in a relatively short time scale, and a move to services-based calculations is likely to follow.
It is easy to talk about what can be achieved without reference to practical constraints, but reality imposes many such barriers. What's important is to buy in to a sustainable long-term partner who will help you plan for and deliver the future challenges as well as the current ones.
Borst: As the life sector continues to face a volatile and challenging risk environment, exacerbated by regulatory change, the drive to achieve high-level organisational objectives of profitable and sustainable business has translated into financial, risk and actuarial goals that have increased the emphasis on speed, efficiency and cost. While the balance between the three depends on individual company circumstances, innovative and cutting-edge capital modelling solutions remain key for improving business performance.