A step change in nested stochastic modelling
As actuaries in the insurance industry address the shift to increasingly complex modelling of stochastic components, the demand for nested calculations is growing, with the accompanying computational burden expanding exponentially.
Nested stochastic pricing and models are used to produce balance sheet projections – an explicit requirement under Solvency II, IFRS 17 and other legislative programmes. Nested stochastic models allow insurers to manage the business appropriately, price new products, project earnings and measure risk.
This is a more common challenge for life insurers than it is for carriers in the property casualty markets, as while the technique may be used for any actuarial assumption, the interest-sensitive nature of life insurance business means that stochastic modelling is particularly useful for this segment.
Due to runtime and scenario generation issues, running these types of projections and analyse the resulting information require significant changes in hardware and software infrastructure.
At a time when the demands placed upon insurers by regulators have exploded, and the availability of data and computational power offered by software providers and modellers have grown, a major step change in the development of nested stochastic modelling is emerging.
Although nested stochastic modelling has become increasingly popular throughout the wider global financial sector in recent years, the availability of such tools for consistent use at scale has traditionally been lacking, the result being a great deal of time spent by actuarial functions on complex calculations, but creating inconsistent results. The application of nested stochastic techniques in insurance comes with its own technical and operational challenges.
Through our R³S software suite, RNA Analytics has long offered clients the ability to perform nested stochastic calculations, and our Solvency II FLAOR model package includes examples of our curve fitting / LSMC proxy models (as well as including a European liability style nested stochastic model).
Now, as a result of ongoing dialogue with clients, our modelling experts have developed a new tool to address the analytical and computational challenges that insurers are increasingly experiencing as they seek to fulfil ever demanding and complex forecasting requirements. The R³S Nested Stochastic Model Package has been developed by our market-leading and multi-award-winning modelling teams for initial roll out to clients operating in the APAC region, with a view to expanding the solution globally.
New demands, new models
Importantly, the new model package allows IFRS 17 model integration, making the transition to the new accounting rules seamless, with unparalleled end-to-end functionality, and offering greater accuracy and faster run times than ever previously achieved.
Standard features of the new package address traditional challenges of PV discounting (only projected cashflow amounts are read in as data); input projected cashflow data is already grouped into IFRS 17 reporting portfolios; and input projected cashflow data is stochastic by RN scenario.
When it comes to IFRS 17 NSP Input Sources, the values (including the discounting later) vary by RW scenario. The projected CFs sublayer averages out the RN scenario valuations.
Further benefits of the new R³S Nested Stochastic Model Package include application lifecycle management (ALM), incorporating software testing, architecture and maintenance.
Because all insurers' books of business and output requirements are unique, the new package is fully customisable.
This release is the first of a series of nested stochastic solutions in the pipeline for RNA Analytics’ growing client base, which includes introducing existing R³S Modeler functionality to the package, such as explicit asset CF modelling to create fully dynamic ALM, replacing NSP liability values with curve fitting proxy solution to create semi-dynamic ALM; and beyond this, asset hedging.
Our modelling experts are committed to breaking new boundaries in the nested stochastic arena. Contact us to find out how we can help your team achieve new outcomes.