Best tech vendor for Solvency II: IBM

Paolo Laureti, IBM Risk Analytics

Buy-side Awards 2016

With Solvency II now in effect, insurers are beginning to ask whether the time and investment they put into meeting the regulatory requirements might also have value for the business.

"Our clients made significant investments in enterprise risk management [ERM], especially the ones subject to Solvency II. Now they want to show their spending is not only for compliance – it can also bring business benefits," says Paolo Laureti, offering manager of insurance ERM solutions at IBM Risk Analytics, based in Rome.

To serve the range of size, complexity and level of sophistication of potential clients, IBM's Algorithmics Economic Capital, ERM and Solvency II solution comes in two versions – the Compliance and Reporting Edition, and the Enterprise Edition – with users able to upgrade from the first to the second as required.

"Some companies are starting to use their ERM models to calculate the impact of business decisions on capital and risk profiles. This includes the introduction of new products, a different asset allocation or potential mergers and acquisitions," says Laureti.

Such tasks often present challenges, he explains: "If you are looking to do what-if analysis in real time with hundreds of thousands of scenario simulations, your hardware and software has to be fast."

The performance can be improved by variance reduction techniques for speeding up Monte Carlo simulations. One of the advantages of IBM Risk Analytics is it can draw on the wider company's deep resources of expertise and technology, including data extraction, transformation and loading – a ubiquitous issue in ERM and Solvency II compliance – as well as the installation of high-performance hardware.

Italy-based insurance group Generali has called on a wide range of IBM resources and capabilities in preparing for Solvency II. In 2013, it bought IBM Algorithmics Economic Capital, ERM and Solvency II to help it build an internal model for Solvency II and to industrialise its standard formula process. It also aimed to explore the further business benefits of this infrastructure once its compliance process was in place.

With the IBM Algorithmics platform, we have been able to create a new capital system from scratch and attain regulatory approval in less than two years. Our priority is to maximise speed and performance without losing accuracy, and that is where we see the strength of the IBM solution
Gerardo di Filippo, Generali

The implementation at Generali was challenging not only because of the complexity of its organisation and consequent large number of hierarchy nodes required for its reporting structure, but also because of the numerous cross-links between them, which had to be fully modelled to obtain a realistic representation of the overall balance sheet position. Furthermore, Generali wanted to model losses down to each line of business, rather than at legal entity or other higher organisational level, resulting in a complex reporting hierarchy that is granular and therefore demanding in terms of workflow and computation.

The IBM platform was able to incorporate more than 500 risk drivers in a correlation matrix and more than 2,000 different risk factors to model the various Generali businesses across different geographies. All balance sheet items are modelled via proxy functions for greater flexibility and reduced implementation time, while still giving robust results. Generali is using approximately 100,000 proxy functions to mimic the movement of the different balance sheet items, enabling it to meet its precision and granularity targets.

Given the size of the hierarchy and the number of loss functions, keeping simulation time down to a manageable level while ensuring an acceptable level of stability of the results was a challenge. Generali adopted IBM's Sobol quasi-random number generator to generate stochastic scenarios, with significant performance gains – Generali is now able to achieve the desired level of stability of both standalone and diversified risk measures using around half as many scenarios compared with the standard Monte Carlo scenario.

Gerardo di Filippo, head of group risk capital calculation and reporting at Generali, says: "With the IBM Algorithmics platform, we have been able to create a new capital system from scratch and attain regulatory approval in less than two years. Our priority is to maximise speed and performance without losing accuracy, and that is where we see the strength of the IBM solution. The flexibility and reliability of the platform allowed us to place risk at the heart of the business, and to incorporate risk into our business decision-making."