Modelling Solutions – an IT Problem or an Actuarial Opportunity?
Solvency II (SII) has led to a dramatic increase in the workload associated with the actuarial functions within insurance companies in Europe in the last 5 years. An impact assessment report for the adoption of the Solvency II Directive estimated that the one-off net cost of implementing Solvency II for the whole EU insurance industry will be around EUR 3 billion to EUR 4 billion. Whilst the SII’s Pillar 1 modelling or quantitative aspects have become more demanding (see the figure below), the biggest growth in costs has been the qualitative and data management aspects. These requirements have led to large IT projects to develop systems that provide timely, validated, transparent data to the actuarial models, run the models and then aggregate, validate, integrate and report the results. The costs associated with these IT projects have run into hundreds of millions for some global insurers.
The Actuary’s Role in Modelling Projects
Historically, actuaries focused on the models, being their main area of expertise. Data required by the models was dealt with by using tools such as spreadsheets and end user databases such as Microsoft Access and FoxPro. However, due to greater data volumes and SII’s requirements for transparent, auditable and robust data, this approach was no longer adequate and specialist IT skills and implementation projects were required to focus on the flows of data into and out of the models. This has changed the nature of the actuary’s role from previously owning the whole valuation process to being part of a much bigger project. So does this mean that valuation actuaries have been relegated to an IT support role providing input for requirements, specifying calculations and checking results?
Sources of Friction
Not necessarily. A partnership between the IT and actuarial departments can work well for the delivery of modelling solutions. However there are plenty of points which can become a source of friction. These include
- pressure to use systems which are already part of the IT estate,
- competing priorities with other IT projects,
- increased turn-around time between requesting a change to the model and the actual implementation,
- additional time and effort required to communicate requirements to IT in a formal manner via requirements documents,
- depending on the system used, there is sometimes a loss of visibility to the end-user (the actuary) of the details of the modelling process once it has been implemented by IT.
These large scales IT systems take time to implement and change. However, actuarial requirements continue to increase and change. This can lead to a gap between the IT needs of the actuarial department and the IT department’s ability to address these in its core systems. This gap typically continues to be addressed using the same end user computing tools referred to above. However, these lack the attributes of the core systems such as control, auditability, transparency, documentation, scalability and automation. In order to address these, actuarial departments add people with IT skills to the team potentially leading to an overlap or conflict with IT departments.
Opportunities to Regain Control
This issue is not unique to actuarial departments and the scope for overlap is growing due to profound changes that have taken place with how IT services are delivered to end-users. Cloud-based computing means that the hardware and other infrastructural elements of a modelling solution are readily available, and in a way in which subject matters experts can utilise and configure these resources, without having to involve an army of business analysts, programmers and testers.
More recently, software is now available that operates on these cloud systems that enable subject matter experts such as actuaries to create the “plumbing” that is required to collect, prepare, validate, combine, map, cleanse and store the data that feeds into models. These platforms also maintain versions and documentation automatically and also collect, aggregate, summarise and report on results that come out of the models. This approach enables actuaries to do many of the tasks that would otherwise need to be done by the IT department in traditional IT development projects.
These new modelling platforms place the actuaries back at the centre of the solution. This enables actuaries to take control over both the modelling and data flow aspects of their processes thereby reducing the implementation times and communication costs involved in larger cross-disciplinary projects. It is likely that the internal IT department would still be involved in this process, including vetting that the third party provider complies with their policies for security, business continuity and data governance.
Take Advantage of New Tools
In an article in August 2014, John Rowland, the global leader for life insurance capital modelling at Towers Watson states “One area definitely worthy of consideration is data handling, particularly cleansing and validation. It is surprising how much manual manipulation of data goes on in many firms when simple tools can be deployed to improve automation, transparency, audit trails and validation of data as it is prepared for model runs”.
This is echoed in the EY European Solvency II Survey 2014: “Almost one in six companies have 60% of these processes automated. This demonstrates that although some progress has been made, there is substantially more potential —perhaps in alignment with other activities—to improve the information flows throughout the organization. Leveraging technology and data progress to improve risk management cost effectiveness is an area of opportunity for many insurance companies.”
Therefore despite the large investment in actuarial and IT activities and the significant progress made, there is much more that can be done by actuaries using new technology.
There are parallels for this shift in other disciplines such as data analytics where cloud-based tools are enabling analysts to collect, manage and analyse large volumes of data without the need for IT departments to create databases, write software or install any hardware.
As with many other cloud or Software-as-a-Service (SaaS) applications, these new modelling platforms help to level the playing field, enabling small and medium size insurers to achieve the same results and efficiencies of some of the largest insurers but without the very large, upfront development cost. They enable actuaries to address data issues without the need to set up large IT projects to take on these tasks. In due course, the compelling cost advantage and security of these tools will mean that even the largest companies will adopt these technologies. This is already happening with other SaaS products from companies such as Salesforce, now used by many global businesses as their main Customer Relationship Management system.
Actuaries should seize the opportunity presented by these new cloud based technologies. Doing so will help them retain their position as the pre-eminent managers of risk within insurance companies as well as reducing the cost of implementing and maintaining Solvency II systems.