Solvency II:
turning financial regulation into business opportunity
The Solvency II regime is here to stay. European pensions and insurance companies have no other choice but to absorb and align the new legal framework within their IT systems and infrastructure. This article examines how Danish pension fund ATP used its operational platform as a point of departure to embrace the regulatory challenges embodied in Solvency II, transforming legislative requirements into a business opportunity rather than treating them as an obstacle.
by Jacob Elsborg, MBA, M.Sc.
In July 2001, the Danish Financial Supervisory Authority (FSA) introduced new financial regulations, which fundamentally changed the rules of the game for Danish pension funds, including the country’s largest – ATP. The centrepiece of the financial reform was the introduction of mark-to-market valuation of pension liabilities. The FSA also tightened required standards of risk management, risk assessment, and transparency. A key element was the requirement to conduct resilience tests, or so-called traffic lights. The net effect of this new procedure was to tighten the overall solvency requirements for all Danish financial institutions, including ATP.
The FSA regulations rather than the Solvency II regime were the catalyst that led ATP to rethink its mission and its strategy for accomplishing it. Looking back, the process ATP started would be very similar if the Solvency regime1 had come prior to the challenges mentioned above. The consequence has been that adoption of the Solvency regime has been aligned with the business process re-engineering (BPR) that has been undertaken over the last years.
The integration of the requirements stipulated by Solvency II is seen as a natural extension of developing the ATP business model, and engendering a competitive advantage rather than a regulatory burden. This is also the reason why we start by taking a closer look at the development of the ATP business model and the BPR, which has led to an operational platform that allows ATP to turn Solvency II into a business opportunity.
THE ATP BUSINESS MODEL
At the start of the millennium, the challenges of increased market turbulence and repeated financial crises together with a fair value disclosure regime led ATP to conclude that its approach to pensions needed to be reappraised. The change was not a minor tweaking in investment strategy - it was to rethink the organisation’s business model from the ground up. The goal was to reconcile a return-seeking investment strategy with safeguards to pensions and pension pledges through sustainable guarantees and effective risk management.
Rethinking the business model entailed a multi-year BPR involving 12 elements organised into four categories. First, on the overall business side, ATP adopted a new and integrated view of pension management, created a new business area called liability hedging, and renewed its risk management practices to provide timely warnings of changes in risk patterns. Second, on the investment side, the overall strategy was redesigned by adopting an absolute return strategy, separating beta and alpha portfolios, and entering into strategies to hedge tail risk. Third, on the liability side, ATP developed and implemented a new pension accrual model preserving important features from the old model, while designing the new model to continually balance the accrual of new pension rights with economic realities and the investment policy. Also, a new mortality model was designed to capture and address the longevity risk, and a daily valuation of the liabilities was introduced. The fourth category involved strategic management of the operational platform, defining the operational platform, setting up the organisation and responsibilities, and developing an operational framework.
An integrated view of assets, liabilities and objectives
A pension fund’s overall objectives and risk tolerance, investment policy, and pension policy constitute the three key focus areas of pension fund management (see Figure 1).
Integrating decisions in these three areas is a critical governance challenge. For example, changes on the liability side can impact investment policy. The most dominant example of this is that life expectancy increases, pension liabilities increase and reserves decrease, and thereby threaten the solvency degree. This in turn reduces the ability to act. Overall, a risky long-term investment strategy makes little or no sense if a fund’s tolerance for a red-light risk2 is low, while the policy of indexation on the other side of the balance sheet involves expedient consumption of the reserves.
The challenge is to design and implement strategies and policies that are consistent with the overall objectives and that take into consideration the relationship between the asset and liability side of the business model. ATP’s reaction to this challenge was to develop an in-house asset and liability management (ALM) model, which ensures that allocations to risky assets are dynamically controlled as a function of the size of the reserves and ATP’s risk tolerance. This approach includes a total (assets and liabilities) mark-to-market valuation on a daily basis.
In brief, the four guiding principles of the ATP business model are defined as:
- ensuring an appropriate risk level -> the investment risk is defined in view of ATP’s free reserves;
- avoiding uncompensated risk -> liabilities are hedged in full;
- diversifying aggressively -> the portfolio should do well in any circumstances;
- hedging tail risks -> solvency should be protected by hedging against black swan events.
Fig. 1. The interdependence of the three key focus areas in pension management.
Source: ATP.
Investment objectives
ATP’s two main investment objectives are to protect reserves against adverse developments in the financial markets and to create excess return in order to ensure the purchasing power of the pensions. These objectives have led ATP to divide assets into two independent portfolios – a hedge portfolio and an investment portfolio.
The hedge portfolio’s aim is to eliminate mark-to-market risk of the ATP liabilities. This portfolio is not expected to generate excess return. In contrast, the investment portfolio is designed to generate excess returns, which in turn requires taking on investment risk. As the hedging portfolio predominantly consists of derivatives, it does not in itself consume liquidity.
So in principle, all ATP assets are available for the investment portfolio. Both portfolios are managed through risk budgets defined as dynamic risk budgets. These are set up to handle the trade-off between the aim of producing higher pensions for the members against the risk of losses, and thereby compromising further pension payments. The essence of dynamic risk budgeting is that the risk is reduced before a threat materialises.
A central challenge for an asset manager in the pension industry is the quest to produce high stable returns in order to secure the purchasing power of pensions. One of the biggest threats facing funds takes the form of severe losses, because such losses reduce reserves. This again reduces the ability to increase pension benefits in the future, and having a solvency margin to protect it can force the fund to sell off risky assets at an inconvenient time.
Along with adopting a new approach towards risk management, ATP changed the concept of return target from a relative return target to an absolute return target. As a consequence, the new definition differs from the traditional benchmark-based framework, where the focus is on creating a return above a benchmark portfolio. The absolute goal ATP has defined for the investment portfolio is to achieve a return net of taxes that is at least equal to liability hedge funding costs, changes in longevity, and indexation of pensions and pension rights in line with inflation.
The liability side of the coin
Changes were also made on the liability side. ATP developed a new pension accrual model that was able to adapt more readily to the complex realities of market fluctuations and mark-to-market valuation. It has to be noted that the social objectives and values from the old pension models were maintained. The new model provides for a lifelong guaranteed pension, with accruals based on collective insurance principles.
In effect, ATP will remain a defined contribution model in the sense that benefits reflect individual contributions made, while resembling a defined benefit scheme through the applied guarantee. The new model was designed in respect to valuation within the framework of the mark-to-market valuation. Since 2001, ATP has on a daily basis calculated a full valuation of its assets, and since 2003 has on a daily basis calculated a valuation of the entire reserves defined in the balance sheet. The new pension model made valuation of the reserves even more specific and in line with valuation of the assets. From 2010 on, a full valuation was made based on daily updated data.
ATP’S OPERATIONAL PLATFORM AND ITS STRATEGY
Along with the overhaul of the business model, the operational platform of ATP and its strategy were defined as integrated parts of the business strategy. The definition included a strategic framework for the operational platform to redefine and establish data, processes, information and system management within operations in order to ensure quality and performance in each area. Besides this, the challenge was to create an operational platform that was flexible and scalable while reducing the total costs of operations.
The creation of a new operational platform and its strategy comprised: defining the operational platform and its implementation within the overall business strategy; setting up the organisation and responsibilities; and developing an operational framework.
For ATP, implementation of the operational platform strategy thus became an integrated part of the business strategy (see Figure 2).
Some asset managers define an operational strategy as the strategy that incorporates the tactics and processes that support the investment strategy. The operational platform strategy expands the operational strategy by including the functional setup of the platform as a part of this strategy. Sometimes the functional setup is defined within the framework of the IT strategy, but, far too often, the functional definition is left in no-man’s land (i.e. not forming part of any strategy).
Fig. 2. Schematic illustration of an operational platform strategy as an integrated part of the business strategy. Source: ATP.
Further, ATP has decided to define an operational platform in the following way: “An operational platform in the asset management industry is where management of the organisation’s data and information together with the execution of decisions take place.”
Besides defining the operational platform and its strategy, the organisation was aligned in order to ensure that only one team handled all functions (see Figure 3). Many teams in ATP’s Risk and Operations department had been working with data quality, however, during development and implementation of the operational platform, the data management function was set up – ensuring that only one group was handling the entire dataflow for users of the operational platform.
Fig. 3. Functional setup of ATP's operational platform. Source: ATP.
A NEW OPERATIONAL FRAMEWORK FOR SOLVENCY II
Having defined an operational platform and its strategy, Solvency II now presented ATP with the need to develop an operational framework which could handle its requirements, and here the main tasks constituted: preparing data for the calculations; implementing calculations and processes; and reporting. The first step to be considered was gathering and washing of data, by many described as an overwhelming burden. However, ATP was well prepared for this.
As early as 2000, ATP had taken a strategic decision to store all investment data within the investment management system SimCorp Dimension. The decision in practice made this system function partly as a data warehouse for ATP. The approach has been followed consistently, and today the data management team ensures that all data from various vendors are entered into this data warehouse and that the data has the right quality. The setup ensured that the Solvency II implementation project only demanded minor extensions concerning the new data needed, and the existing framework could easily handle the new tasks.
In the past several years, ATP has worked on integrating the liability and investment sides of its business which includes a total (assets and liabilities) mark-to-market valuation on a daily basis as described earlier. As a consequence, the data on the liability side was available prior to implementation of the Solvency II regime, executed on a shared operational platform, and therefore it was not a question of how to produce or access data.
The second step is the development and implementation of the calculation processes. One part of the calculations is handled by an external partner already implemented on the operational platform at ATP. The second part is the liability side, and this application was already in place as concerns fulfilling Solvency II requirements. Data from both applications are then used within an ATP-developed Solvency II application. The programme is developed as an executable programme and the functional flow is set out in Figure 4.
With the development of the ATP operational platform, process management was defined as critical for the operations. Accordingly, the way automated processes were operated changed, and a group within the Risk and Operations department was made responsible for operating all automated processes in the Pension and Investments area as a whole.
For these operations, ATP has built a data model handling both the data and process flow. In brief, all data entries and processes are defined as events. When an event happens, the result can be either a success or a failure. This leaves the daily quantitative data management (i.e. managing incoming files and data) and daily process management as exception handling; if an event fails, the process management or the data management team takes care of this. When an event fails, the process flows are set up to wait until the event is defined as ‘OK’, and then the flows continue which is crucial for the time spent on daily operations.
Besides handling the automated flow, the data model has made it possible to incorporate non-automated processes in scheduled flows. In order to incorporate a non-scheduled task in a flow, the non-scheduled task has to be defined as an event, and when the task is done, the result (success or failure) is written into the status table in an Oracle data warehouse. In this way, the non-scheduled tasks are represented on equal terms with the scheduled tasks.
The third step in implementing Solvency II is the internal reporting. Also with regard to handling this challenge, ATP’s operational platform already has a setup in place, as ATP has a portal where all information shared within the Pension and Investments area is presented. Therefore the task was to define the reports based on the data stored by the Solvency II application within the Oracle data warehouse.
Fig. 4. Solvency II functional flow on ATP operational platform Source: ATP.
OPPORTUNITY SURMOUNTS OBSTACLE
By way of summary, ATP’s implementation of Solvency II has underpinned the strength of its new business strategy and the operational platform. ATP has absorbed Solvency II within its overall risk management, and due to the management and development of the new operational platform, the focus of the implementation of the Solvency II requirements has been on the definition and development of the calculation engine, not the processes surrounding them.
In terms of business strategy and the operational platform, it is clear that Solvency II implementation is merely a step in the direction that ATP has taken in the last years. Due to the work of the previous years, where ATP management has worked closely with the board in developing the business model, the Solvency II regime has been based on this, and therefore the conceptual framework is basic knowledge.
The way risk is measured in the risk budget model means that the ATP risk management protocol avoids solvency traps, such as being forced to reduce unhedgeable risks in times of high volatility and negative returns, or being unable to take on additional risk when risk decreases and mean reversion sets in.
The focus of measurement is not on absolute solvency, but on the risk of experiencing a red-light situation within a three-month period. So even under severe adverse conditions sufficient reserves are available to take on additional risk. The aim of implementing an internal model under Solvency II is – so to speak - to replace the former red-light model with the model used for solvency calculation, and thereby to use the model in the daily risk management.
Implementation of the Solvency II requirements as a part of the operational platform was a perfect match. Platform operations were prepared for the change. The main processes and responsibilities were in place in order to support a daily production.
No changes within the area of responsibilities were made in order to support the daily production of Solvency II. The data gathering and quality assurance were already in place, and with a few changes to the existing framework the data production was also in place. Processing the calculation and reporting went equally smoothly due to the flexible setup defined as engines on the platform.
Implementing the Solvency II regime has therefore focused on the definition and development of calculation and reporting, and ATP has not experienced any operational challenges implementing daily calculation in the organisation. All in all, the main part of the work already done in respect to implementing a Solvency II regime has enhanced ATP’s risk management and therefore ATP perceives the implementation as a business opportunity and not as a regulatory burden.
Jacob Elsborg, MBA, M.Sc., is Head of Technology for ATP’s investment department, a position he has held since 2000 and is responsible for the department’s operational platform. He holds a Master’s degree in Economics and Mathematics from the Copenhagen Business School and an MBA from the Henley Management College in the UK. Jacob Elsborg has worked in the financial industry for much of his professional life, starting his career as an IT economist for Danmarks Nationalbank, the central bank of Denmark, from 1995 until 2000.
