 # UCM: Performance and Flexibility

Keylane | The Keylane’s solution to the uniform calculation method is developed with care for both flexibility and performance. With a configurable calculation model any product can be forecasted. The performance is optimised by combining our expertise in actuarial science and software development. With our critical eye on actuarial field and interpretation of the law, we have also added an elaborated and more consistent version of calculation method 1. Read on to find out more.

#### Uniform Calculation Method

On April 23, 2018, the ‘Staatscourant’ published the decision regarding the calculation methods for the representation of retirement pensions in scenarios. Within the law it is prescribed how and which calculation methods should be used. This is part of the law on improved premium arrangements. The prescribed calculation methods are summarized in the Uniform Calculation Method, abbreviated UCM.

As Keylane, we have managed to implement the Uniform Calculation Method in our software while retaining the flexibility to suit for all arrangements and by optimising the performance such that large batches can easily be calculated. The calculation tool is available via a web service and can be used by any life and pension software solution. The optimal software solution is achieved by combining our expertise in both the actuarial field and software engineering.

Within the Keylane solution, we have implemented the three calculation methods as prescribed by law (generic calculation method, calculation method 1 and calculation method 2). Additionally, we have implemented a detailed version of calculation method 1.

##### The generic calculation method

The generic calculation method is implemented as highly flexible solution, fitting to any arrangement. The flexibility is reached by using a configurable calculation model (tabular structure) where each projection month is evaluated in a row, according to the formulas defined for the columns. Any information from the life and pension software can be used by the calculation model to evaluate the pension payment amount over the projection years. After a one time configuration, the system performs 2000 runs through the calculation model(s). The input of each run is participant, arrangement and scenario dependent (for example the participant dependent retirement date and the scenario dependent interest rate of the assets). Aside from the flexibility, the system performs fixed calculations. These fixed calculations ensure that the law is obtained. Although high flexibility is guaranteed in the generic calculation method, there is no lack of performance. The webservice performs the calculations and returns the positive, expected and negative scenario payments within a split of a second.

##### Calculation method 1

Calculation method 1 is implemented as high performing solution, using the formulas as prescribed by the law. This means that the 2000 scenarios lead to three calculation method scenarios. From these good-, expected- and bad weather scenarios (based upon the purchasing power factors), the expected real pension payment amounts are returned.

Within calculation method 1, we have added an extra feature. It is possible to perform the calculations without the simplification of 2000 actual scenarios to three calculation method scenarios. This extension can be interpreted as a ‘generic calculation method calculation’ with the performance similar to calculation method 1.
We have added this feature because the reduction in calculation method scenarios leads to inconsistency of using the scenarios *. By sorting the actual evaluated pension payment amounts, all scenarios remain separate. Additionally, we believe that the simplification is introduced due to foreseen performance issues. This argument is however invalid as the performance drop is nearly unnoticeable (approximately 5% drop).

##### Calculation method 2

Calculation method 2 is, similar to calculation method 1, implemented as high performing solution using the formulas as prescribed by the law. This calculation method is defined for arrangements and pension products for which no feasibility test results are available. Each time new company information is stored in Axon, our solution performs new initial calculations. These initial calculations lead to output of a simplified feasibility test. This output is subsequently used in the calculations for each participant.

We have implemented the UCM solution as a stand-alone calculation tool. This means that it is available for both, Keylane customers and non-Keylane customers. By our vision, the stand-alone component should perform the UCM calculations whereas the life and pension administration systems should communicate the results.

#### In the Keylane UCM module, performance and flexibility are key:

• In the generic calculation method, any arrangement can be forecasted
• For both, calculation method 1 and calculation method 2, the performance of the calculations is within a number of hundredths of a second