Research, Calibrate, Simulate, Validate

Predictive Analytics
Data analytics from professional hands

ROKOCO Predictive Analytics combines actuarial and financial market expertise, econometric and statistical expertise and computer science know-how. Questions in the area of data analytics are answered in such a way that practical benefits can be drawn directly from them. We combine experience in the insurance industry with current research. In close exchange with academic science, we develop and implement calibration and simulation methods.

Economic Scenario Generator (ESG)

Capital market scenarios are used in many fields and sectors. Real-world expectations of capital markets are playing an increasingly important role in the crucial issue of corporate planning. General economic key figures also have a considerable influence on projections for corporate management. In modern simulation programs, such real-world scenarios are occasionally used stochastically, whereby the deterministic form is often used.

Another prime use of capital market scenarios is the valuation within risk-neutral models. This applies particularly to companies in the financial industry. The calculations and simulations to be created are predominantly stochastic. For insurance companies, there is a need to create a market value-oriented (economic) view in addition to the German GAAP-oriented, deterministic reporting. This can have an impact on the investment universe, including targeted financial market products. Consequently, institutional investors are affected with regard to making appropriate strategic asset and risk allocation decisions.
Based on the stochastic simulation methods used for the analysis, scattered results that depend on the postulated interest rate structure follow. Particularly in life and health insurance, but also in the long-tail business of property / casualty insurance, this can have significant consequences for risk and capital management. The interaction of investment income and actuarial practice produces increases in fluctuations in the results, especially those of the solvency capital.

To be able to correctly evaluate asymmetries and key risk figures, the respective parameters must be determined with as many yield developments as possible. The best possible depiction of the investment is required for this purpose. Interest rate structures alone are not enough. Risk factors such as stock indices, real estate performance and credit spreads, as well as other economic factors affecting the claims area, including feedback across the causal network (correlation and causality), need to be simulated.

Data Science Data Science

Our established procedure has been certified by a leading auditing firm to guarantee our clients greater audit security in their use of Solvency II reporting. Because the generation of capital market scenarios containing several thousand paths is very computationally intensive, ROKOCO Predictive Analytics offers the option of producing capital market scenarios on a company’s behalf.

Scenarios for the following purposes can be ordered:

  • risk-neutral valuation (e.g. for Solvency II)
  • real-world valuation (e.g. for planning)
  • PRIIP calculations
In this case, the produced scenarios can be conveniently and easily retrieved via a file server for use in the company. This saves the client tedious tasks, such as the search for suitable simulation seeds or carrying out the complex validation process. In addition to the scenario files provided, supplementary documentation, such as detailed validation reports, method documentation, etc. are also included.

We support our clients with these services:

  • validation
  • calibration
  • simulation
  • training on ESG-processes

PRIIP-ESG – comming Q2 2019

Generating the key information documents (KIDs) for PRIIP has meanwhile become a fixed component in the process of new business for life insurers. The simulation of capital market scenarios with an ESG is a key component of this process. Due to the relevance towards the end-customer, the quality requirements are immense. For this specific reason, ROKOCO has developed a web-based PRIIP-ESG, which not only simulates the scenarios, but also validates them. The output can be directly plugged into the preferred calculation system/tool for PRIIP. The validation results can be used to document that the used scenarios fulfill the regulatory requirements.
The PRIIP-ESG as a web-app will be available by mid of 2019. A cost-free version will enable users to generate all scenarios required for PRIIP (including stress scenarios). It includes the validation function, which checks the simulated seed. This version is ideal for companies, which use generic KIDs for their products.

For insurers that count on specific KIDs and a large selection of funds an extended version is available. There, supplementary functionalities such as the download of validation reports, additional configuration options and increased performance are included.

PRIIP – category 2

For the calculations needed for PRIIP Kategorie 2, the PRIIP Scenario App „PRISCA“ was developed.
For more information about PRISCA please refer to our cooperation partner f-fex AG.

Artificial Intelligence & Machine Learning

For more than 7 years, ROKOCO has put mathematical concepts like neural networks and clustering to work in its tools (see „Portfolio Compression“). The challenge with machine learning is the combination of mathematical/IT-technical methods with the specific expertise in the respective field.
ROKOCO is currently engaged in research and development projects about using artificial intelligence in the field of insurance and capital markets. Here we work together with reputable insurance companies from Germany.