Requirements for an economic scenario generator

Overview

The purpose of this technical principle is to draw the attention of actuaries working in Germany to what they should bear in mind when using capital market models.

Capital market models are generally used to project capital markets or investments to be found there or risk factors to be derived from them into the future over time. In the actuarial context, there are basically two different paradigms and areas of application for this projection (for more precise definitions, please refer to the underlying main sections of this document):

• Realistic (“real-world”) projection of the capital market
• Risk-neutral[1] Projection of the capital market

The central topics for both applications, which we will discuss in more detail later, are the following questions:

• Which mathematical models are fundamentally suitable for modeling the individual capital market instruments and risk factors?
• How can these models be calibrated, i.e. how should the parameters of the models be set?
• How can these models and the resulting simulations be validated?

This note is therefore structured as follows:

First, meta-topics of capital market modeling are discussed in section II. The aim is to deal with overarching topics of modeling, calibration, data and validation and to address introductory issues relevant to the other sections. In particular, it also deals with calibration issues in inactive market segments.

Sections III and IV then deal with separate corresponding modeling, calibration and validation topics in the context of real-world and risk-neutral modeling.

Adoption

This note came into force with its adoption by the DAV Executive Board on 27.11.2023 and replaces the editorially revised note of the same name dated 01.08.2017. Furthermore, this note includes the contents of the former DAV technical principle “Calibration in inactive market segments” dated 04.12.2014, which is thus incorporated into this technical principle.