Andreas Mitschele

A multi-objective approach to integrated risk management

The integrated management of financial risks represents one of the main challenges in contemporary banking business. Deviating from a rather silo-based approach to risk management banks put increasing efforts into aggregating risks across different risk types and also across different business units to obtain an overall risk picture and to manage risk and return on a consolidated level. Up to now no state-of-the-art approach to fulfill this task has emerged yet. Risk managers struggle with a number of important issues including unstable and weakly founded correlation assumptions, inconsistent risk metrics and differing time horizons for the different risk types. In this contribution we present a novel approach that overcomes parts of these unresolved issues. By defining a multi-objective optimization problem we avoid the main drawback of other approaches which try to aggregate different risk metrics that do not fit together. A MOEA is a natural choice in our multi-objective context since some common real-world objective functions in risk management are non-linear and non-convex. To illustrate the use of a MOEA, we apply the NSGA-II to a sample real-world instance of our multi-objective problem. The presented approach is flexible with respect to modifications and extensions concerning real-world risk measurement methodologies, correlation assumptions, different time horizons and additional risk types.

Kernel Methods in Finance

Kernel methods (Cristianini and Shawe-Taylor 2000; Herbrich 2002; Scholkopf and Smola 2002; Shawe-Taylor and Cristianini 2004) can be regarded as machine learning techniques which are “kernelised” versions of other fundamental machine learning methods. The latter include traditional methods for linear dimensionality reduction such as principal component analysis (PCA) (Jolliffe 1986), methods for linear regression and methods for linear classification such as linear support vector machines (Cristianini and Shawe-Taylor 2000; Boser et al. 1992; Vapnik 2006b). For all these methods corresponding “kernel versions” have been developed which can turn them into non-linear methods. Kernel methods are very powerful, precise tools that open the door to a large variety of complex non-linear tasks which previously were beyond the horizon of feasibility, or could not appropriately be analysed with traditional machine learning techniques. However, with kernelisation come a number of new tasks and challenges that need to be addressed and considered. For example, for each application of a kernel method a suitable kernel and associated kernel parameters have to be selected. Also, high-dimensional nonlinear data can be extremely complex and can feature counter-intuitive pitfalls (Verleysen and Francois 2005).

Integrated Risk Management: Risk Aggregation and Allocation Using Intelligent Systems

In recent years integrated approaches have become state-of-the-art practice for risk management in financial institutions. Contrary to the still common silo-based approach where risk categories and business lines are predominantly analyzed separately, an integrated risk management system adopts an enterprisewide perspective to appropriately account for cross-sectional dependencies between all significant banking risks. In this contribution an application of intelligent systems that provides management with risk-return efficient bank-wide asset allocation strategies is outlined. It is based on multi-objective evolutionary algorithms and considers different market risks and credit risk as well as position volume constraints. The presented novel approach is not only able to integrate the differing goals concerning the risk management function but also to partly overcome the obstacles for risk integration and aggregation. Using real market data a sample portfolio analysis is performed and possible conclusions for a bank risk manager are drawn. The approach is extendable concerning for instance advanced risk measurement methodologies, correlation assumptions, different time horizons and additional risk types. Further real-world constraints, such as regulatory capital, portfolio or P&L restrictions can also be easily integrated into the model.

Comparison of multi-objective evolutionary algorithms in optimizing combinations of reinsurance contracts

Our paper concerns optimal combinations of different types of reinsurance contracts. We introduce a novel approach based on the Mean-Variance-Criterion to solve this task. Two state-of-the-art MOEAs are used to perform an optimization of yet unresolved problem instances. In addition to that, we focus on finding a dense set of solutions to derive analogies to theoretic results of easier problem instances.

Heuristic Optimization of Reinsurance Programs and Implications for Reinsurance Buyers

Reinsurance contracts represent a very important tool for insurance companies to manage their risk portfolio. In general, they are used if an insurer is not willing or not able to hold certain risk exposures or parts thereof on its own. There exist two main contract types to cede claims to a reinsurer, namely proportional and non-proportional ones. With the quota share reinsurance, a well-known variant of the former ones, a fixed percentage of the claim sizes is ceded to the reinsurance company. Excess of loss and stop loss are non-proportional types and the reinsurer is only liable to pay if certain losses are exceeded. In practice insurance companies usually place a number of different reinsurance contracts, a so-called reinsurance program.

Intelligente Methoden im Integrierten Risikomanagement

In dieser Arbeit wird ein multikriterielles Modell zur Integration des Risikomanagements auf Basis von Kredit-, Markt- und operationellem Risiko konzipiert. Der Ansatz approximiert die Losungen des Problems mittels multikriterieller evolutionarer Algorithmen. Seine Anwendung wird fur eine Beispielbank aufgezeigt mit besonderem Fokus auf die ansprechende Visualisierung der Ergebnisse.