ALM - Tillgång/skuldmodell för riskberäkning och portföljoptimering

Sammanfattning: Asset management in insurance companies differs from conventional asset management to the extent that respect has to be taken to both assets and the commitments the insurance company has towards its customers. A model that has proven to fit well regarding the matching of assets and liabilities is the Asset Liability Management Model (ALM model). In addition to the matching in the balance sheet, the ALM model can be used in a company's work with strategic portfolio allocation by applying it as a basis for analyzing investment strategies with expected risk and return. From this, the ALM model also becomes relevant for calculating key figures according to the legal framework Solvens II which includes laws and regulations regarding the demands on economical strength (solvens) of insurance companies. Hence, the goal of this masters thesis has been to, on behalf of Bliwa Livförsäkring, create an ALM model to support the asset management department of Bliwa in their work with defining a credible way of analyzing the future risk and return of Bliwa's asset portfolio and insurance undertakings. The ALM model generally consist of four submodels, the scenario model, the liability model, the asset model and the company model, where the scenario model often is named as the core of the ALM model. The course of action has been to develop these submodels individually, with focus on the scenario model. The purpose of the scenario model is to generate credible future economical scenarios in the form of stock development and interest rate curves based on historical market data. In the asset and liability models, asset and liabilities are valued from these scenarios. The function of the company model is thereafter to compare assets and liabilities by calculating key figures and analyzing investment strategies. The resulting ALM model has thereafter been used to study different investment strategies and their impact on key figures and the future portfolio development with respect to current legal frameworks in combination with assets and liabilities. The final version of the created ALM model is by the authors considered to have reached its purpose with some limits. These limits are found in the scenario model and are partly based on the unique interest rate situation that prevailed during the writing of this masters thesis and partly on the limited amount of documentation regarding chosen mathematical model. Apart from these limitations, both the authors and the supervisor at Bliwa, Andreas Johansson, consider the ALM model as a whole to function in a correct manner and will be of assistance to the asset management department at Bliwa Livförsäkring.