Epsilon Archive for Student Projects

Abstract

insurance companies within the EU. It has had extensive criticism before, during, and after its introduction. This paper focuses on the Nordic equity markets and studies how the standard formula compares to other regulator-accepted methods of risk quantification, known as internal models. The Nordic equity market is quite different when compared to European market indices, the latter being the main focus for earlier studies. This paper uses data from listed Nordic equities, listed private equity companies and of unlisted Nordic equities from the Aland mutual insurance company. Data of the latter kind is rarely publicly available for research. The chosen methods in the internal models are Value-at-Risk and Monte Carlo simulations. In the Monte Carlo simulations, a stochastic drift in the form of a geometric Brownian motion is used. The results indicate that the difference of using Value-at-Risk instead of the standard formula is not too large, when only investigating the solvency capital requirements. They produce results in close vicinity of each other. However, the standard formula may force the insurer to choose inefficient portfolios, which is against modern portfolio theory. Monte Carlo simulations will indicate significantly lower solvency capital requirements, close to half of the standard formula. There will be prevailing uncertainty whether the historical average return, which is used as the deterministic drift, will be an adequate estimate of future expected return. Given this critique, the Monte Carlo simulations may be biased upwards.