Andreas Nordvall Lagerås: Asset Liability Management for a Large Insurance Company

The asset liability management and choice of strategic asset allocation at a large insurance company entails several theoretical and practical challenges. I will describe some of these which I think may be of interest for both the theoretically minded practitioner and the researcher in financial mathematics.

* Complex objectives and constraints apply when optimizing an asset portfolio with regard to liabilities and regulations.

* Financial time series: heavy-tailedness and non-stationarity require robust and/or Bayesian methods.

* The risk premia of different assets are both the hardest and the most important parameters to estimate. On a related note, we find that "real world" models, that are of use to us, are often not as sophisticated as the "risk neutral" ones used in derivatives pricing.

* Size: some instruments are simply not available in large enough quantities to hedge some liabilities or risks. Being large also means that any change in your portfolio may affect prices even in normally liquid markets.

* There are risks with the liabilities that are not diminished by having a large collective.