ALM study
The purpose of Asset Liability Modelling, or ALM, is to find the combination of financial assets best matched to the Fund’s future obligations (pension disbursements) in the pension system. The analysis horizon is therefore very long, up to 50 years.
The Fund’s ALM study is based on
- the overall mission for asset management assigned by the Swedish Parliament – for the Fund to maximize return with a low level of risk
- the investment rules and other guidelines dictated by the Swedish Parliament
- information about future inflows and outflows in the pension system over time (yearly forecasts provided by the Swedish Social Insurance Agency)
- expectations about the future returns and risks of different assets (based partly on the Fund’s own assessments and historical data)
Model links together development in the Swedish economy, capital markets and the pension system
Using financial theory, the Fund has created a model that links together development in the Swedish economy, the capital markets and the pension system. With the help of this model the Fund can study how its investment assets and the pension system are affected by projected economic trends given a certain asset allocation in the Fund’s portfolio.
The Fund’s task
The Fund sees it as its task to minimize the risk for, and extent of, future automatic balancing of the pension system. Preferably, this risk should be eliminated entirely.
The model developed by the Fund is used to calculate a number of key variables related to balancing risk, such as
- the probability of automatic balancing in a given year
- average expected balancing over the next 50 years (known by the Fund as balance loss)
- the extent of balancing in a worst-case scenario
Key variables are calculated for a large number of asset mixes (and a large number of possible economic trends). More than 5,000 different scenarios are simulated to find the combination of assets with the best key variables.
Sensitivity analysis
In order to test possible outcomes for the portfolio with the best key variables, the Fund carries out an extensive sensitivity analysis in which the assumptions in the model are altered. Only the portfolio that both produces the best outcome in the Swedish Social Insurance Agency’s base scenario and has positive outcomes in several other scenarios will be chosen as the strategic benchmark.