In this paper we investigate the effectiveness of the Minimum Funding Requirement (MFR) (pre-1999 version) under random future investment conditions. In particular, do the various liabilities calculated under the MFR deliver a suitable level of security to pension scheme members?

The paper considers active (and deferred) members and current pensioners separately.

For active members, Monte Carlo simulation is used to compare how the proceeds arising from investment of the pre-1999 MFR liability in an equity-backed personal pension would compare with the deferred pension promised by the scheme. It is found that, while there may be the intended 50% chance that the personal pension has a better outcome, there are very substantial downside risks for younger members. Investment strategies based upon index-linked gilts are shown to result, on average, in lower personal pensions, but they provide an extremely effective means of limiting the downside risks.

For pensioners, Monte Carlo simulation is used to find the distribution of the initial amount of assets required to pay off precisely the pension liabilities as they arise. This investigation considers the effectiveness of the 12-year rule and different investment strategies. The 12-year rule in combination with an equivalent investment strategy is shown to provide reasonable security for a typical group of pensioners. However, if investments are restricted to an appropriate mixture of gilts then the level of security for pensioners is much increased. In particular, we investigate the use of value-at-risk reserves. It is found that 95% value-at-risk reserves are lower for a mixed conventional/index-linked gilts strategy than a mixed gilts/equity strategy.

It is argued, because of the high degree of downside risk, that there is no place in MFR calculations for high, anticipated equity returns. Instead rates of interest should be based solely upon fixed-interest and index-linked gilt yields.