Cardano on the future of pension fund liabilities

No comments

Kerrin Rosenberg and Theo Kocken draw startling conclusions about the future shape of pension fund liabilities

As our pension system enters the payment phase over the coming decade, funding ratios will suffer great instability. The greatest threat is not the current underfunding problem; instead, it is the substantial outflow of money that lies in store for the funds. That outflow will render a fund management strategy based on expected returns ineffective; strategies based on contributions to control the funding ratio have already proven ineffective, given the enormous size of the pension fund in relation to most parent companies.

It is crucial to understand this instability in order to increase support for amending pension contracts, thereby ensuring the defined benefit system can be retained in the Netherlands. It also has implications for whether the popular UK concept of 'dynamic de-risking' will work or not, and for pension fund risk management generally.

Many in the pensions industry are aware that funding ratios will deteriorate rapidly in a seriously underfunded system that pays out a significant percentage of its liabilities to retired members each year. Even if returns are high, funding ratios will inevitably plummet if retired members continue to receive payment of their full nominal entitlements.

However, it is often thought that as long-term investors, pension funds are not really substantially underfunded. The expected risk premium on investments like equities is seen as a way of ensuring that funding ratios are raised to sufficient levels and stay there.

That strategy could succeed provided that at least one of the following conditions is fulfilled: that the markets are not volatile around the (high expected) average at any time; or that the funds remain in the accumulation phase, where incoming contributions outweigh benefit payments. Unfortunately, presently neither of these conditions is fulfilled for the majority of funds in the UK and the Netherlands.

This means that even if a fund is not currently underfunded and was guaranteed to earn a very high (albeit fluctuating) return over the next 20 years, it will still run substantial risks upon entering the payment phase.

In order to illustrate that instability, we examine the effect of two 20-year scenarios for a fully-funded pension fund (funding ratio: 100%), in both cases assuming an (arithmetic) average return of 8% on the higher-risk investments. These higher-risk investments constitute 50% of the investment portfolio; the rest matches the liabilities.

Using these two scenarios, a comparison has been made of the effects on a young fund (typical for the years 1960-80), and a more mature fund (typical for the coming decades).
The first scenario (good early returns) represents in chronological order what happened during the period 1989-2009, when there was a cumulative return of 8% per annum on equities.

The second scenario (good later returns) basically contains the same returns, but four annual returns have been swapped. Ultimately this has no effect on the money saved - in both scenarios every pound or euro invested trebles over 20 years.
The results of these two healthy long-term scenarios on healthy pension funds with clearly distinct levels of maturity is interesting; in fact, it is almost paradoxical.

Figure 2 shows that the funding ratio of the young fund (which has a long liability duration of 28 years and relatively few payments in the first years) develops almost in line with the investment index, although with lower growth since the liabilities also rise in line with the discount rate. Because this immature fund does make some payments, differences do arise after 20 years. Yet, regardless of the scenario, the young fund has a healthy level of overfunding; the funding ratio is between 110 and 150 at the end of the period.

When a fund enters the payment phase the picture is totally different, as figure 3 shows (duration of liabilities is 13 years, with substantial payments in the first few years). Given exactly the same two scenarios, there is much greater divergence. In scenario 1, with good returns in the initial phase and shocks somewhat later, the funding ratio rises, and this effect is reinforced by the payments - an increasing amount of capital therefore remains, to the delight of the remaining small group.

In scenario 2 there are setbacks at the start - the funding ratio plummets and the pension fund has to fill the gap with inadequate capital at its disposal. In this phase, very high returns later in the 20-year period are not enough to save the fund from depletion; by 2027 its coffers are completely empty, despite having had a funding ratio of 100% in 2010.

So two funds having the same returns over 20 years produce totally different results. In practice, there is a substantial risk that the consequences could be dramatic. This cannot be considered a solid foundation for a pension system in a period of population ageing.

In conclusion, the research highlights the instability of more mature pension funds. Once our pension system enters the payment phase, it can no longer fall back on the old argument of relying on the long term. It is therefore important to focus more attention on risk management today, rather than dynamic de-risking at some uncertain point in the future.

In addition, where possible, there may be a need to amend the design of pension funds quickly to mitigate this instability. Giving guarantees to all members, and therefore paying out 100% even during periods of underfunding, creates excessive uncertainty about the pensions of younger members. Similarly, where there is underfunding, taking a long-term, wait-and-see approach with the prospect that one day liabilities might have to be cut substantially for everyone, creates excessive uncertainty among retirees.

Everybody will benefit from a solution that is clearer and flexible, since guarantees cannot be given and the great instability highlighted here is in danger of leading to uncertainty and an unfair distribution between generations.