Gareth Connolly

Gareth Connolly

Scheme Actuary

Willis Towers Watson

Towers Watson Media

Simon Eagle

Scheme Actuary

Willis Towers Watson

Preliminary results have been considered for many ’tranche 10’ valuations – those with effective dates in the year to 21 September 2015 - and some have already been signed off. Gareth Connolly and Simon Eagle take a first look at what the early indications show.

Discount rate approach – more schemes are planning for investment

The ‘dual discount rate’ approach has ceased to be the most popular for the first time.

The initial indications show a continued shift in the way that discount rates are set - the ‘dual discount rate’ approach (explained below) has ceased to be the most popular for the first time, and has been overtaken by the ‘term dependent’ approach in tranche 10. An analysis of discount rate approaches used over the last 10 years, based on a survey of over 1,100 valuations, is shown in Figure 01.

Figure 01. Discount rate approach: Trend over time

Towers Watson Media

What is the ‘dual discount rate’ approach? Let us explain:

  • For many years this has been the dominant approach – one discount rate is used for the period before a member’s retirement, and another is used for the period after retirement.
  • The post-retirement discount rate is usually lower, representing the return on the assets the scheme might hold in the very long term when all the liabilities relate to pensions in payment.
  • So this approach builds in an allowance for de-risking over time because as more members retire, and others get closer to retirement, more weight is placed on the (lower) post-retirement discount rate.
  • However, the approach can be unclear as to how quickly the scheme is being assumed to de-risk, and how that might compare with a de-risking plan or any views on longer-term covenant strength.

Alternatively, a term-dependent approach can be used – this is where a discount rate is set for each future year, for example based on the gilt yield plus an assumed pattern of outperformance over gilts which varies over time. This allows trustees and sponsors to set a discount rate allowing explicitly for any planned de-risking in any given year, and is usually entirely transparent around the rate of assumed de-risking.

The term-dependent approach may have become more popular due to an increased number of schemes putting in place de-risking plans as they mature, or from a desire for greater transparency.

Recovery plans – use of flexibilities is increasing

We noted in our April 2015 article on early predictions for tranche 10 valuations that, despite material asset returns over the last three years, many schemes’ tranche 10 valuations will show an increased deficit compared with the tranche 7 valuation due to material reductions in gilt yields. The additional deficits will need to be met by increased annual contributions, a longer recovery period, additional assumed asset returns, or a combination of these.

Initial indications of average recovery periods are shown in Figure 02.

Figure 02. Recovery periods

Towers Watson Media

This indicates that for tranche 10 valuations the average new recovery period will end 3.6 years after the date on which the average tranche 7 recovery period ended. This is very close to the 3.5 years median extension estimated in the Regulator’s May 2015 analysis for the scenario where annual deficit contributions were maintained at their current rate. This is also strikingly longer than the average recovery period for tranche 9 valuations.

Initial indications of the proportion of valuations allowing for asset outperformance in the recovery plan, as described in our 6 October article, are shown in Figure 03.

Figure 03. Proportion of recovery plans where allowance made for asset outperformance

Towers Watson Media

This indicates continued increases in the proportion of schemes making allowance for asset outperformance (compared to the discount rate) over the recovery plan, which reduces contribution requirements.

Many trustees and sponsors may be aiming to avoid the newly disclosed deficit leading to increased annual contributions.

Figures 02 and 03 imply that many trustees and sponsors may be aiming to avoid the newly disclosed deficit leading to increased annual contributions. This may be with regard to the Regulator’s 2014 updated funding Code of Practice suggesting use of the flexibilities of the funding regime to ensure recovery plans are appropriately tailored to scheme and employer circumstances.

Mortality assumptions – improvements in funding positions

Another striking feature of the initial survey data relates to mortality assumptions. Figure 04 illustrates the trend in average life expectancies over the last ten years.

Figure 04. Mortality assumptions - implied life expectancies: Trend over time of average

Towers Watson Media

This indicates that the small fall in life expectancy experienced in tranche 9, reversing the previous trend as reported in our July article, has levelled off in tranche 10 for males and has partly reverted for females. All tranche 9 and 10 valuations in the survey used the CMI model for assumed future improvements in mortality. The CMI model was updated in September 2015 and showed decreased life expectancies for males and females due to heavy UK mortality rates in early 2015. The new CMI model will not have been reflected in many of the initial results shown above, therefore a later cut of data for tranche 10 valuations may show a further reduction in life expectancies.

The average tranche 10 life expectancies valuations shown are lower than for tranche 7, and so many tranche 10 valuations are likely to show a funding improvement from changing mortality assumptions.

Closing remarks

  • Scheme funding trends can be useful information for trustees and sponsors when making decisions for their funding valuation.
  • Scheme funding for T10 valuations remain challenging in the current climate, with funding positions often worse than at the last valuation.
  • There is evidence of schemes making changes to funding approaches due to changes in circumstances and in reaction to the Regulator’s updated Code of Practice.