Estimated reading time: 10 minutes
Longevity (how long members are expected to live for in retirement) is a topic of significant importance to defined benefit pension schemes and their actuaries. Pensioners living for longer means pensions being paid for longer, and hence higher overall costs.
What information might we be able to glean from the latest longevity research and what does it all mean for LGPS Funds?
2020: a very unusual year for mortality
As shown in the chart below, the Covid-19 pandemic has led to a sharp increase in reported deaths in the general population, with the total number of deaths in 2020 being significantly higher than the total number of deaths reported in previous years.
Source: Office for National Statsitics data for England & Wales
There were around 73,000 more deaths in the UK from the start of the pandemic to 1 January 2021 than if mortality rates (i.e. the probability of dying in a particular year) were similar to those experienced in 2019.
Mortality rates in England and Wales in 2020 were estimated to have increased by 11.8% over the year. For historical context, the last year annual mortality rates increased by this much was in 1929, a time when annual mortality was much more volatile than it has been in the last c. 40 years and before the establishment of the NHS and the significant recent medical advancements that have been made since World War II.
How will this mortality affect pension schemes?
Higher mortality can affect pension scheme liabilities in three specific ways:
- An immediate reduction (e.g. where a lower spouse’s pension becomes payable rather than the member’s pension) or removal (where there is no dependant’s pension payable) of liabilities where a member has died. This is a “direct impact” as it has already materialised in higher numbers of member deaths, and the abnormal mortality observed in 2020 is an example of this.
- A less optimistic view on current mortality rates (the base table mortality assumption), leading to a lower estimated cost of liabilities. This can be thought of as a step change in our view of mortality rates today, and is an “indirect impact” as we expect more members to now die at younger ages than our previous expectation.
- A less optimistic view on future mortality rates (the mortality improvement assumption), again leading to a lower estimated cost of liabilities. This is also an “indirect impact”.
Direct impact of the pandemic
The high number of excess deaths experienced over the last year is likely to directly translate into reduced liabilities for pension schemes. However, for LGPS Funds, this is not expected to be a significant reduction.
As an indication of the possible average direct impact the increased deaths in 2020 may have on LGPS liabilities, if the 73,000 increased deaths in 2020 were spread evenly across the c. 12m pensioner population in the UK, this would lead to an average fall in pensioner liabilities of c. 0.6%. This will of course vary by Fund and certainly will vary on an employer level.
Indirect impact of the pandemic
Although the number of actual deaths as a result of the pandemic will be known, the long term indirect impacts on mortality remain very uncertain. There are a number of competing (and often interlinking) factors that will affect future mortality rates as a result of the pandemic. These include:
Positive impact on life expectancy (i.e. lower future mortality rates)
Negative impact on life expectancy (i.e. higher future mortality rates)
|The Government’s vaccination programme of targeting individuals in the highest risk groups has so far been successful, and no vaccine resistant strains have been identified.||There may be further waves of Covid-19 as lockdown restrictions are eased, or due to seasonal reoccurrences of the virus; this could lead to a greater reduction in liability values for the average pension scheme.|
|It could be argued that the pandemic has simply brought forward the deaths of those in poor health that were likely to have died over the next few years. In this case, mortality over the next 2 or 3 years will be lower than it otherwise would have been.||During the pandemic, many surgeries were postponed or cancelled and it is likely that a significant number of diagnoses for certain diseases (e.g. cancer) will have been missed. This will likely lead to an increase in future mortality rates.|
|The population surviving following the pandemic may be “stronger” (i.e. be expected to live longer) on average than the population prior to the pandemic, if the abnormally high deaths observed during the pandemic have occurred more significantly in the frailer parts of the population.||The increased public spending during the pandemic and the impact of an economic contraction (likely reducing tax revenue for the government) is likely to result in a reduction in future healthcare spending. This may lead to a deterioration in future health outcomes and an increase in mortality rates.|
|There may be long-term public health benefits; for example if washing hands or wearing a face mask in public become regular habits.||The virus could lead to long-term health implications as people may be frailer due to organ damage incurred while suffering from the virus. There are also indirect health consequences due to the pandemic; for example the impact on mental health as a result of successive lockdowns or reduced physical activity due to the closure of gyms.|
There are also factors where the impact on future life expectancy is uncertain (i.e. could be positive or negative). For example, it is unclear whether influenza in the 2021/22 winter period will be worse (due to the influenza vaccine being ineffective as a result of low levels of influenza in the southern hemisphere, where the influenza vaccine is first developed for the northern hemisphere) or better (due to better public health, such as washing hands and wearing masks in public) than a “normal” year.
In the context of this significant uncertainty, our view is that over the short and medium term it is likely that mortality experience will continue to be heavier as a result of the pandemic. The short term view is based on having already seen excess deaths continue since the start of 2021. In the medium term (2-10 years), mortality rates could be expected to be higher (than would otherwise have been the case) due to the effects of economic contraction and the long-term health implications of lockdowns.
Modelling future mortality improvements and CMI_2020
The Continuous Mortality Investigation (CMI) provides a model that can be used to project future mortality rates. An annual update of the model is produced, updating each year for the latest mortality data in the national population of England & Wales. CMI_2020 is the latest model (incorporating data to the end of 2020) and was published in March 2021.
The underlying principle of this model is to smooth recent rates of mortality to determine an initial level of mortality improvements and then to project these initial rates into the future, based on a range of different input assumptions (the long-term rate of mortality improvement is an example of one of these parameters).
This smoothing is appropriate in “normal” years, when the change in mortality between successive years is relatively low and a relatively smooth trend can be fitted to historical data.
As a result of the extremely abnormal mortality observed in 2020, the CMI model does not fit well to the data. There is of course the question as to how much weight we place on the abnormal data in 2020, and how much it should be considered a “blip”.
Because of this, the CMI introduced a “2020 weight parameter” into CMI_2020. This parameter allows a user to vary how much significance (or weight) to place on the mortality data in 2020, so that the exceptional mortality experienced due to the coronavirus pandemic can be incorporated without having a disproportionate impact on results. Placing more weight on the abnormally high mortality data observed in 2020 serves to reduce life expectancy (i.e. increase mortality rates). However, the impact of the 2020 weight parameter on future mortality improvements “dissipates” over time, with the effect completely disappearing by 2040.
The CMI has confirmed the core (i.e. default) value of this 2020 weight parameter will be 0%, which means that no account will be taken for data in 2020 during the model fitting process. Compared to the previous version of the model (CMI_2019), which was published prior to the pandemic, adopting this default version of CMI_2020 will lead to a fall in life expectancy (and therefore liabilities) of c. 0.2%.
However, the CMI has encouraged users to consider the 2020 weight parameter in detail before adopting a certain value, and not to take the core value as the CMI’s “recommendation”. In particular, if a user’s view is that the pandemic will detrimentally affect life expectancy by more than 0.2% (due to adopting the default version of CMI_2020) then this can be reflected by increasing the 2020 weight parameter adopted.
"As detailed above, in our view the overall outlook for future mortality improvements looks more negative than implied by the default version of CMI_2020, with the adverse consequences of the pandemic seeming to outweigh the positive ones."
We discuss below our recommendations for the adoption of this parameter.
However, we reiterate that the situation remains uncertain, with new information emerging regularly. As a result, we recommend that the 2020 weight parameter adopted (and indeed, the other parameters in the model) is kept under review.
Although data analysis has already completely revolutionised our everyday retail journeys, in the workplace many key decisions are still based purely on anecdotal evidence or instinct alone. Slowly but surely, the juggernaut is turning and analytics is on the rise in the workplace. Indeed, we are heading towards a new destination where Employer DNA will deliver sustainable, robust and innovative strategies.
"We are heading towards a new destination where Employer DNA will deliver sustainable, robust and innovative strategies."
I started by saying that “DNA is the very material that defines our uniqueness – the very substance that carries the information we need to survive and to thrive.” The same is true for Employer DNA. As you work your way up the rungs of the data analytics ladder, the closer you get to the top, the more you’ll realise that your Employer DNA really does contains the insights you need to both survive and to thrive.
What does this all mean for your numbers?
Employer accounting disclosures
Overall, based on the data currently available, we believe it’s appropriate to reflect the view of a deterioration of future life expectancy by making a specific adjustment to the “best estimate” (as required under the IAS19 and FRS102 accounting standards) future mortality improvement assumption.
As discussed above, we believe that the heavy mortality experienced in 2020 and change in future outlook (compared to pre-pandemic expectations), should be captured in the mortality assumption to some extent.
"Based on the reasons set out above, we believe a 2020 weight parameter of 25% would be an appropriate assumption for use in the accounting disclosures."
The magnitude of the impact of the reasons given above for the deterioration in future life expectancy cannot be easily broken down by any scheme-specific characteristics (e.g. location or average socio-economic class). We are therefore recommending that all employers adopt the new CMI_2020 model with a 2020 weight parameter of 25% for the March 2021 accounting valuations.
At this time, we do not believe that the pandemic will affect the other parameters adopted in CMI_2020, and we have captured the expected post-pandemic impacts in the 25% 2020 weight parameter. For English funds, we therefore recommend that all other parameters remain unchanged from those adopted for last year’s accounting disclosures. For Scottish funds, the mortality assumptions will be updated in line with the assumptions adopted for the 2020 funding valuation, with CMI_2020 and a 25% weight parameter adopted in addition.
This update in assumption is expected to reduce future improvements in life expectancies and reduce the value of liabilities. In terms of accounting liabilities, the defined benefit obligation could reduce by between 0.5% and 1.5% as a result, depending on the characteristics of the membership (e.g. gender and age profile) and the other elements of the mortality assumptions adopted.
With the next triennial valuation for English and Welsh Funds scheduled for 2022, we will be monitoring the situation and will carry out a review of the mortality assumptions for the LGPS funds in advance of the 2022 valuation.
If the valuation were scheduled at a current date, we would recommend adopting CMI_2020 with a 25% weight parameter. However, as noted above, the situation remains uncertain and it is likely that our understanding of the future impacts of the pandemic will change before the 2022 valuation date. (Indeed, it may be that a later version of the CMI model, such as CMI_2021, is available by the time of this valuation.)
As yet, it is too early to say what parameters will be adopted as part of the 2022 valuation and there will likely be more information available at the time to allow us to make a judgement on how best to incorporate recent experience into our assumptions.
If your LGPS fund would like to understand the effect of recent mortality experience on their liabilities; we can carry out an analysis in advance of the 2022 valuation. Please get in touch with your usual Barnett Waddingham contact or myself below so that we can confirm data requirements.
If you would like to talk about this topic, please get in touch with your usual Barnett Waddingham contact to find out how we can support you. Alternatively, please contact the authors below or Ross McBride, a key contributor to this blog.
Stay up to date
Get the latest independent commentary and exclusive insights from a range of experts at the forefront of pensions, investment, insurance and risk – tailored to your preference.Subscribe today