23 March 2016
It is hard to escape references to longevity risk at the moment. Last year saw over £20bn of longevity swap and bulk annuity transactions from pension schemes trying to manage it (including £12bn from buy-in and buy-outs alone). And the Pensions Regulator is increasing its focus on ensuring pension schemes carry out complete and robust assessment of their risks, with even an explicit reference to longevity risk in its December 2015 guidance on Integrated Risk Management.
But what exactly do we mean by “longevity risk”? And how do you know how much your scheme is running?
The majority of pension schemes haven’t yet sought to address these questions. For the few that have, most are likely to have been advised that their longevity risk “is about 5%” of their liabilities. Across the pensions industry, this standard answer of 5% rarely changes; whether your scheme has 100 members or 100,000 members, whether you are measuring the longevity risk over the next year or the next 20 years, or whether you are considering a 1-in-20 likelihood or a 1-in-200 likelihood. Can 5% really always be the right answer?
To help answer that question, it is useful look at the underlying components of longevity risk. In general, longevity can be categorised into three main risks:
- Base table risk – the risk that current rates of mortality (often determined by looking at recent scheme experience) have been mis-estimated;
- Improvements risk – the risk that future improvements in mortality rates have been mis-estimated; and
- Individual risk – the risk arising due to the timing of individual member deaths (eg a high liability member living longer than expected)
The size and relative weightings of these three risks can vary significantly across different types of pension scheme. For example, a scheme with very few members is likely to find it is exposed to a material amount of individual risk, improvement risk will tend to dominate for younger schemes and base table risk will be lowest for large pension schemes with lots of historical mortality experience data.
As well as the size of your scheme, your benefit structure can play a part too; a scheme with generous inflation-linked pension increases will be more exposed to longevity than a scheme with non-increasing pensions.
Also, longevity risk isn’t simply a short-term risk. Fundamental changes in longevity eg due to medical advances are likely to take several years to translate (with confidence) into increased pensions costs. That means that considering longevity risk over too short a time horizon can be misleading if you don’t also think about risks over the longer term – for example, if you have a 10-year plan to full funding or self-sufficiency, what risks are you exposed to over that journey?
Feedback is that LCP LifeAnalytics is the most robust tool available in the market for helping pension schemes to analyse how much longevity risk they are running and the most cost-effective ways of hedging it.
Using this tool we have seen longevity risk vary from as little as 2.5% of liabilities (for a large scheme looking at longevity risk over a one year basis) right up to 15% of liabilities (for a 100 member scheme considering longevity risk over the lifetime of their scheme). Only by getting a complete and bespoke picture about the risks you are running can you make truly informed decisions about how those risks should be managed.
My colleague LCP Investment Partner Ian Mills discusses how understanding longevity risk leads to better investment decisions in his blog - read it here.
Do you know how much longevity risk your pension scheme is running?
By answering three simple questions, you can get an overview of how longevity risk might look in your scheme - start here