Disclaimer: Views in this blog do not promote, and are not directly connected to any Legal & General Investment Management (LGIM) product or service. Views are from a range of LGIM investment professionals and do not necessarily reflect the views of LGIM. For investment professionals only.
Dynamic discount rates: wider benefits for risk management?
A dynamic discount rate can help trustees better reflect the true risk of a scheme and make improved decisions about other risks, such as how much longevity risk to hedge. However, these enhanced decisions do not arise for quite the reason you might expect.
DB pension schemes are now very well-funded on average, sharpening their focus on their endgame options and the associated risk management. As part of this, we believe that adopting a dynamic discount rate (DDR), that’s sensitive to the level of credit spreads, can often make sense.
A key potential benefit[1] of a DDR is that it can better reflect the true health of a pension scheme. The idea is that only downgrades and defaults in the matching portfolio (a CDI strategy) should impact the funding position, as opposed to any movements in spreads that don’t impact the ultimate cashflows generated.
Moreover, there are potential secondary benefits of a DDR, as Adam Boyes and Alasdair Macdonald of WTW recently pointed out with a longevity hedging example. Let’s now take this concept further by calculating optimal longevity hedge ratios, and showing how and why they are impacted by a DDR.
Mind the costs
The decision to hedge longevity risk of pensioners via a longevity swap needs to reflect a potential cost that reflects profit margins and prudence factored into insurers’ pricing. There could be roughly a 5% total cost[2] in respect of pensioners, which might work out at around 0.5% per annum.
One way to approach this decision is to consider that choosing not to hedge longevity risk could be viewed as both a potential source of risk and a source of return, in the sense that you save on some costs. In our view, this risk-return trade-off should be balanced alongside the other trade-offs available to trustees, such as investing in return-seeking assets. For the purposes of our analysis, we’ve assumed longevity is an uncorrelated risk.
A strategic longevity hedge ratio
For a given overall net return target, we define the ‘strategic longevity hedge ratio’ as the proportion of funded liabilities to hedge that minimises funding level volatility.
The figure below shows how risk changes with the longevity hedge ratio for an example scheme targeting a net return of cash plus 0.8%.
At higher hedge ratios, the scheme has less longevity uncertainty but incurs higher hedging costs, so must typically hold more in return-seeking assets to maintain its cash-plus-0.8% target. The strategic longevity hedge ratio of about 40% in this example represents the ‘sweet spot’ that minimises overall risk for the return target.
We repeated the optimisation at other return targets, shown below:
As you can see, the strategic longevity hedge ratio drops as the return target increases and the optimal strategy maintains a balance between the two potential sources of return.
Switching on dynamic discount rates
The above assumes a credit-insensitive discount rate, but what if we switch to a DDR? We see the strategic longevity hedge ratios leap up:
Why is this? Using a DDR helps recognise CDI as an efficient way to target excess return. This recognition results in tilting towards it and away from the other potential return source: under-hedging longevity risk. The upshot is a higher longevity hedge ratio.
Don’t feel marginalised
Often the decision to hedge less in the presence of other risks is not well-motivated, in our view, as it could lead to poor decision-making. An argument is made that with other risks at play, the incremental impact from the unhedged volatility in question becomes smaller. For example, statistics tells us that introducing an uncorrelated risk of 3% to an existing risk of 4% leads to an overall risk of 5%[3] i.e. an increase of only one percentage point. The 4% risk ‘masks’ much of the 3% risk. This is true, but we are keen to point out that this does not mean it matters less!
The trouble is that not all increments are made equal. To explain by analogy, imagine you have a fortune of £1 million and you then suffer a loss of 1%. That drops your fortune to £990,000 which isn’t a big deal for your lifestyle. But compare that with if you’ve already lost 98% of your fortune and face the prospect of losing another 1%. That drops your remaining fortune from £20,000 to £10,000 – a far less trivial impact for you.
For risk-averse investors, increases in potential losses matter more if they’re already starting from a greater potential loss. Indeed, a jump from 4% to 5% volatility should be considered to be broadly as painful[4] (all else equal) as a jump from 0% to 3%.
As such, the fact that longevity risk becoming less ‘masked’ when there’s a DDR isn’t actually a good reason to hedge more of it in our view. Rather, the real reason schemes may wish to hedge more longevity risk is if they now recognise CDI as a relatively efficient way of targeting return, compared with saving on longevity hedging costs.
As we’ve seen, holistic modelling can help here, to ensure trustees are managing risks effectively.
If you’ve enjoyed this blog post, please click here to discover more of our content that’s specifically tailored for DB schemes considering their endgame options.
You can also sign up here to receive the latest content most relevant for you via email.
For professional investors only. Capital at risk
Appendix – key model assumptions:
Return on longevity swap |
-0.5% p.a. |
Longevity risk |
3.0% p.a. |
Scheme duration |
10 years (all pensioners for simplicity) |
Correlation between longevity risk and investment risk |
0% |
Efficiency of CDI expressed as a Sharpe ratio of the FL return to FL volatility |
0.30 without a DDR 0.50 with a DDR |
Assumptions, opinions and estimates are provided for illustrative purposes only. There is no guarantee that any forecasts made will come to pass.
[1] Watch this space for a blog from my colleague Ian Blake who will take a deep dive into the construction of dynamic discount rates.
[2] This is purely for illustration - 5% is only a rule of thumb and the cost may have fallen recently. In general, the percentage will vary based on market conditions, specific terms negotiated and the scheme’s unique circumstances.
[3] Sqrt(3%^2 + 4%^2)
[4]Note that 3%^2 = 5%^2 – 4%^2.