There is an increasing weight of discussion around the challenge of schemes being cashflow negative (i.e. pension schemes are net disinvestors as benefits are greater than sponsor contributions). This was specifically raised in the Pension Regulator’s Annual Funding Statement with subsequent research from Hymans Robertson highlighting this as a worry for CFOs.
The apparent risk of cashflow negativity is one of being a forced seller of assets (potentially at a bad price) in order to meet cashflow. One of the obvious responses to this concern is to refocus investment strategy (or part of it) towards income generating assets such that cashflow is met by income rather than disinvestments.
As is alluded to in this piece in Pensions Expert I think this approaches it the wrong way round. I think that cashflow negativity is simply a symptom of a maturing pension scheme and the real risk is that of pension schemes maturing and running out of time (as I have mentioned before).
To try and illustrate this principle I have done a quick bit of desktop analysis. I have taken a real pension scheme set of cashflows and contributions and modelled the required return to meet all benefits. I have then created a dummy, zero cashflow, pension scheme that only has one cashflow in 20 years’ time (broadly the average term of the real scheme) but has the same required return – i.e. one scheme has benefits to meet on a regular basis and the other is simply investing for 20 years with no cashflow requirement. In this example the required return comes out as 1.5% above Gilt returns.
I then look at how sensitive this required return is to shocks in the asset value at different points in the future. For this analysis I have used a shock of 10%. Chart 2 shows the result.
The first interesting observation from chart 2 is that the sensitivity of both schemes to an immediate asset shock is broadly the same.
The second observation is the deterioration of that sensitivity (i.e. the increase in return required as time goes on for the same shock) – the scheme with no cashflow requirements is actually more sensitive to asset shocks than the real pension scheme that is cashflow negative. In my simplified example it is obvious why this is the case – the real scheme has cashflows that go out for 100 years whereas the zero cashflow scheme has a fixed end date in 20 years. This means that for a shock in asset value in 15 years’ time, the zero-cashflow scheme only has 5 years to fix it whereas the average term of the real scheme is still 13 years so has longer on average to make up for the shock.
Extending this second point, also shown in chart 2 is the same analysis but where the zero-cashflow scheme has its payment in 30 years rather than 20. This illustrates the points quite well. Firstly with this zero-cashflow scheme it has a longer average term than the real scheme and therefore its immediate exposure to shocks is lower (as it has longer to recover from shocks). Secondly, for the same reason, that sensitivity does not increase as rapidly as in chart 1 as the scheme is maturing at a slower rate.
So pulling all of this together, this analysis indicates is that it is the maturity of the scheme that drives the exposure to asset shocks rather than the cashflow position of that scheme.
One of the solutions presented to this problem is using assets that pay income rather than rely on a scheme’s disinvestment to pay cashflow. For me, the desire to not disinvest is simply a way of avoiding market shocks rather than an economic need. Put another way – a trustee should be indifferent between an asset that pays a given income vs. one that has capital certainty. And, for me, where capital certainty is the challenge that maturing schemes are facing, that feels like a more appropriate response than focusing on short term income.