This week we published the FAB index (FABI) which shows the way we see the state of this nation’s defined benefit pensions. “We” means First Actuarial, I work for First Actuarial.
We want to paint a picture of the pensions landscape. If you go to a gallery and look at a Turner or a Constable, you immediately recognise a unique sensibility. Turner looked at the sky one way, Constable another; the same sky -different eyes!
Two actuaries can look at a state of affairs and judge how it will turn out differently, because they have different assumptions. Some of those assumptions are minute- collectively, all the small assumptions can make a big difference- especially over time. But sometimes the assumptions are paradigmatic; meaning that two actuaries are seeing things from the opposite ends of a telescope.
The paradigm shift in the way that First Actuarial is asking you to look at the issue of defined benefit pension funding is away from a world where we want to stop paying pensions as soon as possible, to one whether we want to continue to support pensions over time. Those who want to stop paying people’s pensions are the trustees and sponsors of defined benefit schemes who want to buy out their liabilities (or see them transfer to the protection fund). Those who want to keep them open are trustees and employers who see a benefit in paying pensions to people to the very last payment due!
We are trying to encourage people to see the payment of pensions to the very last payment as being preferential; from a societal, economic and personal basis!
But there is a temptation among those in the opposite camp, to think that we are twisting the facts, frigging the numbers and hiding our assumptions to make our position look better than it really is.
Now you didn’t ask Turner or Constable to write down why he saw clouds the way he did but you can ask an actuary why he sees the future the way he or she does. So for all the people on this blog and for all the journalists and for the civil servants and for the actuaries in the opposite camp, here are the assumptions that lie behind the blue lines.
Just Fab -remember!
FAB INDEX–ASSUMPTIONS ; 18 OCTOBER 2016
Method
The FAB Index is calculated using publically available data underlying the PPF 7800 Index which aggregates the funding position of 5,945 UK defined benefit pension funds on a section 179 basis, together with data taken from The Purple Book, jointly published by the Pension Protection Fund (“PPF”) and the Pensions Regulator (“tPR”).
The liabilities are calculated by switching the data underlying the PPF 7800 Index onto a set of assumptions derived using First Actuarial’s in-house “best estimate” assumptions (see below), and adjusted to allow for full scheme benefits.
First Actuarial’s in-house “best-estimate” assumptions as at 30 September 2016
The “best estimate” assumptions as at 30 September 2016 used in the FAB Index are described below. All of the assumptions exclude any allowance for prudence.
The discount rate is set equal to the weighted average of the expected future investment return on the assets actually held by the 5,945 DB pension funds included in the PPF 7800 Index, using the average asset allocation published in The Purple Book as weights.
As at 30 September 2016, the weighted average discount rate was as follows:
| Average asset allocation in total assets
Source: Figure 7.2, The Purple Book 2015, PPF and tPR |
First Actuarial “best estimate” expected return as at 30 September 2016 | |
| Equities | 33.0% | 7.47% |
| Gilts and fixed interest | 47.7% | 1.53% |
| Insurance policies | 0.1% | 1.53% |
| Cash and deposits | 3.5% | 0.25% |
| Property | 4.9% | 7.47% |
| Hedge Funds | 6.1% | 7.47% |
| ‘Other’ | 4.7% | 7.47% |
| Weighted average discount rate | 100% | 4.4% |
Other key assumptions:
| First Actuarial “best estimate” assumption as at 30 September 2016 | |
| RPI | 3.43% |
| CPI | 2.43% |
| Post-retirement mortality | 100% S2PA
CMI_2015 [1.25%] |
| Proportion married | 85% |
What should you make of that?
For the average Joe, like the Pension Plowman, it is hard to challenge these assumptions, because Joe has neither the data nor the analytic skills to do so. I expect that some actuaries will look at these numbers and challenge them as either too optimistic or too pessimistic, but there will be general accord that they are reasonable. Actuaries are reasonable people who do not take outrageous positions.
Why the blue line and the red line on the graph are so divergent is not because we are using different assumptions, not because we are looking at the world in slightly different ways, like Turner and Constable.
It is because we see the social purpose of funding these defined pensions as being positive to our society, economy, to our way of life. We don’t see these pensions as socially divisive , a limiter on growth or as obstructing personal financial empowerment.
We want defined benefit pensions to have freedoms too!
In our opinion, the voices of those who want to see pensions level up to the quality of the best have been drowned by those who want them dumbed down to the worst. The risk transfer to DC has been badly handled so that we do now have a them and us culture.
In political terms , those “just getting by” aren’t getting enough of the pie and we’d like to see people looking at private pensions, the way we do – as making a meaningful contribution to people’s later life income. That cannot be achieved without efficient distributive structures (which these kind of pensions provide) or without a great deal of pay being deferred.
People and employers will not tolerate a great deal of pay being deferred unless they have confidence in the method of deferral. The Gilts + valuation methodology does not give employers that confidence, it just gives them big bills. It does not give people big pensions, it loses them their jobs.
The only way we can return to the days of confidence in pensions – is by having confidence that pensions work. Consigning pension strategies to the shackles of the gilts + funding methodology that drives the red line is to put not just our pensions but our aspirations in chains.
We see the world as we do, because we want the world we live in – especially the world we are going to live in – to be a better place.
Targeting a better pension
As I feared, the devil is in the detail.
Asset class returns to two decimal places, but are they before or after costs? Costs of managing an equity portfolio, a property portfolio or a hedge fund tend to vary by quite a lot.
There are really only two “best estimates”, one for so-called “matching assets” and another for so-called “growth assets”, if you ignore cash as I think you should rather than assume that everyone gets 25 basis points on deposits. We don’t.
Don’t you expect closed DB pension schemes to change their asset allocations over time?
Do you expect open DB pension schemes to continue with their present asset allocation forever?
What also seems to be lacking to me, Henry, is “Dear Prudence”.
At least I can say I’m one of the 85%.
I am not going to defend the minutiae – that is for my colleagues – but on your more substantive point – I don’t see the composition of a defined benefit investment portfolio having to change – unless the objective of the pension scheme changes; the paradigm shift is in looking at the pension scheme positively rather than as an encumbrance!
For everybody’s information, the assumptions will be reviewed monthly and we will republish them, along with the updated index.
• Before or after costs? RESPONSE: Expected returns shown are AFTER costs.
• Cash should be ignored. RESPONSE: Only 3.5% of total assets are in cash, and a 0.25% long-term assumption is arguably prudent (a best-estimate return could be higher).
• Don’t we expect closed DB pension schemes to change their asset allocations over time? RESPONSE: The FAB Index provides a snapshot position. It does not attempt to suggest how schemes should change their allocations over time.
• Why no prudence? RESPONSE: The FAB Index is not answering the question of how schemes should be funded (which we agree should include prudence in the assumptions). It is answering the question of how well-funded are schemes on a best-estimate basis based on their current asset allocation.
Ok, so help me out here.
The current approach is to use a gilt yield to discount the liabilities. This says there is a glaring hole, so we then add a realistic assumption to account for the ‘realistic’ expected return on assets and then reduce it a bit to ensure we are prudent. We then say we need a funding policy to meet the shortfall.
In an alternate reality, you say, we can take a ‘realistic’ return assumption for the assets to discount with (which looks very like the last 20 year real return) and say there is no hole. We then apply prudence to account for the fact that we can’t be sure we’ll get our realistic return assumption and create a funding policy on this basis.
At this point, my simple mind says, “yeah but there the same thing” – unless you are being more or less ‘realistic’ and/or more or less prudent under each methodology.
My mistake of course is assuming that your expected return on equities and all the other asset classes is in any way related to the expected return on gilts. But of course it is not. Gilts are way overvalued and we’re paying too much for the security of the income stream. Equities, by contrast, offer some fantastic additional compensation for the risk that they carry. Such that when you shake a bit of realism and sprinkle a bit of prudence one approach arrives at a better answer than the other.
I’d rather see the honest position then find a plan to solve the challenge. Relying on one person’s ‘realistic’ expectations only serves to deceive. You demand transparency elsewhere, why not here?
Let’s hope equities deliver an almost identical real return as they have over the last 20 years (3.7% from Barclays equity gilt study) from near all time highs.
But surely funds are only holding. Gilts because they’re all mad. We should assume that the 47.7% moves straight to equities and use a 7.47% discount rate for everything. Problem is more than solved.
Hide the problem, take ever more risk and last one out turn the lights off. If only the contributions hadn’t stopped, we could keep the Ponzi scheme going for ever.
This takes the actuarial profession backwards 20 years.
The purpose of the FAB Index is to raise awareness that Trustees need to know the full story when setting funding plans. That is, they need to know the “buyout” position AND the “best-estimate” position. They can then use their view of the strength of the employer (the “employer covenant”) to decide how much prudence to introduce that takes them from the “best-estimate” position towards the “buy-out” position. The employer covenant will also influence how much investment risk they are prepared to take.
That process works in reverse too. Trustees can start with the “buy-out” position and decide how much risk they want to take on the employer covenant as they currently do. If you keep the same risk and reward trade offs and are consistent in your application of actuarial prudence, you’ll end up at the same place. If you are being consistent in your assumptions, they are commutative.
Herbert
To your question..
“I’d rather see the honest position then find a plan to solve the challenge. Relying on one person’s ‘realistic’ expectations only serves to deceive. You demand transparency elsewhere, why not here?”
the point of FABI is to create an argument (I think dialectic sounds too academic – but that as well!). In practice, First Actuarial is quite transparent with its clients, it advises them on the options for valuation and of course it’s the trustee’s choice. We are keen they use the appropriate valuation method for their needs. Quite often they will look for the greater security of gilts plus , but there are often risks attaching to demands against the sponsor (pre-packs and the PPF through to a loss of investment and jobs).
We hope that by exciting this argument, we can widen the current debate – read Con’s two articles this morning over another glass of Shiraz tonight!
Pingback: Do Valuation Assumptions Matter? | P-Solve US
Pingback: Do Valuation Assumptions Matter? - River & Mercantile Solutions