This blog looks at the various options available to the British pension system as alternatives to the current investment pathways, retirement “savers” have when looking at ways to access their pension pot. The options are laid out in this document which you can download or read from this link.
I was thinking of entitling the table using Steve Grove’s phrase “choose your poison”, but I prefer “you can’t have it all”.
Of course you can have it all, you can have an efficient income funded out of earnings which lasts as long as you do , it was called SERPS and then S2P but it did not win the support of the financial services industry who thought they could do better. So far they haven’t.
Let’s start by concentrating on the policy objectives that the OECD driving member’s rights (we’d say “outcomes”).
If you want to maximise retirement income and limit benefit volatility in retirement , you go the collective route. If you want to maximise individual equity (no cross subsidies) , be transparent , achieve simplicity and give individual’s flexibility then you go the individual drawdown route. Both collective and individual approaches to providing-guaranteed lifetime retirement income can limit member choice – the collective approach has to!
So in a conversation between John Ralfe and Simon Eagle – as happened last November, there must be space for tolerance of both approaches. The two do not seem mutually exclusive, there is no cuckoo in the nest of golden eggs.
Is maximising retirement income the main objective?
It’s too easy to suppose that the point of a pension is to maximise retirement income. The simple challenge is to use “utility” as your measure, perhaps the point of a pension is to maximise someone’s enjoyment of retirement. This is the point made by my friend John Quinlivan who argues that people should have their money when they have best means to spend it and “fuck the future”. If the Government is so libertarian as to consider fecklessness a crime, it doesn’t even have to provide a poorhouse for the destitute but most people agree that there are limits to liberty.
But I can see that where there is already a state pension, the safety net of Pension Credit and the last resort of a Citizen’s pension , we do not “need” to maximise retirement income, it should be the individual’s choice whether to maximise pension or maximise utility (if utility is perceived to be pension “freedom”).
Similarly, the arguments for an individual approach to turning pot to pension have their utility (bring happiness) through greater transparency (for the untrusting and the analytic), flexibility – for those who like control and like a simple concept – drawdown is simple.
Maximising income in retirement may not be the main objective for everybody but as with so much else in pensions, it looks like the objective for most of us. Individual flexibility – in policy terms – looks like the opt-out, collective pensions look like the default.
When would you want to a to and through retirement income strategy?
The OECD answer this very well. Running a whole of life retirement income strategy which pays a non guaranteed pension out of money accumulated in the same arrangement (like Royal Mail) makes sense when
- You are looking to maximise retirement income
- You are looking to limit benefit volatility – maximise certainty
- Keep it simple
- Limit member choice
This is how you’d characterise CDC as considered in the DWP’s regulations (PSA 2021) and TPR’s CDC code. But OECD make it clear that there is an alternative way, which is to provide a non-guaranteed income from a pot rather from a stream of contributions. This has advantages in terms of implementation – you only have to look at the time it’s taking to implement Royal Mail to see that the whole of life model is a challenge.
In practice , policy making tends to be pragmatic and looks to the short term issue. Right now people are getting to retirement with decent sized pots and making a mess of pension freedoms, so I expect policy making to focus on at retirement or “decumulation” only non-guaranteed solutions. That does not preclude further Royal Mails, it simply means the focus is shifting.
How do you set benefits?
If you are running a whole of life model, you can adapt the defined benefit accrual model and target a percentage or fraction of final salary or career average earnings, that’s based on contributions received. But if you are simply turning a pot into a pension, the formula must be like an annuity equation, you get x£ income for y£ capital.
Ideally, policy would focus on the Royal Mail model as it maximised return over the whole of life (the perfect model for “patient capital”. The OECD’s table recognises that converting a pot limits the maximisation of returns and the creation of stable benefits but it scores higher for limiting inter-generational squabbles, transparency and simplicity. I agree, the decumulation only model is a version of a flex now fix later approach recommended by Steve Webb – though “fix” is less firm in a non-guaranteed world and the point at which fix happens – is earlier than in the Webb model.
How do you adjust benefits?
The OECD are right to point out that this is the area of greatest contest between those who see collective and individual approaches as competing.
I sense that the “individual model” in this table , doesn’t actually exist – unless there are IFAs who have the capability of managing the process in the bottom row. Managing a drawdown using continuous mortality adjustments and the measurement of achieved versus anticipated investment returns is too time consuming, too resource intensive for all but larger pots.
The adjustment mechanism for individual drawdown can use collective buffers (for instance relying on Prefund’s distribution strategy but here we move into a demi-monde of “with-profits” that ultimately relies on insurance company guarantees (which is really outside the scope of “non guaranteed pensions”.
In practice, drawdown benefits are adjusted in consultation between client and advisor, there is no model which can really require many individuals to move in lockstep other than a collective model (the idea of a collective drawdown policy is as much a fudge as a “group personal pension”, it is little more than a marketing conceit with a collective payment system.
Too little is made of the advantage that a collective approach has in terms of managing benefits. In the conversation between Simon Eagle and John Ralfe, no mention was made of the resource and time needed to manage a drawdown, or the risks of not managing the process.
What do we learn from the OECD’s paper.
The OECD has done, what the PPI has yet to do, which is to think through the icy options for Government’s struggling to get beyond the guaranteed annuity as the way to turn pots to pensions.
It assumes that pensions are preferable to pots – though tax law in this country makes it better for people with pots to draw income from anywhere but a pension pot (at least till 75). (MPAA, death benefits and utility)
If the policy consideration is to maximise income from savings, which I think it still is, then the OECD is siding with a collective model (and pragmatically – a decumulation only model). If on the other hand, freedom of choice and the supposed utility it brings, is to the fore, then the report implies we should stick with investment pathways and individual drawdown from pot or bank account.
The bulk of chapter 5 rehearses the requirements needed for non guaranteed annuities to become acceptable , focussing on governance, communications, transparency and most of all the willingness of both the financial services eco-system and the policy makers to make it happen.
It remarks on one local issue in the UK , which I’ve talked about on this blog before , specifically that the detail of the regulations and CDC code is so precise that it may prevent other forms of non-guaranteed pensions taking hold. My feeling is that the DWP/TPR may have over-prescribed on Royal Mail to a point where any development must be via alternative legislation and regulatory guidance. I hope that Jo Gibson can prove me wrong in this.