This year will be a significant one for pensions and particularly for retirement planning.
The pension freedom changes in April are expected to revolutionise how people will manage their pension funds and other financial resources during the retirement period without any need to buy an annuity.
The investment problem, therefore, will be to look at the retirement needs for the particular individual – be they income, capital or to leave a legacy – and then to manage the money to achieve these goals with an acceptable risk profile, and within the possibility of an increasing life expectancy.
This has been part of the approach since drawdown was introduced in 1985, but perhaps not in as many cases as it should – in many cases drawdown was seen as annuity deferral, with an annuity being the final investment destination.
I am sure annuities will still have a part to play, but if we look at experiences in different countries then there may well be other lessons we can learn.
Let’s look at an example – in the US, the retirement planning market is already well developed, with a lot of academic thought given to how money in retirement is managed.
The US approach to retirement planning has broadly divided into two clear schools of thought – the ‘probability-based’ approach and the ‘safety first’ approach – and from these have spun off other ideas such as the ‘bucket approach’ and even the ‘utility-based approach’.
In all probability
Broadly, the probability-based approach invests a portfolio in a range of assets based on an assessment of risk tolerance. The retiree then takes the income generated from the portfolio and might even have to dip into the principal in the future should any of the assumptions not be achieved.
This approach is familiar in the UK, as it is really the basis of financial planning in the accumulation phase – the setting of a fund or income target and an asset allocation to meet a risk profile.
The safety first approach is geared to how people say they plan to use their limited resources to obtain the most satisfaction across their lifetimes. Targeting specific investments, it tries to match the proposed patterns of spending on a more granular level – asset to liability.
In institutional terms this equates more to liability-driven investment, looking more specifically at an individual’s spending patterns and their need for specific capital sums.
Once those basic requirements are covered, then the rest of the retirement portfolio would be allocated to other investments based on a person’s risk tolerance and time horizon for those objectives.
This has been refined into ‘modern retirement theory’, with a pyramid of funding hierarchy and risk matching – it starts with basic needs, then builds a contingency fund, then a fund for discretionary expenses and finally a long-term legacy fund.
The bucket approach seems to be a further iteration of this, with retirement money being segregated into different ‘buckets’, based on how soon the money will be needed.
An alternative approach in the US has also considered maximising the satisfaction you get from your withdrawals across your retirement. This has become known as the utility-based approach.
As the retirement planning concept means a limited amount of capital, there will need to be choices and the utility-based approach seeks to measure the utility of the chosen options to justify them as opposed to others not chosen.
The finer detail
So, having agreed on an approach as to how needs and goals will be aligned with investments, we move on to some further ‘nitty gritty’ issues which will again have relevance to the UK market.
In 1994, William P Bengen wrote an article in the Journal of Financial Planning considering, from a retirement planning perspective, how much of a fund could be spent each year without outliving the fund or losing buying power due to inflation.
Bengen statistically tested lots of data and situations from the last 70 years or so and concluded with a ‘rule of thumb’ – if you spend 4.5% of the assets in your fund in the first year of retirement and increase that annual amount each year by inflation, the fund will last for at least 30 years under all the historical scenarios tested.
In the US, this rule of thumb has been subject to scrutiny and has become known as the ‘SAFEMAX’ or ‘the 4% rule’. In recent years it has been attacked as producing too high a withdrawal rate given current unsettled economic times. It also has been accused of being purely probability based, too focused on the historic figures and even too country-specific.
Other relevant theories that flow from the SAFEMAX include decisions around how spending is varied if the SAFEMAX is exceeded or if growth is not as anticipated (Guyton and Klinger 2006).
My final academic reference in this article would be consideration of some of the issues considered by Moshe Milevsky, described in some circles as “the world’s leading researcher and authority on retirement income strategies”.
One of the subjects that Milevsky has covered, and which has started to gain some coverage in the UK in the light of the new pension freedom rules, is how an unfavourable sequence of investment returns (averaging out at the required rate but in an uneven order) can cause real damage to a retirement portfolio. The title of his research paper ‘Retirement Ruin and the Sequencing of Returns (2006)’ sums up the issues.
This could be particularly relevant should the UK’s Financial Conduct Authority, as promised, look at some of the potential pitfalls that could arise from a non-advised income drawdown regime. Alongside this he has given consideration to other issues such as the role of annuities in retirement planning in his paper ‘Life Annuities: An Optimal Product for Retirement Income 2013’.
Throughout all of the above, we can see something of a pattern – the quantitative modelling overlaid with the qualitative decision making and objective setting. Neither can ever be an exact science, as they are based on the vagaries of the investment markets and of putting together a portfolio to meet a specific return alongside the different vagaries of human needs and behaviour.
This is the way the UK market has been developing over the years, with the focus on lifetime planning and cashflow modelling as opposed to transactional sales.
It is important that we learn lessons from like-minded retirement systems, particularly the US and Australia where annuities are part of the financial planners’ armoury but are not necessarily regarded in the best light. This will mean the design of new products – variable annuity type products with guaranteed income levels or guaranteed death benefits.
Perhaps deferred annuities that come into payment in a retiree’s 80s. I am sure there will be new funds, perhaps targeting the SAFEMAX and/or even offering downside protection.
From equities to bonds
Similar considerations are emerging from Europe and, as long ago as 2008, A Report on Rethinking Retirement Income Strategies – How Can We Secure Better Outcomes for Future Retirees was published by EFAMA. Written by Professor Maurer and Barbara Somova, it concluded that retirement income options were often constrained by a requirement to buy an annuity, and that better outcomes could be achieved by holding equities in early retirement with a switch to bonds and annuities over time.
The report focused on the quantitative side of things and not the behavioural bit, but the work being done on the behavioural side is quickly catching up. More and more is being done on behavioural issues and the FCA has pointed out its increased focus on this.
Add this to the work done on the quantitative side, and I think the whole retirement income world in the UK will change dramatically.