Drawdown danger: how to avoid the post-Budget pitfalls
The Budget’s retirement reforms will dramatically increase scope for setting up innovative drawdown plans, but because managing drawdown is technically very difficult, one false move could send the whole line of dominoes tumbling down.
With drawdown the personal consequences and regulatory risks of getting it wrong are high. However, advisers should not expect to rely on a readymade solution. On paper it might look like the solutions will be institutional, as life companies replace annuities with a product in which the rate of draw is a managed outcome.
This is the sort of solution the Department for Work and Pensions prefers, hence it is backing collective defined contribution plans in which both investment and longevity risks can be pooled.
But this misses the point that funding retirement spending is a process not a product. Besides, after with-profits nobody really trusts investment risk pools.
Life after GAD
Throwing away the Government Actuary’s Department (GAD) rules is an opportunity to define a new approach to achieving capital efficiency for a household.
Advisers will be setting-up drawdown plans for client piece by piece and must avoid any wobbles along the way. As part of a holistic solution to retirement funding involving pension assets, other savings and even (perhaps particularly) freehold property, it has to be specific to each individual in a way products cannot be.
The technical challenge of managing drawdown calls for stochastic modelling. I am convinced it is the only way to make a highly-customised approach cost-effective; quite apart from the other advantages of consistency and objectivity that quantitative decision processes can bring.
In our case, the same model is used both to plan a holistic retirement goal and to manage the assets.
There is no reason why the planning and management functions, though sharing access to the same model, should not be performed by different firms. The two have very different economics so ‘separation by collaboration’ can make good business sense.
A helpful development is the platform technology that portfolio models depend on.
Enough of the customisation benefits offered to wealthy clients can be captured at much smaller portfolio sizes, by assigning a plan to one of five risk ‘channels’ each with five stages with varying mixes of risky and risk free assets – 25 models, plus one holding only risk assets.
Passage between the stages within a risk channel (the ‘glide path’) can be managed systematically.
The information need to match a client to a model at each stage consists of the planned spending profile at different ages (hence the ‘duration’ of the plan), risk tolerance and market conditions.
A modelled approach to drawdown is compelling on many fronts. It can make portfolios more resilient to stress, which is important to investment professionals. Changing the entire conversation to one about spending outcomes turns what looks like an investment-management method into a continuous process of financial planning.
The conversation is not just based on changes in the client’s circumstances but also progress towards their goals and it puts the adviser at the top of the value chain.
Compliance risk is also better managed because the process generates its own audit trail of logical and consistent discussion and decision.
The inputs needed from clients will be familiar to advisers accustomed to cash-flow models. But there the comparison ends. What need to be modelled are the outcomes of a lifetime plan, not a static portfolio. It can start at any stage, when accumulation begins, and continue as long as an annuity is avoided.
Gross drawdown outcomes are presented in real terms. Only an effective tax rate is necessary to turn this into real spending amounts at each horizon, and the plan can have whatever time profile the client wants.
Modelling forces consistency between each of three parameters: resources, time horizons for spending and risk taking.
Risk is not abstract but specific to the goal and parameters. Defined in terms of the range of probable outcomes, it addresses not just generalised volatility but horizon-specific return uncertainty, inflation and longevity.
Outcome projections can be continuously benchmarked against both a level annuity and an inflation-protected annuity as well as against a riskless portfolio in which only longevity risk remains exposed, identifying the cost and benefit of insurance.
Apart from not running out of capital, a common constraint will be avoiding the pinch points where bad investment outcomes seriously affect wellbeing (what the regulator means by ‘capacity for loss’).
The importance of never being forced to trim spending (except as planned) varies with individuals. So too is tolerance for when and how much housing equity might be relied on. Also, in drawdown, the consequences are radically different when there are no children.
It is because drawdown cries out for technology that we think it will be in the vanguard of new forms of collaboration between advisers, platforms and either discretionary fund managers (DFMs) or providers.
These new forms may not fit the distinctions advisers recognise today between outsourcing to a DFM, controlling the asset allocation or selecting multi-asset funds or managers of managers.