We welcome the Chancellor’s decision, subject to legislation, to scrap the Government Actuary Department (GAD) rules that were designed to ensure (in most cases, still) that savings in a personal pension plan last a lifetime. GAD-constrained drawdown (the process of extracting capital from the pension in stages) was directly comparable with an annuity as means of turning accumulated savings into an income stream to fund spending in retirement because both were designed to last a lifetime. Anything that encourages wider adoption of drawdown instead of buying an annuity at retirement is welcome because people will generally enjoy higher lifetime spending if their capital ‘retires’ later than they do. It is a great opportunity for individuals but also a challenge for the professionals they will need to turn to for advice.
Scrapping rules that limit the possible profiles for extracting monies from pension accounts may not sound very radical yet has been widely reported as if it was. And in one sense it is. But it is not a logical part of a radical rethink of the nature of the pension contract, which uses tax relief to encourage socially-desirable outcomes. That contract is clearly not working as intended. There is a gap of about £15 billion per annum between the value of tax relief on pension contributions and the amount of tax recovered from pensions in payment, a shortfall of over 50%. The gap has been fairly constant for about 10 years in spite of successive governments bearing down on new tax relief given. If the NSO numbers (collated by HMRC) are correct they imply a substantial wealth transfer between generations of taxpayers which is not what was intended and is very hard to justify morally. These new changes, subject to some important decisions about how to tax death benefits, seem likely to make the transfer even larger. This will probably come out in the process of consultation and debate that precedes any legislation.
The Chancellor claimed that the GAD rules smacked of ‘paternalism’ and protests from the other side of the House that the freedom would be abused were therefore ‘condescending’. But the Government may also have felt a sense of its own responsibility for the particularly poor value of annuities as a consequence of its policy of quantitative easing. This had led to the extraordinary situation, with very low bond yields, that the cost and profit of building an annuity (the choice of over 90% of retirees) overwhelmed the gain in utility from laying off longevity risk. An individual could (theoretically) fund retirement from a ladder of gilts of different maturities through to age 100 and enjoy a higher ‘income’ than an annuity. Meanwhile, the FCA had been busy reviewing the lack of effective competition, not helped by shocking customer inertia, in the retirement annuity market. So perhaps the real reasons are pragmatic and freedom of choice is just the spin.
The implications for retirement advice
As both financial planners and discretionary managers specialising in retirement portfolios for high net worth clients, Fowler Drew operates in a world in which GAD constraints have relatively little impact. As flexibility ceases to be the privilege of the wealthy, lessons can be learned from how the wealthy plan. The business opportunity is massive but the technical challenges are also great and there will be competition for solutions from insurance companies and money managers as well as advisers.
This is not just a challenge for the industry. The Chancellor’s proposals, if enacted, pose a problem for the Financial Conduct Authority’s approach to regulating pensions advice. It is really only GAD rules that give a registered pension scheme its own integrity as a product or ‘wrapper’ separate from any other sources of possible spending money. The FCA’s approach relies implicitly on that separation.
Without GAD rules, the following logically apply:
the optimal rate of pension extraction should be arrived at independently of the rate of spending – it’s just a tax strategy, calculated to maximise the present value of future post-tax cash flows (the death benefit rules may alter after consultation but the option to exploit them is less realistic in the mass affluent market and the obvious solution is anyway to equalise treatment with IHT)
the risk approach should be determined by reference to the retirement plan as whole, not treating the pension wrapper as requiring its own risk approach and suitability assessment
the pension wrapper is better used to hold certain of the plan assets, not others (so diversification has to be viewed differently)
there is no basis for biasing to an annuity as the default option, any more than biasing to a purchased-life annuity for non-pension capital (advisers believe an FCA bias exists, even if it does not)
investment return illustrations prescribed by the FCA are misleading when the time horizons are multiple but indeterminate; critical yield calculations are meaningless when the product no longer has to provide whole-life benefits.
That is quite a lot right there for firms and the regulator to take on board!
Extending pension drawdown to the mass market as part of a broader, holistic drawdown plan to fund retirement spending will tax the skills of the advice industry as well as portfolio managers. Drawdown is one of the most technically challenging of investment tasks.
It is much more explicit than other goals, with quantified outcomes at multiple dates subject to defined tolerances, such as the chance of sustaining spending in real terms, avoiding cuts (other than those tolerated because they fit a profile for spending at different ages).
Client ‘utility’ is both idiosyncratic and complex because of tension between maximising lifetime spending and, for instance, avoiding or postponing the sale of property (or equity release) and the wish to leave a bequest.
Outcomes are path-dependent.
A job for modelling
It is probably impossible to do any meaningful planning of a drawdown goal without ‘stochastic’ or probabilistic modelling that expresses chances of different outcomes. By randomly simulating the volatile market return paths that might arise throughout the period of draw, it is possible to measure the sustainability of a pre-determined rate of draw and to stress test the plan. It is important that the simulations include uncertainty about future inflation rates since an important gain in welfare from drawdown relative to a ‘level’ annuity is that inflation risk is dealt with better. Inflation can be as threatening for retirement living standards over long periods as market risk. So simulating path-dependent outcomes (sustainable real rates of spending) is key to giving realistic advice.
Just as important a benefit of stochastic modelling is that probabilities changes the nature and clarity of the conversation between adviser and client. Clients can only make rational choices, and exhibit their ‘true’ risk preferences, using information about chances. The resulting risk preferences may well conflict with what the client has said and done in the past, if only because the conversation now focuses on outcomes instead of volatility (the bumpiness of the path which is how risk is mainly experienced). It is intuitive that we can learn to live with volatility but will live off outcomes.
Not much goal planning for private clients uses stochastic models. However, for the probabilities used in planning to be relied on to deliver the benefits desired by clients within their constraints, it is also imperative the investment implementation, as a goal-specific ‘virtual portfolio’ of financial assets in different wrappers and possibly with different owners, uses the same modelling as the planning. This is highly unusual if not unique to Fowler Drew.
Lessons from Liability Driven Investing
If capital does not fully retire when its owner retires, the time horizon for encashment or retirement of risk assets is always a rolling window into which only a relatively small proportion (depending on risk tolerance and market conditions) will fall. It also starts to fill well before drawdown starts. If that spending window was to be backed only by risk-free assets with short duration, it functions (ironically) like a DIY temporary annuity. The rest of the capital, held to match longer liabilities, needs to take equity-type risk.
This separation between hedges for near liabilities and risky assets only for longer liabilities is intuitive and increases clients’ composure. Their experience of bad markets is different when they know that all of their next 7 or 10 years’ spending is funded by assets with no risk of loss or inflation erosion.
The principle of managing risk by separation between risk-free hedges and risky asset proportions is well-established in the institutional market but retail wealth managers have been slow to adopt it for either its technical robustness or its communication benefits.
Without fit-for-purpose technology, the industry will predictably fall back on minimising volatility across all the assets, as this minimises the consequences for sustainability if continuing to draw at a pre-determined rate during a long period of poor market returns. The asset relied on to reduce volatility is fixed income. But if retirement outcomes are planned in real terms, exposure to nominal bonds merely replaces equity or business risks with unrewarded inflation risk. Drawdown is therefore a special case of the general industry problem of relying on diversification across asset classes to ‘control’ risk.
It seems likely that solutions will also be offered that cushion the consequences of continued risk taking, holding out more upside potential than downside risk. There are two versions of products claiming asymmetrical returns:
Insurance-based: part of the return is applied to buying insurance, using options or buying guarantees, against bad markets (examples are ‘third-way’ annuities and structured products)
Skill-based: judgement is used to ‘time’ exposure to risk assets (examples are hedge funds and some absolute-return unit trusts).
Though the sales pitch can be appealing, it is not realistic to assume these will get the job done. Insuring is a zero-sum game because the cost is broadly equivalent to the return premium from taking risk. Market timing skills that persist over the period required by a retiree are simply implausible.
The business opportunity
As to who is best positioned to exploit the drawdown opportunity, financial planners look to be in a strong position. They have the advantage that they usually see the entirety of a client’s balance sheet and are used to thinking holistically. Their weakness is in their decision-making technology. This is the strong suit of the insurance industry and could be acquired by both non-life fund managers and wealth managers.
But the technology that integrates planning and management can also be used as the basis of collaboration between planners and managers, with each using the same modelling to perform the different functions that each does best. We think this collaboration has the best chance of delivering success to all the parties and are talking to IFAs about making our model available to them as planner with us as manager of the assets assigned to a drawdown goal.