Retirement planning: an income strategy for old age
We have been looking into what people from around the world do with their defined contribution savings after they retire*. No market has yet found the ideal strategy. The risks and degree of uncertainty involved often lead to conflicting objectives. Last year, we suggested some principles that any investment solution would need to follow to reconcile these conflicts^. Having since drawn on the experience of other markets, we now propose an income strategy that combines stable, real growth in early retirement with some sort of longevity insurance in later old age.
In arriving at this solution, we discovered that certain risks and requirements were common to pensioners the world over. Four key areas of uncertainty stood out:
1. Investment – the risk of earning less than expected
2. Longevity – the risk of living longer than expected
3. Inflation – the risk of prices rising faster than expected
4. Consumption – the risk of spending more than expected.
Of these risks, the only one where the individual has any direct control is consumption. The relative importance of the other three changes over time. Figure 1 shows the impact of a small change in each of these key variables. Early in retirement, the risk of not achieving sufficient returns is the major worry, as there is still a significant period of time over which to grow the assets. The threat from inflation is also at its highest early on for the same reason: there is a long period of time over which the uncertainty can manifest itself. Longevity risk starts out relatively small, owing to the high probability of survival through the early years. However, this risk grows with age, reflecting the fact that longevity is selffulfilling, i. e. the probability of reaching 90 is much higher at 89 than at 65.
This insight helps to focus the solution on the appropriate risk at each stage of retirement. When the savings are largest, generating strong real investment returns with limited losses will provide the biggest benefit. As the pensioner ages, and withdraws income from the account, protecting them against the risk of their outliving their savings should be the main focus.
There is a clear gap between what individuals need in their post-retirement solution and what they want. To manage the economic and actuarial risks outlined above, individuals require a solution to provide certain features. Their most basic needs would include flexibility, stable returns, inflation protection and longevity protection (primary criteria). But they are also likely to want predictable income, simplicity, adequacy and legacy benefits (secondary criteria).
Clearly a number of these factors conflict with each other (for example, predictability versus flexibility) and it is therefore difficult to rank them in order of importance. As with many investment decisions for individuals, a balance of factors is likely to be most effective.
Components of a post-retirement solution. There are essentially three components of post-retirement income provision:
1. Cash lump sum
2. Investment accounts providing nonguaranteed income
3. Longevity protection, likely to include some type of annuity.
Any of these components is unlikely, in isolation, to be sufficient on its own. In Figure 2 we analyse how they stack up against our primary needs criteria.
In our opinion, the ideal solution should have as many “green lights” in the primary criteria box as possible. Since no single product meets all of these criteria, a combination of components is required. The solution should focus on maximising risk-controlled growth opportunities in the early stages before adjusting to protect against longevity risk later on, fitting the pattern of risk sensitivities as the pensioner ages.
There are a number of ways to hit these moving targets. We believe all would probably need to combine guaranteed income from an annuity with a non-guaranteed, variable income from an individual account. Here we have outlined three possible “blends” of these two components:
1. Account-based income with deferred annuity. All income in the first few years would come from an investment account, supplemented in later (“frail”) old age by an annuity bought at retirement but whose guaranteed income is deferred until needed. In our modelling1, we have assumed 30 % of the account would be spent at 65 on a deferred annuity which begins to pay out at 80. If the pensioner were to use a balanced portfolio of equities and bonds for the remaining savings, we found that account-based income would typically run out at around 89 using this approach, although the annuity would obviously continue to pay out.
2. Account-based income converted to an annuity in later years. All income for the first few years would come from an investment account, later transformed into guaranteed income from an annuity bought when needed in frail old age. We again assumed the pensioner would start drawing on the annuity at 80. The risk here is, of course, that annuity rates move unfavourably in the years between retirement and reaching 80. Our backtest showed that the remaining “pot” of savings would typically only purchase an annuity sufficient to provide four-fifths of the income target of 60% of pre-retirement earnings.
3. Account-based income with immediate annuity. All income from start to finish would be provided by a combination of an investment account and an annuity. In this case, we assumed that half the initial income would come from a level annuity, with the investment account providing a rising top-up to that. We found this continued to pay out until after 90 in half the backtested scenarios, although in 30 % of cases it had run out at 85.
Weighing up the pros and cons, we considered that option 1, which relied on an account-based income and a deferred annuity, came out ahead. Testing this strategy against the pensioner’s “needs” and “wants” mentioned earlier, we see an increased number of green lights, as shown in Figure 3.
Principles for a successful post-retirement solution
There are many variables in retirement: how long people will live for, the costs of goods and services, interest rates and so on. Faced with this degree of uncertainty, people tend to make poor decisions.
In light of this, we believe that pensioners need guidance on what constitutes a good quality retirement solution. Given the general absence of such guidance, some have argued for the creation of a post-retirement “default strategy” to provide pensioners with a better starting point for these decisions.
However, it is not an approach we would advocate for several reasons. Clearly, every individual’s circumstances differ and no single default can fit all. As a result there is the potential for misbuying/ mis-selling risks. At the same time, although financial literacy remains low in many markets, greater access to the internet and more sophisticated financial analysis using the web make it easier for people to make investment decisions. And, of course, there is always a minority who welcome choice so that they can tailor their own finances. Last, and by no means least, it is difficult to see how one would get agreement on what should constitute a “default” approach.
Instead, we favour an approach that seeks to establish a set of principles which are the necessary conditions for good quality retirement solutions. In the UK there have been preliminary discussions about awarding a “kitemark” to suitable funds which meet certain standards. This would be comparable to “QDIA-approved” funds in the US which are officially authorised for use as a default in the pre-retirement stage.
However, an overarching solution is far broader than simply a fund or insurance product. Where a fiduciary is involved, for example in a corporate pension plan, a pensioner could be given a shortlist of suitable investment funds and a shortlist of suitable longevity protection options from which to choose (see table below). The individual would choose the proportion to allocate to each list, subject to a minimum. If permitted and tax- efficient, a partial cash lump sum might also be taken. A typical set of choices might be as follows in Figure 4.
Using technology and real-world assumptions, individuals could assess the likely impact of different choices on the illustrative outcomes they receive, with a clear distinction between guaranteed and non-guaranteed benefits, and the purchasing power of future income. This choice could be revisited on a regular basis. At some point, as our earlier analysis showed, there is a tipping point beyond which the protection component becomes far more valuable.
The options in list A would provide stable, real investment returns and be able to adapt to changing requirements. In reality, this means that the funds in list A are likely to be well diversified and fairly liquid. Clearly the kitemarked/ approved components in both lists should offer value for money, but this should not be confused with “cheap”. A purely passive strategy is unlikely to deliver the outcomes that pensioners need. Individuals should be encouraged to make their selection with guidance or advice at this important point in their lives.
For pensioners where no fiduciary is involved, providing guidance and having approved choices should help them with this difficult decision and improve outcomes for them. Asset managers and insurers should take some responsibility for the thoughtful design of these strategies.
We looked at how different parts of the world tackle the conundrum of how to create a safe and sufficient income in retirement. We concluded that a successful post-retirement strategy needs to adopt principles that address certain common risks: stable, real investment returns, flexibility to adapt to changing requirements, simplicity in implementation and communication, and reliable protection against longevity risk. The resulting solution will therefore combine investment and insurance components. With lengthening life expectancies, we anticipate account-based strategies will blend a growth and income approach for the first 15 – 20 years after retirement, with longevity protection engaging in later life. We believe this approach would help shift the post-retirement conversation towards a long-term investment solution more likely to provide an income that would match pensioners’ realistic expectations.
*See “Global lessons in developing post-retirement solutions”, Investment Perspectives, May 2015, for the longer version of this article with the detail on individual countries.
^See “Global lessons in developing post-retirement solutions”,Investment Perspectives, May 2015, for the See “Pointers towards a better pensions landscape”, Investment Horizons, issue 2, 2014.version of this article with the detail on individual countries.
1 See “Global lessons in developing post-retirement solutions”,Investment Perspectives, May 2015, for the See “Pointers towards a better pensions landscape”, Investment Horizons, issue 2, Assumptions used. Pensioner: final salary of $ 50,000; aiming to replace 60 % of work income; savings equal to 12 times final salary; retirement age of 65; annual withdrawals midway through year; mortality probabilities from UK life office male pensioners combined lives’ data, Institute and Faculty of Actuaries. Investment: period covered, 1952 to 2013; cash, all invested in US three-month deposits; equities, 70 % S&P 500 Index, 30 % MSCI EAFE Index; balanced fund, 50 % S&P 500 Index, 50 % US 10-year Treasury bonds; inflation, US Consumer Price Index All Urban, seasonally adjusted.of this article with the detail on individual countries.