A dynamic value-based framework for retirement

That always struck me as odd. We are very good at recognising, storing and exchanging financial wealth, but much less good at recognising other forms of value that people create through their lives. Someone can spend decades creating social value for others, yet very little of that value is carried forward as something they can later draw upon.

It is a question I have kept coming back to. Eight years ago, in Pension Power: investing with attitude, I argued that pension members already had values; the problem was that pension systems often made those values difficult to express. People routinely reflect their values in everyday decisions - what they buy, what they avoid and which businesses they support - yet their pension savings can feel largely disconnected from those choices.

A few years later, writing in Professional Pensions – Why we are looking at DC value for money in the wrong way – I returned to the same underlying problem from a different direction. We were becoming increasingly good at measuring the things that are easiest to measure in pensions - costs, charges and historic investment performance - without first asking a more fundamental question: what do members actually value?

From value exchange to retirement

I now think that question becomes even more important in retirement.

Retirement values are not fixed. What matters to someone at 65 may be quite different from what matters at 75 or 85. Yet retirement is still often approached as a product problem: drawdown, annuity, collective defined contribution (CDC), retirement CDC (R-CDC) or some combination of flexible and fixed income. We ask which solution is best when perhaps the more useful starting point is to ask what forms of value the retirement system is trying to provide, what members give up to obtain them, and whether those trade-offs should change as retirement unfolds.

That leads me to what I would call a dynamic value-based framework for retirement. It builds on the same values-first thinking, but applies it to the design of retirement systems rather than to a one-off product choice.

That was part of the thinking behind value exchange ideas I began exploring in 2019 in conversations with Clare Wood, formerly global head of product and head of investment assurance at First Sentier Investors. The question was whether financial wealth was only one form of wealth, and whether other forms of value could be recognised more explicitly.

Consider a successful fund manager and a care worker. The fund manager may accumulate substantial savings, housing wealth and pensions. The care worker may spend a working lifetime supporting people who are elderly, disabled or otherwise dependent on help.

On a conventional financial balance sheet, the fund manager is clearly wealthier, but that immediately raises some harder questions. Which has contributed more to society? Which has created more value? And are the financial rewards attached to their contributions necessarily an accurate measure of the value created?

The point is not that the care worker is morally superior to the fund manager. It is that financial reward and social value are not necessarily the same thing. Our economic system happens to be extraordinarily effective at pricing, recording, storing and exchanging one kind of value.

That thinking later developed into the Digital Village concept, which Clare and I presented at the Library of Mistakes in 2023. The concept was much wider than pensions. It envisaged financial, social, nature and climate value, a wallet for owning, owing and offsetting, and an exchange through which different forms of value could potentially be recognised and transacted.

The analogy was the traditional village. Hundreds of years ago, if a neighbour helped bring in your crops, the exchange did not necessarily involve money. You might help with their harvest later, repair something, provide care or return the favour in another form. Trust, memory, reputation and reciprocity acted as part of the infrastructure of exchange. Money solved the enormous problem of allowing transactions to move beyond those small communities, but in doing so financial value became much easier to carry across people and time than many other things we value.

Technology may now allow us to recreate some of the characteristics of the village at scale. We can already see fragments of this in time funds, where help provided today can generate a claim on help from someone else later (a time-banking system such as the municipal Zeitvorsorge project in St. Gallen, Switzerland).

Air miles and loyalty points demonstrate that people are perfectly comfortable earning a unit of value in one context, storing it and exchanging it later. Environmental markets increasingly try to attach transferable value to carbon, nature or biodiversity outcomes. None of these systems is the Digital Village, but collectively they demonstrate that value exchange need not always mean a conventional cash transaction.

This brings the argument back to my mother. A better value exchange system would not necessarily have meant she received care for free, nor does the idea require every act of kindness to be assigned a financial price. But it might have recognised that she had spent a working lifetime contributing to a valuable social resource and allowed some form of that contribution to persist rather than disappear at the moment it was delivered. When she later needed care, financial wealth would not necessarily have been the only form of accumulated value that counted.

That remains the idea I keep coming back to: value created earlier in life should not necessarily disappear simply because it was not created in money.

From value exchange to retirement

More recently, I have started applying that thinking to retirement. A pension pot is financial wealth, but financial wealth is rarely the ultimate objective. Its usefulness comes largely from what it enables someone to do.

In retirement we exchange wealth for income and security, but also for independence, care, time, experiences, family support and inheritance.

Liquidity has value because it preserves the ability to respond to things we cannot know today. Secure lifetime income has value not simply because it produces more pounds each month, but because it can provide independence, confidence and freedom from worrying about whether money will run out. An inheritance is measured in pounds, yet its value may really be about family or legacy. Care appears as expenditure in a financial projection, while to the person receiving it its value may be dignity and quality of life.

This makes comparisons between retirement arrangements more complicated than they can initially appear. CDC may produce more lifetime pension income, but the member is exchanging individual pension wealth and liquidity for longevity protection and a collective income stream. An annuity makes another exchange explicit: capital is surrendered for guaranteed income for life. Drawdown preserves pension wealth, liquidity and optionality for longer, but leaves more investment and longevity risk with the member.

These are not simply alternative products performing an identical task. They package different forms of retirement value. My recent retirement work – Best value pension arrangement in retirement? – made the same point in a simpler way: pension income and pension wealth are both valuable, but they are not the same thing. Liquidity, security, longevity protection and the ability to change course also have economic value.

The practical implication is that we should define value before we start ranking products against it.

A small set of retirement values

This does not require us to build a psychological model of every pension member. That would be both unrealistic and unnecessary. The more practical opportunity is to identify a relatively small set – perhaps ten or twelve – of pension-related values that retirement arrangements exchange between themselves.

Income adequacy, income security, longevity protection, pension wealth, liquidity, flexibility, ability to change course, care optionality, simplicity, family or legacy, control and participation in investment upside would provide a reasonable starting set. Different members may place different weights on them, but the important point is that the industry can explicitly test retirement strategies against those values rather than quietly assuming that one of them - often income - represents value in its entirety.

Different forms of retirement value

This is why I prefer the term value-sensitive design to personalisation. The objective is not to construct a bespoke portfolio for every retiree. It is to recognise that retirement strategies embed different value trade-offs and to make those trade-offs visible when schemes, providers and policymakers design defaults.

There is also a quantitative dimension. The value of an additional £10,000 of accessible wealth is unlikely to be the same for someone with £20,000 of liquid resources as it is for someone with £1 million. Likewise, another £2,000 of secure annual income may be extremely valuable if it brings essential expenditure within secure-income coverage, but much less valuable once needs are already comfortably met. Economists would call this utility; in practical retirement design it simply means that the value of an outcome depends on what the member already has.

Under the bonnet, a quantitative engine could model a range of possible retirement outcomes and apply different weights or utility functions to the values being delivered. The mathematics can become sophisticated without the member experience doing so. A trustee board could use such an engine to ask whether a proposed default is robust across different value profiles, whether a CDC solution sacrifices too much liquidity for certain cohorts, whether staged annuitisation improves overall value, or how much security can be added before the loss of flexibility becomes disproportionate.

The immediate application is therefore not personalised advice. It is a better way of designing and testing retirement systems.

Dynamic in value, not just investments

The second part of the framework is that values - or at least their relative importance - are capable of changing through retirement. Someone at 65 may place considerable value on liquidity, flexibility, investment participation and helping children. At 85, the same person may place more value on simplicity, secure income, care resilience or independence.

This does not mean schemes should continually ask members to complete questionnaires. People can be fickle. A market fall can suddenly make security feel overwhelmingly important, just as a long bull market can make investment risk feel benign. Short-term sentiment is not the same as a fundamental change in what someone values.

A more human approach may be to distinguish between financial circumstances, which can be updated regularly, and personal values, which need only be revisited when there is a genuine reason to think they may have changed. Retirement itself, bereavement, a significant deterioration in health, the emergence of care needs, inheritance, separation or moving home are obvious examples. These are life events at which asking, 'Has what matters to you changed?' feels natural rather than intrusive.

There have, of course, been dynamic pension products before. Managed funds, lifestyle strategies and multi-asset funds can all alter investment allocations over time. But that is not quite the point here. Those approaches are primarily dynamic in asset allocation. A dynamic value-based framework is dynamic in what the financial resources are trying to achieve.

Changing the investment portfolio may be appropriate, but so might retaining more liquidity, altering withdrawals, introducing secure income, delaying an irreversible exchange or simply doing nothing. The innovation is therefore not dynamic asset allocation; it is the dynamic allocation and use of financial resources against retirement values.

A dynamic value-based retirement framework

The circularity of the second diagram matters. Retirement is not 'solved' on one date. A retirement system can have a default pathway while retaining the ability to respond when the economics or the underlying value trade-offs materially change.

Gradual exchange rather than one big decision

That becomes particularly important where decisions are difficult to reverse. Annuities provide an obvious example. Instead of asking someone at retirement whether they should annuitise and what proportion of their assets should be committed, a retirement design could allow secure income to be introduced progressively.

A relatively small portion of assets might initially be exchanged for lifetime income, with the position considered again later. More could be annuitised in a subsequent year, or nothing might happen. The pace could depend on what has actually occurred: investment experience, remaining assets, spending, annuity pricing, age and the level of secure income already provided by DB or State Pension.

This is what makes the approach path-dependent. Decisions made later respond to the path actually travelled rather than the path assumed at retirement. Flex-and-fix might sometimes emerge from that framework, but it should not be hard-coded into it. Drawdown, annuities, CDC, R-CDC and future products are tools. The value framework should sit above them.

Complexity belongs under the bonnet

The obvious challenge is implementation. A framework involving changing values, utility and multiple retirement options can quickly sound like something available only to wealthy clients with financial advisers. That is not the proposition.

Retirement itself is complicated. Markets, inflation, spending, health, longevity and family circumstances all change. Making one apparently simple decision at retirement does not remove that complexity; it simply fixes some choices before we know how those uncertainties will resolve.

The design objective should therefore be to keep the member experience simple while making the underlying system sophisticated. Members do not need to understand utility functions or stochastic models, and they do not need a unique portfolio. The framework can begin by testing whether scheme-level retirement strategies perform reasonably across a relatively small set of retirement values and plausible member profiles.

This is consistent with the argument I made in 2018. The solution then was not to turn every saver into an investment expert. It was to use technology and better pension design to make it easier for people's values to be reflected in mainstream arrangements. The same principle applies in retirement.

Complexity should sit in the system, not with the member.

Nor does dynamic mean constant intervention. Sometimes the right answer may be to leave a retirement pathway unchanged for years. Dynamic simply means preserving the capacity to adapt when doing so improves value.

A framework for all members

This also matters because retirement design cannot be built only around people with large pension pots. A member with a modest DC pot and significant reliance on the State Pension may have quite different value trade-offs from someone with several hundred thousand pounds and substantial other assets.

For the first member, the DC pot may be one of the few meaningful sources of accessible capital available. Liquidity could therefore have very high value. Exchanging much of it for another small stream of guaranteed income might raise measured pension income while leaving the member less resilient to an unexpected expense. For genuinely small pots, sophisticated optimisation may add almost nothing, and good defaults and safeguards may be all that is needed.

This illustrates why value-sensitive design is more important than mass personalisation. The framework can be used at scheme level to test a default, at cohort level where broad differences genuinely matter, and with limited member input where that adds value. It should not require millions of unique retirement portfolios in order to be useful.

The same applies to safeguards. A mathematical model should not recommend exchanging someone's last meaningful liquid reserve for a small increase in expected lifetime income simply because the utility score is fractionally higher. Essential spending, adequate liquidity and foreseeable short-term needs can be treated as constraints before the remaining resources are optimised across income, security, flexibility, growth and legacy.

A mathematically elegant solution is not necessarily a good retirement outcome.

Financial health linked to values

This ultimately brings the argument back to financial health. We usually define financial health using financial numbers: how much someone has saved, how much income they can produce, the investment return achieved or the probability that their money will last.

Those measures matter, but they are not the purpose of money.

I would define financial health more broadly as the ability to use the resources you have to support the things you value, now and through the rest of your life.

That makes values an input to retirement design rather than something considered once a product has already been chosen. It also makes the challenge manageable. We do not need perfect knowledge of every member. We need to recognise that people may attach different importance to a relatively small number of retirement values, that different products exchange those values in different ways and that the relative importance of those values can change through life.

Eight years ago, I argued that pension members already had values and that pension systems needed better ways of reflecting them. In retirement, the question becomes broader. Pension wealth itself is exchanged for income, security, flexibility, care, legacy and other things people value.

The next generation of retirement design should therefore not simply ask which new product is best, nor require millions of people to seek individual advice. It should build systems that are capable of understanding the value exchanges embedded in different retirement strategies, testing those exchanges against what members are trying to achieve and retaining the ability to adapt as retirement unfolds.

Value should drive the design. Products should be the tools used to deliver it.

Brian Henderson is an independent consultant, trustee and non-executive director

This article follows two related pieces Brian has published on CDC and the pension trilemma and Which pension arrangement offers best value in retirement? via LinkedIn.