The Five Years Before Retirement: Decisions You Don’t Want to Leave Until the Last Minute

Why one retirement number is not enough

Suppose two households each have R25 million of assets as retirement approaches. At first glance, their financial positions may look very similar.

But one household needs R70,000 a month, has no debt and holds most of its investable capital in relatively accessible portfolios. The other spends R160,000 a month, expects significant family support obligations and holds a meaningful portion of its stated net worth in a private business it is not yet ready to sell.

These are illustrations, not client cases or retirement-capital targets. The same headline asset value tells us remarkably little about retirement readiness.

The useful calculation needs to connect spending, inflation, tax, medical costs, other income, liabilities, the accessibility of each asset and the time horizon. It must also consider what happens if the future is less favourable than assumed.

As we explored in what changes once you have enough to retire, the answer may eventually be that the household is comfortably funded. But reaching that conclusion is the beginning of transition planning, not the end.

Start with the life, not the portfolio

Before deciding what investment changes to make, establish what retirement is likely to cost.

The present household budget is a useful starting point, not a complete answer. Some expenses disappear or reduce when work ends; others increase. Travel and leisure may be front-loaded into the earlier years. Healthcare and assistance may become more important later. Large once-off commitments, support for family members and possible property changes can also distort an average monthly figure.

A practical exercise is to separate spending into three categories:

• Essential spending: The household’s core living costs, which would be difficult to reduce without a meaningful lifestyle change.
• Flexible spending: Expenditure that matters but can be adjusted if circumstances or markets change.
• Exceptional spending: Travel, property work, vehicle replacement, family commitments or other irregular costs that should not quietly disappear inside a monthly average.

This distinction has consequences. A financial plan that assumes every expense must grow at the same rate every year misses how real households make choices. Conversely, a plan that assumes spending will naturally fall with age can underestimate later-life costs.

It is also worth planning in current purchasing-power terms. A future income quoted in nominal rand can sound impressive until inflation is taken into account.

Know which capital can do which job

In a financially established household, retirement is rarely funded by one account.

There may be pension, provident, preservation and retirement annuity benefits; local discretionary investments; offshore assets; tax-free investments; existing annuity income; property and business interests. These differ in access, ownership, tax treatment, currency, risk and estate consequences.

Before moving money, prepare an accurate household balance sheet that shows which capital is actually available to support retirement and when. A business valuation should not automatically be treated as cash that can fund spending next year. A primary residence is wealth, but not necessarily an income-producing asset. The look-through offshore exposure of local funds should not be counted as another separate asset.

It is useful to organise investable assets by purpose, not merely by provider:

These are planning roles, not prescribed products or asset allocations. One legal wrapper can contain more than one role, and each role may need different liquidity and risk characteristics. The appropriate structure follows the actual financial plan.

Model the transition, not just the destination

Retirement projections are often reduced to a single chart showing a capital balance gently rising or falling until age 90. That can be reassuring, but the confidence is usually in the smoothness of the line, not in the future itself.

Five years out, the more useful approach is to examine several plausible outcomes and the decisions they would require. For example:

• What happens if retirement begins earlier than expected or employment income ends unexpectedly?
• What if the first years of retirement coincide with weak investment returns?
• What if medical spending rises faster than general inflation?
• What if a planned business sale is delayed or delivers less cash than anticipated?
• What would change if one spouse died considerably earlier than the other?

Modelling should distinguish the return the financial plan needs from a return an appropriate investment strategy may reasonably target. Those are not automatically the same. A higher required return is not permission to take more investment risk than the household can bear; it may instead indicate that spending, saving, the retirement date or other assumptions need revisiting.

Investment objectives also need consistent measurement. An inflation-plus target should specify the inflation measure and period and account for all applicable investment and advice costs; personal taxation should be treated explicitly in the financial plan. It is an objective, not a forecast of what will happen every year.

Rethink risk and liquidity without overreacting

It is tempting to treat retirement as a date on which an investment portfolio should suddenly become conservative. But an age is not an investment objective.

Capital that must fund spending soon has a different job from capital that may remain invested for twenty years or be left to the next generation. Equally, holding too much of a long-term portfolio in nominally stable assets can expose retirement spending to inflation risk over time.

A useful pre-retirement review asks which withdrawals are foreseeable, which assets may need to be sold to meet them, how much liquidity is appropriate and how the plan would respond to a market decline early in retirement.

The answer is not a universal number of years of expenses to hold in cash, nor an instruction to sell growth assets shortly before retiring. It is a liquidity and investment policy suited to the household’s specific needs, time horizons and capacity for loss.

That policy should be built before the first retirement withdrawal is due, not improvised during an uncomfortable market.

Understand decisions that are hard to undo

Some retirement decisions have consequences beyond the returns on the underlying investments. The final years are a useful time to understand the options and constraints, even if the actual decision will be made later.

Retirement-fund benefits and income structures

Pension, provident, preservation and retirement annuity benefits may have different rights and histories. The two-pot regime introduced savings, retirement and vested components, with access and annuitisation consequences that depend on the member’s circumstances and fund rules. Do not assume that the same cash-versus-annuity choice applies to every account or component.

Living annuities and guaranteed life annuities solve different problems. A living annuity offers investment participation and flexibility subject to rules, but leaves investment and longevity risks with the investor. Guaranteed annuities can provide contractual income characteristics, but their flexibility, survivor options and inflation treatment must be evaluated carefully. Neither label tells you which combination will best serve a particular household.

Tax on retirement benefits and later withdrawals

Retirement lump-sum tax is affected by relevant prior benefits and cumulative tax rules, while regular annuity income and discretionary investments have different tax treatments. A strategy that looks efficient when each account is viewed separately may be less effective once household income, withdrawals and future estate consequences are considered together.

For deeper analysis once income is actually needed, see our South African retirement withdrawal strategy. The objective five years out is to understand the decision framework, not to prescribe a universal withdrawal order prematurely.

Employment, business and estate consequences

Some clients also need to coordinate employer benefits, notice periods, share incentives, succession arrangements or a planned business sale. These may have long lead times, valuation uncertainties or specific tax consequences. Spousal income arrangements, wills and beneficiary nominations deserve attention before retirement changes the household’s circumstances.

A practical five-year decision sequence

The dates below are not legal deadlines. They are a useful way to put the bigger decisions in the right order and avoid leaving everything until the year you stop working.

Someone retiring in eighteen months can still use the framework. It simply means combining stages and tackling the most consequential gaps first. There is no prize for having started exactly five years earlier.

Make the plan work for the whole household

Retirement plans are often presented as though one person’s age and investments tell the whole story. They rarely do.

A surviving spouse may need income for considerably longer, have different healthcare costs, or be less involved in managing investments. Adult children may rely on support that is not visible in the household’s ordinary spending. An offshore account or family company may add administration and estate complications.

The retirement transition is an opportunity to clarify who understands the plan, who can access the right records, what happens to income if one spouse dies and who will coordinate decisions. Our estate-planning guide addresses the wider estate context and the importance of coordinating the household arrangements.

The aim is a plan that continues to make sense when family circumstances change, not one that works only while every assumption remains convenient.