How Much Money Do You Need to Retire Comfortably in South Africa?

The practical work is to establish the income gap, reconcile tax, test difficult return paths and decide what could change if the plan comes under pressure. “Comfortable” is a household budget, not a national rand amount.

In this article

Start with the spending the money must support

Use today's rands and separate three kinds of expense: essentials that must continue, discretionary spending that can change, and irregular costs that need their own allowance. Include medical aid and out-of-pocket care, rates and levies, home maintenance, replacement vehicles and any support for relatives. A monthly average should include annual bills, but a large imminent expense also needs an actual payment date.

Then list other income, its tax treatment, escalation and reliability. A pension with contractual increases differs from rent that may stop during a vacancy. Part-time earnings that are likely to end in five years should not reduce the income gap for the following twenty-five. For a couple, model what remains after either partner dies, including changes in pensions, tax and household costs.

A paid-off home can reduce housing expenses. Its entire market value is not automatically retirement-income capital. Count only the net proceeds a workable sale or downsizing plan could release, after replacement housing and transaction costs. Apply the same discipline to a business sale that has not yet happened.

Reconcile gross income with spendable cash

Suppose a 67-year-old South African resident needs R30,000 a month after income tax, has no other taxable income and receives all income as a fully taxable annuity. Assume no deductions, medical tax credits or other relief beyond the age-related rebates. Using SARS's 2027 tax-year brackets, the required annual gross income is approximately R425,583, with tax of R65,583, leaving R360,000 to spend. These are the rules for 1 March 2026 to 28 February 2027. SARS individual tax rates

At a hypothetical 4% starting rate, that gross withdrawal implies R10.64 million, rather than the R9 million obtained by dividing after-tax spending by 4%. The difference is approximately R1.64 million. It comes from treating tax consistently, without changing the assumed withdrawal rate.

This example does not apply unchanged to a couple, a younger person, someone with other income or a discretionary investment account. Selling investments can return original capital as well as realise gains; it is not automatically taxed as annuity income. Income tax depends on the nature and total amount of taxable income. SARS personal income tax

What the same tax assumptions imply at different starting rates

These are first-year capital illustrations for the specific taxpayer above, rounded to two decimals in millions. They are not recommended drawdown rates, annuity quotations or forecasts. Investment costs are not deducted from the gross income in this table; they reduce the portfolio separately and must be allowed for in the sustainability assessment.

Test a path, not just a percentage

The same savings can produce very different retirement incomes. This illustration tests what changes when returns arrive in a different order, costs fall, or income is cut. It starts with R12 million to explore those effects, not to set a retirement target for every household.

A retirement plan spends money while markets move. That makes the order of returns important. South African research has tested withdrawal outcomes using local asset returns and inflation, but historical simulations cannot establish a guaranteed rate for a future household. A 2016 study by Eben Maré used resampled monthly South African equity and bond returns from 1950 to 2014. Its conclusions depend on that dataset and methodology; they are not a current product quotation or a complete personal plan. South African Journal of Science

The first annual gross withdrawal is R425,583, equivalent to R35,465 a month before tax. The original income target rises by 5% a year. Read the capital and income outcomes together.

Four fixed illustrations over 30 years. Rand amounts use purchasing power at the start of year 1.

Scenario	Capital after 30 years	Final gross monthly equivalent paid	What happened to income?
Weak returns early	R0	R0	Income only partly paid in year 28; none in years 29–30.
Same returns reversed	R2,853,513	R35,465	Original gross income target paid for all 30 years.
Early losses, lower costs	R29,646	R35,465	Original gross income target paid, with very little capital left.
Early losses, income cuts	R3,925,046	R16,408	Five income cuts; final purchasing power is 53.7% below the original target.

Capital remaining after each year

• Weak returns early
• Same returns reversed
• Early losses, lower costs
• Early losses, income cuts

Income actually received

• Weak returns early
• Same returns reversed
• Early losses, lower costs
• Early losses, income cuts

Keeping more capital required less income. The spending rule cuts the previous nominal withdrawal by 10% in years 3, 10, 16, 21 and 26. The final gross monthly equivalent is R16,408, a reduction of R19,057 or 53.7% against the original target in starting purchasing power. No request goes unpaid in that path because the requested income has already been reduced.

In the weak-early path without income cuts, the year-28 withdrawal is only partly paid and capital is exhausted. No income is paid in years 29 or 30. Lowering costs to 0.5% preserves all original gross withdrawals in this illustration, but only R29,646 of real capital remains after year 30. Neither outcome establishes a sustainable rate for a household.

What the illustration assumes

Withdrawals occur once, at the start of each year. The remaining balance earns that year's gross return, then the annual fee is charged on the post-return balance. Costs are 1%, except for the 0.5% lower-cost path. These are invented cost assumptions, not ClearGauge quotations.

The weak-early returns are −15%, −5%, 10%, 12% and 15%, followed by 25 years at 8%. The reversed path uses the entire same list backwards. Inflation is fixed at 5%. From year 2, the flexible rule tests the previous requested withdrawal increased by 5%: if that exceeds 5% of opening capital, it instead requests 90% of the previous nominal withdrawal. Otherwise it applies the normal 5% increase.

Real capital at the end of year n is nominal capital divided by 1.05ⁿ. Income paid at the start of that year is divided by 1.05ⁿ⁻¹. Both refer to the start of year 1. Dividing annual income by 12 does not simulate monthly cash flows.

This is a hypothetical investment pool, not a living-annuity implementation. Product drawdown limits, future income tax, mortality, changing healthcare needs and asset-specific trading costs are not modelled. Preserving gross purchasing power does not promise constant after-tax spending. Four invented paths provide no probability of success and cannot establish suitability for an individual retirement decision.

See the annual capital and income figures

All amounts below use starting purchasing power. Income columns are gross monthly equivalents. Policy reductions plus unpaid requests equal the gap against the original R35,465 monthly target.

Weak returns early

Thirty annual observations for weak returns early

Year	Capital at year end	Income paid	Reduction by rule	Requested but unpaid
1	R9,276,069	R35,465	R0	R0
2	R7,927,507	R35,465	R0	R0
3	R7,780,567	R35,465	R0	R0
4	R7,766,864	R35,465	R0	R0
5	R7,960,046	R35,465	R0	R0
6	R7,672,237	R35,465	R0	R0
7	R7,379,164	R35,465	R0	R0
8	R7,080,733	R35,465	R0	R0
9	R6,776,844	R35,465	R0	R0
10	R6,467,399	R35,465	R0	R0
11	R6,152,296	R35,465	R0	R0
12	R5,831,430	R35,465	R0	R0
13	R5,504,697	R35,465	R0	R0
14	R5,171,990	R35,465	R0	R0
15	R4,833,199	R35,465	R0	R0
16	R4,488,212	R35,465	R0	R0
17	R4,136,918	R35,465	R0	R0
18	R3,779,200	R35,465	R0	R0
19	R3,414,940	R35,465	R0	R0
20	R3,044,020	R35,465	R0	R0
21	R2,666,318	R35,465	R0	R0
22	R2,281,709	R35,465	R0	R0
23	R1,890,067	R35,465	R0	R0
24	R1,491,263	R35,465	R0	R0
25	R1,085,167	R35,465	R0	R0
26	R671,646	R35,465	R0	R0
27	R250,562	R35,465	R0	R0
28	R0	R20,880	R0	R14,585
29	R0	R0	R0	R35,465
30	R0	R0	R0	R35,465

Same returns reversed

Thirty annual observations for same returns reversed

Year	Capital at year end	Income paid	Reduction by rule	Requested but unpaid
1	R11,786,064	R35,465	R0	R0
2	R11,568,216	R35,465	R0	R0
3	R11,346,384	R35,465	R0	R0
4	R11,120,496	R35,465	R0	R0
5	R10,890,478	R35,465	R0	R0
6	R10,656,253	R35,465	R0	R0
7	R10,417,746	R35,465	R0	R0
8	R10,174,877	R35,465	R0	R0
9	R9,927,567	R35,465	R0	R0
10	R9,675,735	R35,465	R0	R0
11	R9,419,298	R35,465	R0	R0
12	R9,158,172	R35,465	R0	R0
13	R8,892,271	R35,465	R0	R0
14	R8,621,508	R35,465	R0	R0
15	R8,345,794	R35,465	R0	R0
16	R8,065,038	R35,465	R0	R0
17	R7,779,148	R35,465	R0	R0
18	R7,488,031	R35,465	R0	R0
19	R7,191,590	R35,465	R0	R0
20	R6,889,729	R35,465	R0	R0
21	R6,582,348	R35,465	R0	R0
22	R6,269,346	R35,465	R0	R0
23	R5,950,621	R35,465	R0	R0
24	R5,626,067	R35,465	R0	R0
25	R5,295,579	R35,465	R0	R0
26	R5,280,468	R35,465	R0	R0
27	R5,126,759	R35,465	R0	R0
28	R4,875,791	R35,465	R0	R0
29	R3,986,115	R35,465	R0	R0
30	R2,853,513	R35,465	R0	R0

Early losses, lower costs

Thirty annual observations for early losses, lower costs

Year	Capital at year end	Income paid	Reduction by rule	Requested but unpaid
1	R9,322,918	R35,465	R0	R0
2	R8,009,720	R35,465	R0	R0
3	R7,905,560	R35,465	R0	R0
4	R7,938,750	R35,465	R0	R0
5	R8,187,563	R35,465	R0	R0
6	R7,943,833	R35,465	R0	R0
7	R7,694,392	R35,465	R0	R0
8	R7,439,107	R35,465	R0	R0
9	R7,177,841	R35,465	R0	R0
10	R6,910,455	R35,465	R0	R0
11	R6,636,803	R35,465	R0	R0
12	R6,356,741	R35,465	R0	R0
13	R6,070,117	R35,465	R0	R0
14	R5,776,777	R35,465	R0	R0
15	R5,476,566	R35,465	R0	R0
16	R5,169,320	R35,465	R0	R0
17	R4,854,877	R35,465	R0	R0
18	R4,533,066	R35,465	R0	R0
19	R4,203,716	R35,465	R0	R0
20	R3,866,650	R35,465	R0	R0
21	R3,521,686	R35,465	R0	R0
22	R3,168,641	R35,465	R0	R0
23	R2,807,324	R35,465	R0	R0
24	R2,437,543	R35,465	R0	R0
25	R2,059,097	R35,465	R0	R0
26	R1,671,786	R35,465	R0	R0
27	R1,275,400	R35,465	R0	R0
28	R869,727	R35,465	R0	R0
29	R454,550	R35,465	R0	R0
30	R29,646	R35,465	R0	R0

Early losses, income cuts

Thirty annual observations for early losses, income cuts

Year	Capital at year end	Income paid	Reduction by rule	Requested but unpaid
1	R9,276,069	R35,465	R0	R0
2	R7,927,507	R35,465	R0	R0
3	R7,843,623	R30,399	R5,066	R0
4	R7,897,653	R30,399	R5,066	R0
5	R8,167,781	R30,399	R5,066	R0
6	R7,945,679	R30,399	R5,066	R0
7	R7,719,516	R30,399	R5,066	R0
8	R7,489,217	R30,399	R5,066	R0
9	R7,254,708	R30,399	R5,066	R0
10	R7,068,975	R26,056	R9,409	R0
11	R6,879,846	R26,056	R9,409	R0
12	R6,687,258	R26,056	R9,409	R0
13	R6,491,149	R26,056	R9,409	R0
14	R6,291,454	R26,056	R9,409	R0
15	R6,088,107	R26,056	R9,409	R0
16	R5,926,527	R22,334	R13,131	R0
17	R5,761,991	R22,334	R13,131	R0
18	R5,594,447	R22,334	R13,131	R0
19	R5,423,840	R22,334	R13,131	R0
20	R5,250,113	R22,334	R13,131	R0
21	R5,112,195	R19,143	R16,322	R0
22	R4,971,756	R19,143	R16,322	R0
23	R4,828,749	R19,143	R16,322	R0
24	R4,683,126	R19,143	R16,322	R0
25	R4,534,841	R19,143	R16,322	R0
26	R4,417,261	R16,408	R19,057	R0
27	R4,297,532	R16,408	R19,057	R0
28	R4,175,613	R16,408	R19,057	R0
29	R4,051,465	R16,408	R19,057	R0
30	R3,925,046	R16,408	R19,057	R0

Download the retirement scenarios (Excel)

Compare the capital and income outcomes, select a fixed scenario and inspect the year-by-year calculations. Fixed illustrative examples, not a personal retirement calculator.

What changes the answer?

A longer retirement. Test periods that extend beyond an average lifespan. A surviving partner may need income after the other partner dies. A thirty-year horizon is a scenario to test, not a prediction of either person's life.

Income that cannot be reduced. A spending-cut policy has limited usefulness if nearly every expense is essential. Compare suitable guaranteed-income options for that part of the budget, paying attention to escalation, spouse benefits, guarantee periods and access to capital. Obtain actual like-for-like quotes. A higher initial income may buy less protection later.

Other income and accessible reserves. Contractual income can reduce dependence on portfolio withdrawals. Cash can help meet near-term spending without an immediate sale of risk assets, but it has an opportunity cost and must eventually be replenished.

Costs and tax. Compare the full cost stack and the tax treatment of each withdrawal source. A lower fee helps only within an otherwise suitable arrangement; changing products can itself create costs or surrender valuable benefits.

The ability to work longer or spend less. These can change the capital required without assuming better investment returns. Test changes the household could actually sustain. A plan that depends on permanent cuts nobody would accept is not a workable plan.

ASISA's living-annuity standard explains that the customer bears investment and longevity risk and that the selected income level is not assured for life. A permitted product choice and an affordable first-year withdrawal are different tests. ASISA Standard on Living Annuities, sections 1 and 5.1

A useful retirement target has conditions attached

Write down the after-tax spending target, income sources, accessible capital, tax basis, cost assumptions and investment horizon. Add at least one adverse path and the action you would take if it occurred. Review that record against actual spending and balances, and after a material change in health, family circumstances or income.

The judgement is to prefer a range you can explain and adapt over a precise number whose assumptions remain hidden. A larger capital target is not always necessary when dependable income already covers essential spending. A smaller target is not defensible merely because a spreadsheet assumes stronger returns.

For the next questions, read safe withdrawal rates, sequence risk and living versus guaranteed annuities. The existing Retirement Planning Toolkit can help explore stated assumptions. If your decision needs a joined-up review of the household's arrangements, see the scope of a Financial Clarity Review.

Sources and calculation basis

Tax rules above relate only to the 2027 tax year. The ASISA document states a last update of 30 October 2024. The research cited was published in 2016. Sources were accessed on 6 September 2026. All scenario assets, returns, costs and spending rules are original hypothetical inputs. The accompanying annual data and methodology permit reproduction; no client outcome is presented.

General education only. Personal retirement, investment and tax decisions require the household's actual facts, current product terms and appropriately scoped advice.