Does an Offshore Portfolio Meet a South African CPI + 6% Target?

An investor might reasonably expect an aggressive portfolio to aim for inflation plus 6% over the long term. But a portfolio can hold mostly shares, carry substantial risk and still have a modelled return below that objective. The label “aggressive” tells us something about its investments; it does not establish what they will earn.

We use an illustrative mix of 80% global shares and 20% gold to examine that distinction. First we compare its expected return with a South African purchasing-power target. Then we explain why expressing it in rand does not add the full South African inflation rate to its dollar return. Finally, we consider why local assets deserve a place in the comparison, without assuming that adding them must improve the result.

Start with the spending goal

Suppose the aim is to grow purchasing power by 6% a year over ten years. Inflation measures the rise in prices; a real return measures what is left after that rise. If savings only keep pace with inflation, the account balance increases but its buying power does not.

Portfolio Lab’s dated model assumes 4.12% annual South African inflation over this horizon. A 6% annual real return would therefore require approximately 10.37% a year in rand before inflation. The calculation compounds the two: (1 + inflation) × 1.06 − 1. A benchmark defined instead as CPI plus six percentage points would give 10.12% using the same inflation assumption. The distinction matters when specifying a mandate, but this example falls short under either definition.

What the model actually expects

The allocation uses AC World Equity to represent global shares, with gold making up the remaining 20%. Both holdings are unhedged, meaning their rand values remain exposed to exchange-rate movements. We keep the holdings and weights fixed. This is an illustration, not an optimized allocation or a portfolio recommendation.

The ten-year central estimate is 8.99% a year in rand, equivalent to 4.68% after the model’s South African inflation assumption. It therefore represents roughly CPI plus 4.7% in real-return terms, rather than the 6% objective. A central estimate is not a guarantee or a probability of success.

The five views help explain the point. The rand return is higher than the dollar return before inflation, but the expected gain in buying power is almost the same. Selecting the row with the largest nominal number does not improve the investments.

The shortfall primarily comes from the underlying return assumptions. In this snapshot, global shares are modelled to earn roughly 5.1% a year above inflation and gold roughly 1.1%. Allocating 20% to gold therefore lowers the expected real return, although it can provide diversification. Gold’s lower modelled return is an assumption to examine, not a claim that its future performance is known.

For scale, R1 million growing at the model’s 4.68% real rate would be worth about R1.58 million in today’s purchasing power after ten years. At 6% real it would be about R1.79 million. These are constant-rate illustrations with no contributions, withdrawals or additional costs, not simulated outcomes. They show why a seemingly small annual gap matters for a future spending goal.

Why not add South African inflation to the dollar return?

The 7.28% dollar estimate already includes US inflation. Adding the full South African inflation assumption would count inflation twice. The model instead removes US inflation and introduces South African inflation:

1.0728 × 1.0412 ÷ 1.0249 − 1 ≈ 8.99% in rand.

This uses a long-term assumption that exchange rates adjust for differences in inflation, known as relative purchasing-power parity. With approximately 2.49% US inflation and 4.12% South African inflation, it implies about 1.59% annual rand depreciation against the dollar. It does not assume the rand repeats a particular historical decline. Actual exchange rates can move much further, or in the opposite direction.

The same currency effect applies to gold. If gold rises 5% in dollars and a dollar buys 4% more rand, gold rises 9.2% in rand: 1.05 × 1.04 − 1. Rand appreciation reduces the local-currency result instead. The modest gap in our table reflects the model’s currency assumption; it is not mainly a penalty from compounding or volatility.

A South African cash rate available today is also a different comparison from a ten-year compound-return assumption. That rate can change. A meaningful comparison with cash needs a ten-year reinvestment assumption and the same treatment of inflation and costs.

What local assets might change

An entirely offshore allocation is only one way to fund rand spending. A comparison with South African shares, nominal bonds and inflation-linked bonds could show whether a broader mix offers a better balance between growth, income and meeting that spending commitment.

Offshore holdings can provide protection when rand weakness coincides with difficult local conditions. Conversely, stronger domestic conditions can support local assets while a stronger rand reduces offshore returns measured in rand. Those relationships are possibilities to test, not a dependable rule that the two sides always move in opposite directions. Global shocks can hurt both.

Nor does a JSE listing necessarily mean domestic economic exposure. Many South African-listed businesses earn substantial revenues abroad. A useful comparison must distinguish domestically focused companies from businesses whose earnings already provide foreign exposure. Allan Gray discusses this distinction in its Balanced Fund commentary.

The appropriate test would compare an offshore-heavy allocation with mixed local and offshore portfolios against the same real-return objective. It should examine stronger and weaker rand scenarios, inflation surprises, historical periods below the target and withdrawals during difficult markets. Local bonds bring interest-rate and credit risks; local shares bring their own business and valuation risks. Diversification can help, but adding local assets does not automatically deliver CPI plus 6%.

This article has not run that allocation comparison. Its finding is narrower: under the stated assumptions, this particular global-share and gold mix falls short of a 6% real-return objective. That is a reason to investigate both the allocation and the assumptions, rather than to raise an expected return simply to match the target.

What to take from the example

Start with the money the portfolio needs to provide, in the currency in which it will be spent. Compare the expected return after inflation with that objective. Then test how different holdings and adverse conditions could change the outcome. A higher nominal rand return, or an aggressive portfolio label, does not answer those questions by itself.

The figures are forward model assumptions, not historical performance. Equity estimates use an August 2026 benchmark profile and a developed-world structural growth assumption; emerging-market growth is not separately calibrated. Gold uses a dollar money-growth and valuation scenario with a 0.4% annual holding-cost assumption. This gold model is not a validated causal forecast. General fund fees, investor taxes and personal cash flows are not included in the table.

Portfolio volatility is calculated from the assets’ volatilities and correlations. It measures return variability, not the amount expected to be lost or the probability of missing the target. Compound portfolio return is approximated from arithmetic returns and portfolio volatility. Because risk inputs are calibrated separately in each currency, small differences remain between the real-return figures. This is not one simulated sequence of future exchange rates.

Correction, 10 September 2026: the original version used inconsistent currency assumptions for shares and gold. Both now use the same inflation-based translation in the app and the dated evidence. This article has also been reframed around the real-return objective; correcting the conversion does not validate the underlying investment forecasts.

Download the dated calculations and source references, or read the base-currency methodology. To explore your own assumptions, open Portfolio Lab and start with your spending currency and objective.