Tax cap on retirement fund contributions: what should high savers do?

To save within or without a retirement fund? Thus far, this question has required little debate. The tax advantages conferred by retirement funds (contributions are deducted at the tax payer’s marginal rate while proceeds are taxed at an average tax rate) generally increase the pension value by as much as 30% relative to a do-it-yourself approach (all else equal).

The Tax Laws Amendment Act, 2013 which passed the first retirement reforms into law, may change this. The new rules take effect on 1 March 2015 and introduce an annual tax deduction cap of R350 000 on retirement fund contributions (including the cost of risk cover).

Individuals who contribute above the cap can deduct these excess contributions from tax in subsequent years, subject to the applicable limits in those years. Any unclaimed contributions are returned untaxed at withdrawal or retirement.

This begs the question: should members invest contributions above the tax limit (“excess contributions”) in a retirement fund (formally), or in a unit trust/ETF (privately)? This question is even more pertinent for provident fund members, who will also be required to purchase an annuity with two-thirds of the value derived from contributions made after 1 March 2015.

So how do these two ways of saving the excess contribution compare?

  1. In both cases, the employee invests after-tax money. However, by saving formally, excess contributions potentially still qualify for a tax deduction in future years, before retirement.
  2. Private savers do not have to purchase an annuity at retirement.
  3. Private savers are not subject to the restrictions of Regulation 28 and can therefore invest their entire above-threshold savings into equities, the asset class that has historically delivered the highest long-term returns.
  4. Investment income is not taxed in a retirement fund; saving privately attracts a 15% dividends withholding tax and income tax on rental or interest income.
  5. Formal savings can only be accessed on withdrawal (resignation or retirement), the private investment at any time.
  6. The retirement cash lump sum related to excess contributions (net of contributions not claimed for tax) will be taxed at 36%; the private savings incur capital gains tax (CGT), at a maximum of 13,2% (25% inclusion rate taxed at 40%).
  7. If the entire fund proceeds are converted into a compulsory annuity, the excess contributions are paid out first (tax-free); if private savings are used to buy a (voluntary) annuity, then a formula is applied to split the annuity proceeds into a taxed (income) and an untaxed (capital) portion.[1]

The private option offers many advantages: it confers greater flexibility before retirement, in terms of accessing funds, choosing the asset mix and not having to buy an annuity at retirement. And the effective tax rates are lower, both on a cash lump sum and any (voluntary) annuity income received.

But the private option has two critical disadvantages. Firstly private investors pay tax annually on investment income, which can reduce the effective return by up to 1% per annum or more, depending on the asset mix. However with a full allocation to equities, this is limited to the 15% dividend withholding tax, which would reduce the annual return by 0.45% only (assuming a 3% dividend yield). This reduction could be offset by earning a higher long-term return due to the higher possible equity asset allocation.

Secondly, the formal savings portion used to buy an annuity is not taxed whereas the entire private investment is subject to CGT, so any annuity must be bought with after tax money.

The increased flexibility attached to private investing must be evaluated in the context of personal preferences and needs. But how do the different tax treatments impact the relative savings outcome?

To answer this question, we looked at two scenarios, a “cash only” option, for members who plan to withdraw in full before they reach retirement age, and full annuitisation (net of disallowed contributions), using either a guaranteed or a living annuity.

Our base case assumes that excess contributions are invested (both formally and privately) in a balanced high equity portfolio which earns a real return before fees of 6% pa plus annual inflation (also 6% pa). We have adjusted future tax tables – including rebates – in line with inflation. We flexed three factors: the investment term, the real return earned, and the type of fund (provident or pension), to see how these affect the relative outcomes.

As the discussion around these scenarios is quite technical, we only present our conclusions, but you can view the full report.

Taking cash

Assuming that both strategies initially earn the identical return (net of fees and withholding tax), the deciding question is how these lump sums are taxed at retirement. In the case of the retirement fund, the gain on the contributions will be taxed per the withdrawal lump sum tax table, at 36%. In the case of the private investment, 25% of the capital gain will be taxed at the marginal tax rate of 40% (an effective tax rate of 13.2%).

On that basis, investing privately is the more profitable option. The higher the “return” component in the fund value, the more profitable the private investment option becomes. In other words, a higher return or a longer investment returns makes private investing more profitable.

This however assumes that these savers apply the same savings discipline, pay similar fees and adhere to same investment strategy as their fund manager would do.

These are big assumptions: private investors invariably pay higher fees than a corporate retirement fund (in terms of unit trust/ETF fees, platform fees and advice fees); they are at risk of emotional investment decisions (timing markets and switching funds on the basis of past performance) or investing too conservatively (a too low equity allocation).

The results suggest that members who have time horizon less than 20 years, and who could approach their private investing with the necessary discipline and financial awareness (earnings a return no more than 2% pa lower) could match their retirement fund return, but with the benefit of greater flexibility. But it would serve those with a longer-time horizon or less discipline to add their excess savings to their retirement fund.

Buying a guaranteed annuity

For our analysis, we assume that members convert their entire excess contributions into an annuity (other than the contributions themselves, which can be reclaimed tax-free); the lower tax rates attached to a cash lump sum would be exhausted by contributions within the R350,000 cap, so there is little tax incentive to draw a further lump sum.

Retirement fund proceeds are not taxed if they are converted into an annuity, whereas the private investment still incurs CGT. If both investments earned the same net return, then the retirement fund proceeds will deliver a higher transfer value and annuity income. But this would be offset by a portion of the “private” annuity income being a return of capital, which would reduce the tax on the annuity income.

Assuming an equal rate of return (of whatever size), fund members would do better investing privately, for all periods – the advantage of the low CGT rate combined with the capital deduction in the voluntary annuity income outweighs the initial tax shield on buying the compulsory annuity.

But the tax advantage is dramatically reduced simply by earning a 1% lower return when investing privately.  All savers are likely to fare better adding excess contributions to their retirement fund if they cannot at least get within 2% pa of their long- term retirement fund return.

Investing in a living annuity

In a living annuity, the investment income is not taxed, only the withdrawals, per the prevailing income tax tables. With private savings, it is the investment income that is taxed, rather than withdrawals. This tax is as follows: a 15% withholding tax on dividends, normal income tax on interest income, and capital gains tax on realised capital gains. This tax reduces the individual’s investment return and the reduction typically lies between 1.5% and 3.0% pa over the first 25 years, depending on the asset mix (a higher equity portion lowers the tax rate).

The lower return is offset by the lower draw-down required, to match the living annuity after-tax income.

In the case of an all-else-being-equal scenario – the initial amount invested, the after tax income drawn down (assuming an average tax rate of 40%) and the after-fee investment income – then the private option is more efficient: it delivers the same income, but with a higher residual capital. In other words, the tax impact on the living annuity withdrawal is greater than the tax impact on the privately-earned investment income.

This advantage persists even if a lower amount is invested privately. So assuming that someone paying capital gains tax on the same private pre-retirement savings they could transfer tax-free to the living annuity, they would still come out ahead in the above all-else-equal scenario.

However, this advantage is immediately eroded by just a 1% higher annual fee (lower return) on the private income plan. This would happen all too easily, by choosing a too-conservative asset mix, paying higher active management fees or making emotional investment decisions.

Summary conclusion

The choice of savings product for “excess contributions” matters. Without the tax shield on retirement fund contributions, saving privately is more tax-effective and can deliver a better outcome, with more flexibility. The longer the investment term, the more pronounced the advantage of saving privately.

But this is on the premise that private investors at least match their retirement fund return (on identical contributions). Only disciplined, informed savers will achieve this. To do so they must save consistently, at low cost and by-pass the general equity restrictions of Reg 28 (thus earning a higher long-term return, enough to compensate for the withholding tax on dividends). And they must adhere to their strategy through-out and not succumb to periodically-outperforming high cost managers or make emotional investment decisions during periods of market volatility.

In practice, only a small minority will be able to tick all these boxes. For those that cannot, the critical question is not which savings product is more tax efficient, but which is likely to deliver a higher return over time. They will find that a well-diversified, low cost retirement fund using time-driven investing and indexing will deliver an outcome that they cannot match privately. They should therefore direct their “excess contributions” to their retirement fund.


[1] The capital portion of a guaranteed voluntary annuity is calculated as follows: divide the lump sum paid for the annuity by the total expected returns of all annuities in terms of the contract, then multiply this factor by the annual annuity income. The total expected return of all annuities is based on the purchaser’s official life expectancy. In the case of a voluntary living annuity or income plan, only the investment income will be taxed (as under point 4 above).

Get investment and saving tips straight to your inbox.

Related articles

Compulsory annuitisation: leaky arguments can’t plug leaky system

Briefing Parliament’s finance committees on 3 March 2016, Ismail Momoniat (National Treasury Deputy...

Delayed retirement reforms should not delay you reforming your retirement

The retirement reforms set to have taken effect on 1 March 2015 have been delayed, possibly until 20...

Latest statement from National Treasury on retirement reform (issued 16 February 2016)

National Treasury issued a statement on 16 February 2016 on the subject of the retirement reforms du...

Get started or switch to 10X today.