Issue: July/August 2005

Defined Contribution Funds - Advantages and challenges

Deon Booysen, Executive Head – Client Solutions at Sanlam Employee Benefits By Deon Booysen:, executive head of SEB: Client Solutions discusses the advantages and challenges inherent to DC retirement provision.

If you had a choice today, would you prefer to be in a defined benefit (DB) or in a defined contribution (DC) fund? Unfortunately for many there is no choice, as most open funds available in South Africa today are DC funds.


DC funds have a number of advantages, both for the employer and for the employee. The attractiveness to the employer is that the cost of these funds is known and that it is no longer influenced by actual returns. It is really only when the expected long-term real rates of return change, both before and after retirement, that an adjustment has to be considered.

For the member, a DC fund also holds a number of advantages, such as the portability of benefits, control over the choice of investments and the fact that good investment returns accrue to the benefit of the member. It therefore implies that poor investment returns also accrue to the member. In effect, in a DC fund environment the investment risk has been passed from the employer to the employees.

How exactly are the contribution rates in a DC environment decided on? Most likely the fund obtained advice, including an estimate of future expected real returns, replacement ratios after retirement, and the terms on which these could be achieved, seeking a contribution rate that (based on these assumptions) would provide an acceptable replacement ratio on average.

This sounds quite acceptable, but let’s consider what on average implies. In a DB environment, it would imply that a member would get his or her benefits. For some employees the cost of these benefits would be cheaper than the average and for others more expensive than the average, but for the employer these differences would average out over time.

This is not the case in a DC environment:

If a particular member’s returns are below average over the period until his or her retirement, he or she will directly feel the impact of this as a lower retirement income. Members in this position should have tracked their savings versus their target and increased their savings for retirement in order to still reach their target. Where a member failed to do this, or was not provided with this information by the fund, he or she would be retiring on a less than adequate pension.

It is a fact that most people overestimate the likelihood of a good outcome and underestimate the likelihood of a bad outcome – which implies that most people expect their particular returns up to retirement to be greater than the average assumed return required to meet the target replacement ratio. Unless trustees of DC pension schemes provide their employees with access to appropriate information, there is a risk that a large proportion of employees will retire on an inadequate income.

Let’s consider the replacement ratio, which is the percentage of your final salary you will receive as a payment every year, from the lump sum saved to retirement. Put differently: a lump sum of say R1 000 000 will provide R100 000 per annum based on a conversion factor of 10. If your salary at retirement was R200 000 per annum, this implies that the annuity replaces 50% of your salary, i.e. a 50% replacement ratio.

In the National Treasury’s discussion paper on retirement reform, it was indicated that assuming that a real rate of return of 3% could be earned over the term of the savings, a contribution rate of 10% of salary towards savings would provide a replacement ratio of 44,1% over 35 years (i.e. if you contributed from say age 30 to 65) and a replacement ratio of 54,5% over 40 years (i.e. if you contributed from say age 25 to 65).

These assumptions are based on a conversion factor of 13,8 for 100% inflation increases for a pensioner retiring at age 65. This factor of 13,8 is probably sufficient for a male aged 65, assuming that there are no expenses. For a female it would be closer to 16.

The problem is that these conversion factors change every day. For example: If real rates reduced by say 1%, these conversion factors could increase by as much as 10%. This is quite an important point, especially in a DC environment. Even if you carefully plan and monitor your retirement savings to reach a particular replacement ratio, you may find that in the last six months of your journey the target may have moved dramatically due to the movement in the interest rates that influence the value of the conversion factor. Likewise, switching to cash would not necessarily be the appropriate way to protect your expected replacement ratio – even though it is true that cash will not lose any capital and will provide some sort of interest growth.

However, in the example above, if real rates decrease by 1%, the conversion factor will increase by 10%, but the value of cash will remain unchanged – leading to a reduction in post-retirement income. Ideally, one would have preferred an investment that also increases in value by 10%. The converse is also true – should real rates increase by 1%, the conversion factor will reduce by 10% (or become cheaper). An investment with a value that also decreased by 10% would imply that the retirement income you could purchase with this investment would again remain unchanged.

This is quite an important concept – when aiming for a certain replacement ratio, the “safe” investment is not necessarily cash, but rather one that moves in line with the conversion factor. Trustees of DC funds should therefore carefully consider the investment choices made available to members, to ensure that a suitable choice is available to those members close to retirement who want to lock in their current replacement ratio.

The question then is what assistance trustees have to be able to aid members in tracking their savings up to retirement. A fund can use a liability index to illustrate how their returns are measuring up against a stated replacement ratio target.

Trustees should consider setting up such a liability index as part of the measurement criteria to check the extent to which the investments are meeting the objectives of their investment policy statement. Of course, where funds offer individual member choice, an average index would not be appropriate. In such instances members should be provided with access to more advanced financial planning that would address their unique circumstances.