How much income will your pension replace?

It is generally accepted that we do not need an IRR of 100%, to preserve our standard of living. Approaching retirement, our fixed living expenses tend to fall as our debt burden shrinks, children leave home and we no longer save for retirement. The lower annuity income also attracts a lower average tax rate.

But what IRR should we target? The recommendations vary between 60% and 80%. The appropriate number is critical as it defines the investor’s retirement liability, the corresponding investment target and the appropriate saving and investment strategy.

At 10X, we advocate a minimum IRR of 60%; investors can be highly confident of achieving this goal if they save 15% of their income for 40 years. In fact, assuming they earn a 4% net real return (after fees and inflation) over their entire 40-year savings term (and that their salary grows at 1% pa in real terms over this period), they would achieve an approximate 78% IRR.  Future investment returns are not certain however; we should therefore allow for a margin of error in our savings plan.

But this formula may provide a false sense of comfort, as it refers to income, not take-home pay. This income – pensionable salary – is defined by the (specific) fund rules. The Income Tax Act does not prescribe how pensionable salary should be defined, other than that it must be between 60% and 100% of total annual remuneration. So the employer has some lee-way.

It is mandatory for new employees to join an existing retirement fund. This could deter prospective employees who view retirement saving as a cost rather than a benefit. Organizations may therefore define pensionable salary at the lowest-permitted level, or allow employees to select their own “pensionable salary base”, to minimise the impact on the employee’s take-home pay.

What they fail to appreciate is that their definition of pensionable salary also defines the employee’s post-retirement life style, or more pertinently, their change in life-style. Lowering the “pensionable salary” base magnifies this drop in life-style as it inflates the take-home pay and deflates the pension pay-out. And it is a poor trade as the income drop in retirement is far greater than the pre-retirement gain in take-home pay. We demonstrate this by way of examples.

Base case example – saving 15% of total cost to company

Our base case in Fig 1 derives the take-home pay at various income levels. Pensionable salary is defined as 100% of total cost to company (column A). Employees “lose” 5% to taxed risk and medical aid premiums and the employer contributes to the pension fund at a rate of 15%. But many employees will disregard all this detail, and focus on the number that really matters to them: their money in the bank (column I).

Fig 1: Base case – saving 15% of total cost to company (R000s)

Into what take-home pay replacement ratio does this translate? In our example, we assume that the employee buys a conventional annuity with their fund credit. We further assume that this annuity replaces 60% of their final pensionable salary, and that they spend 3% of this annuity income on medical aid premiums (which is tax-deductible). The net annuity income is then taxed per the (present) income tax table, with the pensioner qualifying for the additional R6,750 rebate for over 65’s.

Fig 2: Base case – resultant pension income (R000s)

Column O on the right shows the replacement ratio based on the net take-home pay. Even though the annuity only replaces 60% of pensionable salary, the impact on take-home pay is cushioned, and pensioners, on average, only lose some 20% of their take-home income.

Saving 15% (or less) of less than 100% total cost to company

What happens when employees base their contribution on less than 100% of their total cost to company? Or if they save less than the required 15% off this reduced base? Initially, they will receive more money in the bank every month, but the gains are relatively modest: compared to someone saving 15% of 100% (of TCT), someone saving 10% of 60% will take home only 10% more money every month.

This affords them a slightly higher living standard pre-retirement. Someone on a total annual package of R170k, for example, would increase their take-home pay by around R1,000 pm; someone on a total package of R1m would take home R4,700 more every month. The gains are surprisingly small, even for those saving less than half the recommended amount (ie at 10% of 60%).

Fig 3: Saving less – rise in take-home pay

But what happens to their standard of living post-retirement? Our example assumes that saving at 15% will afford an annuity that replaces 60% of their pensionable salary. Someone saving at only 10% of pensionable salary will see a proportionate (one-third) reduction in their annuity income, so they only replace 40% of their pensionable income.

The 60% replacement ratio is an estimate of what savers could expect, saving 15% for 35 years, earning a guaranteed 4% net real return (after inflation and fees). We have also modeled a scenario for savers who only save for 20 years, and who, for various reasons (paying excessive fees, too conservative asset allocation, poor fund section and switching) only earn a net real return of 2% pa. Unfortunately, this scenario is the rule rather than the exception in South Africa.

So how much less money flows into these pensioners’ bank account in retirement? In Figure 4, we first compare the savings outcome relative to the base case (the annuity income earned on saving 15% of 100% of TCT). We show the percentage reduction in the take-home pay. In the final (boxed) column, we show the impact of saving 10% of 60% TCT for 20 years, earning a 2% net real return.

The impact is dramatic: those who save 15% off 80% of pensionable salary receive, on average, a 17.5% lower take home income. This number increases proportionally as the savings rate and base declines. Employees saving at 10% of 60% TCT – who only enjoyed a 7% higher work take-home pay – forfeit more than a 50% of their post-retirement take-home pay!

Fig 4: Saving less – drop in annuity income (relative to base case)

But this does not yet tell the full story of the beckoning contraction. In Fig 5, we compare pre and post-retirement take-home pays. This magnifies the percentage reduction, as those who save less have to compare their lower pension against a higher pre-retirement base.

The percentage reduction for those saving 15% of 100% TCT reconciles to the replacement ratio shown in Fig 2 above (column O). From there, the reduction again increases almost proportionally. For example, employees who only save 10% of 60% TCT face a two-thirds reduction relative to their work take-home pay.

Fig 5: Saving less – drop in annuity income (relative to work take-home)

Those who saved little, started late and earned only the 2% pa real return, will on average, face an 85% drop in their post-retirement take-home pay. Unless they lived well beneath their means before retirement, and saved extensively in other ways, they will face a severe reversal in fortune.

Conclusions

We can draw three key observations from this study:

1. A final IRR of 60% reduces your take-home pay (relative to your work take-home pay) by only around 20%
2. Saving at a lower rate – either by reducing the contribution rate or by defining pensionable as a lower number – increases your work take-home pay only modestly – by between 5% and 10%
3. Saving at a lower rate reduces your retirement take-home pay dramatically, by as much as 80%
4. The drop in life-style post-retirement is multiples (around 6 times) higher than your gain in life-style pre-retirement.

We would therefore advocate that you do not attempt to reduce your retirement fund contribution, either by lowering your contribution rate or lobbying to compress pensionable salary; the long-term cost far outweighs the short term benefit. You not only forgo the extra savings, but also the compound return you would have earned on those savings over thirty or forty years.

Some words of caution: replacing 60% to 80% of take-home pay sounds like a good income, and generally it is. Of course, if you have long since paid off your bond, and your children left home ten years ago, and you continued to consume all your take-home pay, then a 60% of 80% replacement ratio may not seem like a good deal anymore. And if your health care expenses grow faster than expected, it may feel as though you now have to live on half of what you used to.

You can avoid such negative surprises, by staying informed and using the appropriate retirement planning tools.  Many employees overestimate the value of their retirement savings, in terms of the (consistent) life style it will afford. The 10X Retirement Calculator will project your approximate final income replacement ratio in current money terms; if you believe this won’t sustain your current lifestyle, you have the opportunity to respond: by spending less, saving more, or perhaps delaying your retirement.

You can also hope for higher market returns, but that would be speculative. As Benjamin Franklin observed, “he that lives upon hope, will die fasting.”