In the shift from defined benefit (DB) to defined contribution (DC) funds, many of the risks associated with retirement funding and medical cover have shifted from the shoulders of employers to fund members. More importantly, with employers stepping away, financial services companies now assume a more central role. Perhaps we need to consider new ways of approaching retirement funding, one of which is based on the principles used in a DB fund.
In Benefits Barometer 2013: The employee benefits system - 'Retirement funding' our primary focus was to assess how well members of South African retirement funds were faring, given the dramatic structural shift from the defined benefits (DB) to defined contribution (DC) model. Our answer was: not very well.
DB funds worked when the population they served was substantially more uniform than it is today and only when individuals had long service periods.
A key issue was that DB funds and postretirement medical aid benefits were reflected on the employer’s balance sheet. With rising costs, reducing interest rates and volatility of funding levels, many employers simply did not have the appetite to keep this level of risk on their books. They also had to maintain liquidity for paying lump sums on death or withdrawal. And if employees lived too long, employers had to find a way to fund this. Clearly the model had its flaws.
Given the evolving demands of the changing world of work, DB schemes may no longer be viable. That leaves us with the challenge of refining the newer DC model so that it better serves the needs of both the current and future working population of South Africa.
Given the evolving demands of the changing world of work, DB schemes may no longer be viable. That leaves us with the challenge of refining the newer DC model.
One of the most neglected aspects in the conversion from DB to DC funds is that it subtly changed the intermediation of retirement funding1. Even though financial services have always been a key part of retirement funding, the shift to DC funds moved the bulk of the intermediation from employers to financial services companies.
Under older models, responsibility for an employee’s welfare all fell to the employer. Welfare included retirement savings, risk benefits and pre- and post-retirement medical aid (the most valuable retirement benefit of all, given the constantly rising cost of medical care). In the event of any shortfall, the employer would meet the additional costs. People knew what income to expect in retirement, and as long as the company survived and they remained with the same employer until retirement, that is what they would get. It was a relatively paternalistic system with the power and responsibility resting with the employer.
With the DC model, power is channelled through trustees or management committees within umbrella funds. More often than not these fiduciaries are guided by the financial services industry. Trustees make many of the decisions relating to fund and benefit structures, but some critical ones rest with the employees themselves, again being guided by the industry.
The crux of the problem is that the employer no longer has a full view of what is being presented to their employees, creating a significant human resources management problem. Instead, they are faced with multiple service providers and sets of fiduciaries all trying to address pieces of the problem. Integrating this into a cohesive picture of their employees’ financial, physical and mental health is extremely challenging.
For example, the provision of a medical scheme, along with related HR decisions such as sick leave, remains within the ambit of the employer, at least while the individual is still employed. But setting the risk benefit parameters typically falls to the trustees or management committee members of the retirement fund. The potential for both gaps and significant duplication between those two support systems is real, along with the lingering question about who should be held accountable for making it all fit together.
The consequences of these discontinuities fall on the employer, not on the financial services company:
The great irony here is that interviews with members have highlighted that many employees believe their employers are still looking after their interests and believe that these benefit arrangements are sufficient, when they are in fact insufficient. However, employee benefits are often seen as providing base cover, with employees being left with the responsibility of making any adjustments if it is insufficient. But individuals aren’t always aware of this and more often than not, do nothing.
Under a DB arrangement, employers promised retirement fund members a benefit often based on a percentage of salary and linked to the number of years of service. Unlike DB funds, DC funds do not offer a guaranteed income after retirement. In addition, where employers used to take responsibility for providing medical aid post-retirement, many no longer do and the responsibility of paying for medical care during retirement rests with members.
Simply put, DC retirement funds offer members a tax-shielded savings account that can be used to purchase an income after retirement. As a consequence, ensuring an individual will have sufficient income in retirement is a highly uncertain exercise. There are no guarantees and members bear most, if not all, of the risks once borne by the employer. In particular, members are exposed to2:
Independent actuary and retirement fund researcher Rob Rusconi provides calculations that compare how significant these risks are to a member when we compare the DC structure to the earlier DB structure3. In interpreting this graph, the DC benefit is expressed as a percentage of the equivalent DB benefit: 150% = very good, 100% = breakeven and 70% = poor.
Clearly, investment risk represents the bulk of the risk burden in a DC scheme, with a range of outcomes that can be as high as 263% to a low of 36% of the DB benefit. This certainly provides unequivocal evidence that trustees and management committees need to make communicating the potential magnitude of these risks to members a top priority.
Employee benefits are often seen as providing base cover, with employees being left with the responsibility of making any adjustments if it is insufficient.
But the chart above also illustrates that many funds may not have thoroughly thought through the potential impact of the salary growth and mortality risk on their retirement fund outcomes.
Note that the investment risk shown in this graph includes both returns pre-retirement and the risk of low yields at retirement.In targeting returns pre-retirement, the problem isn’t as simple as producing maximum returns.
Members would benefit from some certainty about the incomes they are likely to generate after retirement. What this means is that whatever strategy the member plans to use after retirement for income provision needs to be reflected in the pre-retirement strategy for a few years prior to retirement. In particular, this strategy needs to be responsive to the changing cost of the relevant annuity so that the member can have a reasonably stable expectation of what income to expect.
Irrespective of the type of annuity chosen, declining interest rates generally result in a reduced level of real income. The chart on the next page shows the averaged inflation-linked annuity prices from several providers to illustrate that falling interest rates have reduced the inflation-linked income that a pensioner can buy. The average actual pension, in nominal terms, that R1 million could purchase has fallen from just over R3 700 per month to just over R3 200 per month – more than a 13% decrease – from August 2007 to February 2014.
In some ways these results emphasise the reality that we may not yet have found the optimal alternative to DB funds and that work on this point is probably still required. Perhaps a more optimal solution would be one where some measure of protection was provided while at the same time, risks might be shared between employers and employees4.
For example, Rusconi suggests that funds could target paying out a lump sum to members at retirement. A formula would be applied that would smooth out the investment returns. In this arrangement, the investment risk would remain with the fund which would have to pay out a specified lump sum, but the mortality and salary growth risks would remain with the member. In the US, this type of arrangement is known as a cash balance plan.
As an industry, perhaps more focus needs to be directed at either finding more explicit ways to mitigate risk or providing a better educational framework for helping individuals understand what they can do for themselves.
The most critical risk is planning risk
While the risks mentioned previously are frequently cited in the literature on DC funds, planning risk is often ignored. In many ways, planning risk encapsulates all the other risks faced by individuals because it speaks to the ability to make the decisions most likely to get the individual to their target.
For retirement funding, a whole series of decisions need to be made – pensionable pay, contribution rates, retirement age, pre- and post-retirement investment strategy, and annuitisation – to maximise the individual’s chances of retiring with an adequate income. Making good decisions at the outset, and then knowing when and how to adjust them, is a critical aspect of planning. Planning risk is the risk that choices made by the individual would result in an inadequate income in retirement.
Getting all of these inputs correct is complex, riddled with risk and long-term modelling complications. Because members don’t understand the implications of this, they often make poor decisions.
In addition, because employees are likely to work for multiple employers over their lifetime and be a member of multiple funds, stitching it all together remains in the individual’s hands.
Members would benefit from some certainty about the incomes they are likely to generate after retirement.
How can intermediaries help?
Can we make a DC fund ‘behave’ like a DB fund? The fundamental thinking behind both types of retirement funds is the same: the task is to spread the income earned during your working life over your entire life5. The experience gained in managing DB funds can still help DC funds.
a. Setting the contribution rate
In DB schemes, the employer determines the benefit they would like to offer their employees when they reach retirement. This is usually set in proportion to the length of service and the member’s salary at retirement. Once the pension promise has been set, the actuary would determine the contribution rates that would have to be paid to reach that goal.
Similar thinking can be used in a DC fund. If we set a reasonable goal for income in retirement, we can calculate the contribution rate that will get us there.
b. Adjusting contribution rates over time
A good DB fund was close to, if not fully, funded at all times. In other words, it generally stayed on track. As investment returns went up and down, plan sponsors made up the shortfalls when they occurred.
Members in a DC fund can also ensure they remain on track by adjusting their contributions when necessary and possible, given affordability.
There’s little room for wishful thinking about the investment markets when funding a retirement benefit6. Individuals must adjust their savings rates if they are falling short of their goals. Putting your retirement savings on autopilot and just hoping things will come right simply isn’t an option.
c. Finding an appropriate payout strategy
In a DB fund, the member obtains an income for the remainder of their lives, with the employer picking up the tab if individuals live longer than expected.
In a DC fund, members need to be as concerned about their withdrawal strategies as they are about accumulation strategies. Retirees can opt for the certainty of a guaranteed income for the rest of their lives, but this is often costly and can mean a substantial decrease in income. To maintain their living standards, they could take a riskier annuity option, such as a living annuity, but this exposes them to the risk of outliving their income.
We are only just beginning to appreciate the burden that has been placed on the employee of managing the risks that were once addressed by the employer. But the purpose of a DC fund and a DB fund is essentially the same: to provide an adequate source of funding when individuals are no longer earning a salary.
However, the intermediation of funds has changed. If members are to get full value from the DC model, experts need to refocus on mitigating the financial planning risks that are embedded in DC funds to provide an adequate source of funding when individuals are no longer earning a salary.
1 Sexauer & Siegel (2013)
2 Future of Employee Benefits Members Survey
3 Rusconi (2004)
4 Rusconi (2004)
5 Sexauer & Siegel (2013)