COVER STORY: Editorials: Edition April / June 2019


Wake up! Pay up!

Were it only so easy.
Proper thought must be given, and appropriate action taken,
over the R50bn of pension assets not being used as they should be.

Terms have meanings, frequently emotive and rarely neutral for conveying an impression.

“Unclaimed benefits” and “unpaid benefits” can be used interchangeably, but they carry different connotations. “Unclaimed” implies that the obligation rests with former members of pension funds that they get off their butts to demand entitlements. “Unpaid” denotes a responsibility on the fund, the employer or the fund administrator to find beneficiaries who’re owed payments.

“Unpaid” is preferable because there are at least 4,5m former members and dependents due for some R50bn in pension benefits. Many of them — including, for example, spouses and minors eligible to receive pay-outs on a member’s death — haven’t a clue that monies await them.

When these people are poor, often pitifully so, in urban and rural areas across SA and in neighbouring states, the human destitution is heightened. A few thousand rand in these peoples’ pockets is meaningful.

There are facilities to initiate the claims process. But feel for the inquirers, especially of low literacy, who attempt to use the search engine provided on the website of the Financial Sector Conduct Authority; first to know that there is such a facility; second to have the internet tools and skills to use it; third to provide such details as an ID number (of a migrant?), the name of the fund (really?) and the name of the employer (accurately?).

Only then will the database be able to match these details with the central database, and only then to supply the inquirers with contact details for the relevant fund or administrator so that they can follow up. Wish them luck as they try to submit a claim without easy and cheap access to a photocopy machine, a commissioner of oaths and telecommunication facilities.

The fault lies not with the FSCA search engine for it’s an honest endeavour to assist members as best it can through the tracing function on its website. By end-June last year, it had received 45 000 enquiries of which 8 000 possible matches were identified to a value of R1,8bn.

The problem has myriad causes. National Treasury has identified several: fund members not being required to supply their funds with the necessary information to facilitate pay-outs once they leave their jobs; employers not providing correct and updated information to the fund; insufficient effort being made to trace beneficiaries, and poor data-management systems at numerous fund administrators.

To this mournful list can be added such other factors as employers failing to inform employees that they’re members of retirement funds in terms of their employment contracts, the names of their funds and the fact that the funds are separate entities that should survive even if the employer folds; let alone for employees being requested to provide details of their beneficiaries in the event of death or disability claims.

So how is a fund or administrator supposed to respond when it hears from a person purporting to be a member but who cannot supply evidence to prove it? Or when the person has a common surname and an identity number or even a birth date that doesn’t match records?

Bluntly, there is no single correct answer. Worse, the environment is ripe for fraudulent documentation and phoney offers from crooked operators demanding upfront fees to assist desperate people.

Stark numbers speak for themselves. A 10-year trends analysis by the then Financial Services Board shows that in 2007 there were 3 753 funds with 9,3m members and slightly over R1 trillion in assets. Of these there were 1 859 funds which had 1,06m members owed R5,8bn in unpaid benefits.

By 2016 the picture had deteriorated markedly. There were 2 053 funds with 14,5m members and assets of R2,4 trillion. Of these, there were 1 327 funds which had 4,2m members owed R42,4bn in unpaid benefits.

This isn’t the end the sad story. Not only would today’s total for unpaid benefits exceed R43bn but they’d be only for those funds supervised by the FSCA. Thus excluded are the huge Government Employee Pension Fund (which alone has reported unpaid benefits of R9,2bn) and the Transnet funds, amongst others.

It’s nonsensical, national priorities for new infrastructure projects and social benefits being what they are, that more than R50bn should be left to accumulate in anticipation of the full amount eventually being successfully claimed. This simply won’t happen.

A proposal is therefore that the amount across the spectrum of institutional and employer-based unpaid benefit funds (UBFs) be consolidated into a central not-for-profit UBF, supported by a central database into which government departments – such as SA Revenue Services, Social Security Agency and Home Affairs — can input relevant identification details. This UBF would be managed jointly as a public-private partnership.

Another proposal is that one of the existing funds, which is effectively already a UBF, be taken over by the state and run as a fund to which closing funds could voluntarily transfer unpaid benefits. Even if the UBF takes all reasonable steps to find and pay all those entitled to the unpaid benefits, it will never be able to find and pay them all.

Then, the argument goes, it could justify holding assets with a value less than the full face value of its liabilities. Particularly if the fund is underwritten by the state, no beneficiary will be deprived of money due.

The balance could be released to finance new development and social programmes. Private-sector participation in governance and oversight is essential, just in case a strained fiscus becomes tempted to fill old holes.

Necessity being the mother of innovation – and necessity being acute – a centralised UBF could offer the further advantage of supervision under the Pension Funds Act requiring that funds supply information by which members can be identified and traced.

Only in 2009 were the regulatory reporting requirements for funds under the Act amended to require all funds to include reports on their unpaid benefits in their annual financial statements. Previously there’d been no provision for them separately to disclose memberships and the assets related to unpaid benefits.

Two years ago, the FSB noted in a September 2017 presentation, it had proposed amendments “to provide for a Central Unclaimed Retirement Benefit Fund to house all unclaimed retirement benefits and possibly be expanded in the future to include other sectors such as insurance companies and banks”.

Meanwhile, endemic problems remain. The amount of unpaid benefits due to a fund member is in many cases gradually chipped away by administration, asset-management and tracing fees deducted from individual ‘member accounts’ rather than from a fund-wide expense account.

Kotze . . . better and better

Not uncommon is that amounts held to provide benefits for people who can never be traced (due to bad records, for example) remain available in an UBF as ‘assets under management’ until depleted. Instead, they could cover expenses attributable to those who can be traced.

Unacceptable are arrangements that create disincentives to trace. They’re suspected to enter the mix usually when relationships between fund trustees and service providers are incestuous.

However, there is a brighter side. Databases are evolving as company records, perhaps as a function of labour legislation, are improving. If only under regulatory pressure, communications with fund members are refining. When a member switches from one employer to another, there’s a better chance of the move being recorded for a tracing agency to follow.

Moreover, at least some institutions have become increasingly proactive. A case in point is Liberty, historically the most impacted by the sheer volume of funds it had gathered. Under testing circumstances, where each fund requires examination, Liberty’s behaviour on the treatment of UBFs presents an illuminating case study.

Van der Merwe . . . unique usage

Its clean-up started off the back foot with the large number of messy employee benefit (EB) books taken under its wing. Some years ago Liberty bought Capital Alliance which had previously acquired the EB books of AA Life and Stangen. Then it took over Investec EB which included the Fedsure and Norwich EB businesses. Big size, shooting Liberty to the uppermost range in terms of fund numbers being administered, caused big trouble in terms of flawed records.

When the FSB commenced its ‘cancellations project’ in 2008, there were more than 6 000 funds which had no boards of trustees. Of these funds, about 80% were on Liberty’s books. Some had no assets or liabilities, but others did. Liberty has been painstakingly tracking the latter.

So far it has stood alone in having obtained a High Court order for the cancelled registrations of 25 ‘dormant’ funds to be reversed. Because these funds were discovered to have assets and liabilities, their registrations should not have been cancelled (TT Feb-April ’18). The Liberty decision voluntarily to commence reversals, by court applications, shortly preceded the landmark Concourt judgment on the FSB’s so-called ‘cancellations project’ (TT Nov ’18-Jan ’19).

That Liberty didn’t await this judgment, on the gamble that a different outcome could have averted a legal compulsion on the FSB/FSCA to reverse erroneous cancellations across the board for all administrators, speaks to a moral sensibility.

“Most important is that benefits are paid to members,” says Liberty Corporate chief executive Tiaan Kotze. He won’t be going to court for reinstatement of other funds possibly deregistered in error but instead “will be working with the regulator to ensure that proper processes are followed”.

The processes will be keenly monitored both by Liberty and other administrators of erroneously-cancelled funds as well as the Right2Know Campaign, the Casual Workers Advice Office and the Sebokeng-based Unpaid Benefits Campaign.

It’s unclear how the flaws can be remedied without court orders of the type for which Liberty had applied. In fact, on March 4 the FSCA suddenly issued a circular that effectively endorsed the Liberty route. The circular instructs administrators, who’d been responsible for erroneous cancellations before April 1 2018, to apply to court for these cancellations to be set aside (see Currents).

Having investigated a sample 500 funds of the 4 600 funds subjected to the FSB’s ‘cancellations project’, KPMG found that the FSB had insufficient information on the disposal of their R2,5bn in assets. In the months since the Concourt judgment last September, remedy appears still to elude finality.

Nonetheless, although the handling of the erroneously-cancelled funds is obviously important, Liberty’s R110m so far detected in some 130 ‘dormant’ funds (including the 25 re-registered) is tiny against the private-sector industry’s R40bn-plus total for unpaid benefits. Kotze reckons that Liberty’s share of unpaid benefits amounts only to around 6% of the industry issue.

As part of its clean-up, Liberty appointed two independent trustees to its UBF. One is a former unionist with extensive retirement-fund experience; the other has long served in asset-management firms. No longer on the UPF board is there a Liberty employee.

Then too, Liberty has re-looked its pricing model. In September 2017 it stopped charging an administration fee to any member whose benefit is less than R800.

Better processes – which come at significantly increased capacity-building cost, borne by Liberty – mean better ability to pay claims. The tumultuous process involves taking somebody verified by a credit agency and getting the person’s dependents to provide documentation that will show the correct benefit for the correct person. The process is supplemented by engagement with community-based organisations and education initiatives.

At the beginning of last year the Liberty UBF had assets of more than R1bn. After investment returns and R175m in payments, it stood at a little more than R900m. Individual balances average less than R10 000 per member, and more than 35% of members have balances under R800. Payment of claims has improved over a year from around 500 to 5 000 people a month.

The major bottleneck is in the ability to trace, Kotze points out: “Of approximately 90 000 members in the UBF, tracing agents are struggling to trace around 27% of them. Where tracing is expensive for the 35% of members with small balances, we’ve been running sms campaigns with some success.”

A curious phenomenon is that, of those successfully traced, roughly 5 000 have declined to submit claim forms. The reasons beg an industry-wide examination to establish, for example, whether members might be fearful of beneficiaries learning about monies due. If this is a factor, then member communications and unpaid benefits have yet another problem.

Also speaking to improved tracing abilities is Fedgroup, pleased with having been awarded the tender by Argen Actuarial Solutions to find unpaid members in the now-liquidated IF umbrella pension and provident funds (TT Aug-Oct ’18). The funds have about R180m for distribution.

Fedgroup Life chief executive Walter van der Merwe is confident of his division’s ability to satisfy the tracing mandate by virtue of its setup for administration of beneficiary funds: “We’re merely extending our capability to a similar process of maintaining a relationship with beneficiaries.” The exercise for beneficiary funds is “vastly different” from pension funds, he suggests, because the latter are mainly concerned with active members.

Van der Merwe explains that Fedgroup’s tracing method starts with information available electronically, moves on to databases and continues to inquiries with former employers. Then it proceeds to visiting places or work and last-known home addresses.

The higher the volumes that pass through its systems, the lower the costs to members. There are no fees for unsuccessful traces.

“Unique is not our technology but the way we use it across the spectrum of financial products,” he explains. “They all require bank interactions.”

Main competitor to Fedgroup in the beneficiary-fund space is Fairheads, necessarily active in tracing, where director David Hurford argues for a centralised administrator of unpaid benefit funds that are separate from fund investment management: “It would charge a modest monthly administration fee, rand-based and not a percentage of the benefit, as well as a termination fee once the member has been traced. The administrator’s revenue stream would therefore depend on finding people.”

What needs to happen, one way or another, is a centralised database that can draw on multiple sources exclusively for its confidential use. There’s a wealth of invaluable contact information held not only in government agencies but also in customer records of retail groups and cellphone operators. Take it as a thought for the FSCA to action, along the lines of the proposals already presented.

Clearly, from what’s being done and can be done, the problem of unpaid benefits is not intractable. Not by a long shot, provided that the public and private sectors collaborate for innovation. That R50bn hangs in the air can be justified by neither, and least of all to fund members.


Michael Prinsloo, head of employee-benefits consulting strategy at Alexander Forbes, comments:

We are not going the Liberty route for the reinstatement of closed funds. Our situation is different from Liberty where closed funds still had assets.

Unclaimed benefit funds, and occupational benefit funds, have typically implemented policies around unclaimed benefits and gone through extensive efforts to trace individuals, then to pay them once traced. Alexander Forbes’ funds are no exception.

At this stage it’s unclear what more can be done as in many cases much time has elapsed. Some unclaimed records result from surplus allocations dating back to 2001. Funds have run multiple tracing attempts but the problem is that 15 years ago funds did not receive ID and contact numbers etc, so the data on older records is patchy.

The Mines 1970 UBFs have put themselves forward as a success story in terms of finding members and paying them over the past few years (TT March-May ’17). They have a tighter pool in which to fish, being largely ex-Chamber of Mines employers who were involved, so it is limited.

Given current legislation, in particular default preservation and changes in tax law which set out when a benefit accrues, theoretically in the future we’ll have far fewer ‘technical unclaimed benefits’. They’ll still happen but will be called something else. So the visible problem will shrink but it’s uncertain whether the actual problem will go away.

In the surplus-legislation aftermath there were funds that submitted valuations to write down the value of the liabilities associated with these types of benefits and essentially release those assets for other purposes (essentially working on the liability being equal to the fund credit calculated into a probability of payment). This was met with resistance from regulators.

But I do think there is merit in considering an approach where a portion of the money is used for specifically identified cases e.g. social development and impact investments. The trustees of UBFs could conceivably follow such investment strategies today, but many would be too nervous to take on risky investments of this nature.

However, if the liability is reduced it may be more palatable for these trustees to “take risk” with essentially surplus assets.