GEPF, budget austerity and the Eskom debt crisis: The R1.8 trillion and growing Elephant in the Room, Part 1:
GEPF, budget austerity and the Eskom debt crisis
Finance Minister Mboweni’s Special Appropriation Bill added R59bn over two years to Eskom’s capitalisation. The government will now have even higher debt service costs.
Given the alternative to negotiate debt relief with a main creditor that in fact is an organ of the state, this choice of policy was irrational already in February when the Treasury allocated R69bn from the national budget over three years to Eskom’s debt servicing. It is, arguably, also unconstitutional, given that the government’s constitutional mandate is to gradually improve the lives of all “within its available resources”.
A resource available to the government in an objective sense can however be politically or even ideologically disregarded, or seen to be politically available for one purpose, but not for the other. This, we will argue, is the case with the Government Employee Pension Fund (GEPF), which funds are managed by the Public Investment Corporation (PIC) that is under investigation for patronage and corruption.
In the 2017/18 fiscal year, the government paid an estimated R25-30bn in interest on its R359bn debt to GEPF. GEPF owned 14 percent of the gross loan debt of the government.
Eskom’s debt to the GEPF in March 2018 was R87bn. This was 20.8 percent of Eskom’s whole R419bn debt last year. The GEPF probably earned a market rate interest of some R8bn on this investment, because of Eskom’s low credit rating.
At least 45 percent of the GEPF’s R72bn cash income from investments in 2017/18 came from the government itself and from the state owned Eskom. A table over GEPF cash in- and outflows is published below.
Does the R1802bn large GEPF (March 2018) need these income streams to pay government pensions and benefits to members at the guaranteed levels? Should the GEPF’s investments in equity be redirected to control the crisis? What is the substance of the GEPF’s “long-term funding shortfall” that gave headlines earlier in the year? Why is the government as big debtor not speaking to itself as its biggest creditor?
From the “funded” perspective adopted in the Government Employee Pension Law of 1996 (GEP Law), the funds of GEPF need cover 90% of all liabilities at any point in time. The actuarial audit of GEPF concluded that on 31 March 2018, the GEPF had R1 802 billion in financial assets managed by PIC. It covered 108% of GEPF’s liabilities to working and retired members.
This means that the market value of GEPF’s financial assets could have been R304bn less in March 2018. It would still have met the 90% legal requirement. At a 100% funding of liabilities, which is the political goal of the board, the surplus was R137bn.
Why does the law stipulate a mere 90% “funding level”? Mustn’t all pension obligations be paid? They have to be paid, but not all of them on the same day.
The statutory audit controls if the market value of the fund’s financial assets at a certain day is enough to pay “the present value” of past and estimated future liabilities to those who are members on that day and until they die. If it is, then the scheme is 100% fully funded. The “funded” pension scheme must have all the money in advance, like a well-paid but lonely individual who calculates how much she should save until she stops working and will start to withdraw money from an account until she dies.
In contrast, the so called “Pay-as-you-go” pension system uses the fact no pension scheme on a certain day has to pay all what it owes and will owe to all its members. Every month “My pension is, because others’ contributions are”. Pension liabilities are paid bit by bit over a time span of 30-50 years. Before neoliberal economic policies became dominant in the 1980s, pay-as-you-go was standard policy for state pension schemes in countries with strong labour movements. Surpluses built buffer funds as protection against shocks and financed investments in social infrastructure.
But no matter how it is constructed, no pension scheme is actually paying the theoretical “present value” of all present and future liabilities to those who are members on a certain day. In this world, it never requires that theoretical “100% coverage”.
This is why South Africa’s GEP Law can set the minimum to 90% funding “in advance” of liabilities and why, for example, the legal demand on the funding level of a private pension scheme in the US is put at 80%, or why the credit rating agencies Standard & Poor or Fitch only regard a funding level as “weak” if the funding level is below 60% in a public pension scheme like GEPF. The “underfunding” allowed for in the GEP Law is a kind of concession to the practical power of an “unfunded” pay-as-you-go scheme.
In the case of GEPF the power of pay-as-you-go reality has been boosted with large cash flows from dividends and interests on investments. For fourteen years, investment cash flows were not touched to pay benefits. They were all reinvested by PIC. Contributions were equal to or higher than benefits paid from 1998 to 2013. The contribution surpluses were also invested. The result was exponential growth of GEPF’s financial assets, only interrupted by the stock market crash of 2008-2009.
For this reason, Finance Ministers and GEPF boards have never followed the auditors’ recommendations to increase the employer contribution rate. The 2018 audit had the same recommendation, but the level stayed at 16% on the salary for service staff and 13% for all others, giving an average of 13.5% in employer contribution. 7.5%, as is the norm in many pension schemes, is deducted from the salaries of the employees.
The change of benefit rules in 2012 can illustrate this play between funding theory and real world practice. Only 19% of the contingency reserve was theoretically “filled” in 2012 year’s audit. The auditor reported a R437bn shortfall in the “long-term funding level”. This concept was introduced after 2006, when the GEPF reached 100% funding of total theoretical liabilities for the first time. It is not in the GEP Law.
The 2012 “shortfall” would be more than R600bn in today’s prices. This beats the theoretical “long-term funding R573bn shortfall” in the 2018 audit that rung alarm bells in media. In 2012, there were no alarm reports. Instead, the GEPF board must have noted that the funds were ballooning as usual. They could be tapped.
A generous rule change from 1 April 2012 let members withdraw their whole “actuarial interest” share in the GEPF if they resigned before retirement day, instead of using a defined benefit formula for early resignations unrelated to developments on the capital markets.
The change led to mass resignations. This was encouraged by the finance industry: The lump sums should of course be moved to private schemes to attract hefty fees. An eyewitness tells us that Finance Minister Pravin Gordhan pleaded with delegates at FEDUSA’s 2013 wage bargaining conference to start a campaign for the GEPF.
But benefits paid doubled from R43.2bn in 2012/13 to R85.8bn in 2014/15 (stabilising at that level). Thus a part of cash flow from dividends and interest incomes had now to be used to pay benefits from 2013/14, for the first time since 1998.
The incident demonstrated the healthy state of the fund even when it is hit by shocks, like unintended consequences of policy changes. The contribution rates stayed the same and yet the GEPF’s cash income surplus in the 2018 fiscal year was R47.5bn. Hence there was no need for additional payments from the government to deal with the shock.
But what about the R573bn “long-term funding shortfall” in March 2018?! Well, there is no basis in the real world for the GEPF having R573bn more in financial assets. It is imali yomoya. And we should not even want the GEPF to be “R573bn bigger”. It would spell disaster.
Who would stand today in the other end of such a R573bn financial claim or ownership title of the GEPF? Do we want the debt of Eskom to be R573bn bigger? Would we like the government debt to be R573bn bigger? Would we like the market Johannesburg Stock Exchange (JSE) to crash from a R573bn higher peak value?
The total value of its shares traded on JSE, the “market capitalisation”, was 315% to Gross Domestic Product (GDP) in March this year. This is a global record. It is a reason for caution, but it also adds to the argument why Eskom’s debt crisis mustn’t be rescued from the national budget; thereby leading to greater austerity.
Judging from the 2015-2018 trend, benefits paid to members will seven or eight years from now again be financed by employer and employee contributions alone, but this also depends on the Treasury.
The defined benefit guarantees in the GEP Law drives the growth rate of benefits paid lower than the usual about one percentage point above inflation growth rate of salaries. But in the 2019 Budget, R16bn is budgeted for offering early retirement packages to 30 000 senior government staff, using an undisclosed amount from the GEPF. The Treasury said it will be paid back later.
According to the Presidential Summit on Public Health in October 2018, there are 37 000 real vacancies in public health alone. But the abundant funds of GEPF are used for austerity measures. In this instance no problem is raised with “the use of workers money”.
The big crisis in ESKOM and the huge impact ESKOM’s collapse would have for all of us requires serious consideration of alternatives.
Renegotiating the debt and claims
In time of debt crisis for governments, private creditors have often been offered a so-called “haircut” or stand the risk of losing all their claims.
A “haircut” can mean to get a lower interest on a loan. Any kind of more or less radical options exists. In our case the government would be negotiating with a state organ. Alternatives can range from simply writing off the debts or delaying the pay back. Ordinary bonds can be converted into “zero coupon bonds” as has been suggested by Magda Wierzycka: The loan and the interest on it is paid when the loan period expires.
GEPF can take a haircut of its R87bn bond claim on Eskom (March 2018), with no risk for the guaranteed pension payments: the minimums are defined in the GEP Law of 1996. It can forfeit some R8bn in interest income from its Eskom bonds, converting it to an interest free loan. This interest is today paid by public sector workers and users. The government finances it with growing debts and more budget austerity.
At the start of the PIC investigation, the GEPF board mentioned that they have a policy of “maximising returns”. This policy can be changed into maximising social utility without any change in the GEP Law. The theoretical determined funding level can even drop below 100%. The GEPF would still be able to fulfill its primary role to pay pensions and benefits, despite corruption and patronage at its manager PIC, which under the flag of BEE has become too large.
A real broad-based black economic empowerment program puts an end to austerity and guarantees all rights in the Bill of Rights, especially socio-economic rights.
To change the terms of Eskom’s debt to GEPF is not to destroy financial wealth. It is to move money from a place where “maximised returns on investments” is not needed, not required by the GEP Law and not rational from the point of view of the vast majority of citizens, namely from the GEPF, to the national budget on which the daily life of tens of millions of South Africans depends as public service sector users and workers.
A change of GEPF’s investment policy from shares to bonds in SOEs and to Treasury Bonds – for example switching the 50% in local equity (R1.034tr in March 2018) and 32% in local bonds (R568bn) to the opposite relation – could offer the government and indebted SOEs up to R450bn more in credit from a creditor that itself controls. The terms of intra-governmental loans can be decided outside the markets, bringing down the interest rates without endangering pension guarantees.
This is also more prudent. From 2016 to 2018, the auditor increased the goal for the solvency fund from R302bn to R402bn. There was no explanation for the 33% increase, but it reflects that the risk to GEPF’s funds has increased. A solvency fund guards pension payments against sudden losses from risky investments and stock market crashes. A riskier investment policy demands a larger solvency fund.
Even when setting to zero and disregard the other political contingency fund requirements added since 2006, the 2018 solvency requirement was only filled to a third: At the “100% funding level” there was R137.5bn left of financial assets in the GEPF to meet the R402bn safety requirement.
Consequently and by their own cherished standards, the Finance Minister and the GEPF board can have no defense for their present investment policy: It is too risky for the GEPF’s finances. As we showed above, the solvency requirement for their investment policy can have no backing, no real world counterparty, in South Africa’s economy. To siphon off even more employer contributions from the national budget to the GEPF is out of the question. The only solution is to shift investments from risky equity to bonds. Corporate indebtedness is a problem. Government bonds are the safest alternative.
For the GEPF’s ability to pay benefits currently this would in fact be a win-win change. The 2017-2019 Budget Review reporting on cash flows from investments read together with the GEPF’s annual reports show that cash flow returns on bond possessions since 2010 has been well above 7%. Dividends to equity held by the GEPF have had an average return of around 3.3%. The weighted average was 4.4%. Changing investment policy from equity to bonds would guarantee that cash flows from contributions each year are higher than benefits paid, even with a necessary rebate on intra-government loans. The surpluses should be used to stop austerity and defuse the debt crisis.
Eskom is too big to fail, but the Treasury must stop to support Eskom via the national budget, borrow more money for this purpose at market rates and unnecessarily increase its debt service costs. The political agenda must change. Austerity must be reversed, not sharpened.
The 2019 Annual Financial Statement of Eskom now reports R119.8bn in debt to “public entities”. This is 27% of Eskom’s R445bn debt in March 2019.
Next occasion to change the debt service regime in the interest of the vast majority, including the active and retired members of GEPF, will be the Midterm Budget in October that sets out the three year fiscal framework.
Dick Forslund is senior economist at Alternative Information and Development Centre. This is Part 1 of a chapter in a coming AIDC booklet about the PIC and the GEPF. It can be read together with Dominic Brown’s article on the PIC published by Daily Maverick. Part 2 will address the “10% public sector wage cut” proposal.
In the days when DB funds were commonplace, an employer ‘contribution holiday’ was often used to absorb unnecessarily high surpluses. Is that not an option in this case ? A two year holiday would make 100m available to support Eskom in a budget-neutral manner without having to tinker with contractual obligations.
That should read 100bn !