Issue: September/October 2007


Short walk to statism

Unless the NSF has trustees independent of government, and immutable rules obliging them to advance member interests irrespective of the government of the day, there is a real risk of national savings being vulnerable to the service of government.

Imagine that Thabo Mbeki and Trevor Manuel are replaced by a president and finance minister of populist persuasion. Imagine that they propound a radical interpretation of the Freedom Charter. Imagine that they consider markets to be “amorphous”, as Manuel in an earlier period described them, capable of being bent into shape for the state’s dictates.

Now imagine the proposed national savings fund (NSF) in such an environment.

It doesn’t require too much imagination, unfortunately. Jacob Zuma has pals and a platform. And the ANC has a long history, passionately advocated in the struggle but practically reversed in power, of wanting to control the economy’s “commanding heights”. Imagine its re-reversal under a new government, taking South Africa back to the uncertainties over future economic strategy that prevailed in the early 1990s.

That’s where the NSF comes in. Under control of the state, the NSF and the state-owned Public Investment Corporation (PIC) can become immensely muscular arms for extending state influence over the private sector. As the size of assets under their control increases, as it inevitably will, their equity investments in JSE-listed companies will increase commensurately. So can their boardroom interventions. The PIC has come to light with shareholder activism, and the NSF would be similarly entitled.

Whether this is desirable or dangerous depends crudely on subjective philosophies of how much is too much influence by the state over corporate South Africa. At the extreme, “too much” would presumably be when a capitalist entity is effectively nationalised by the capitalist mechanism of a dominant state-controlled shareholder asserting its ownership rights. There are many pit-stops between.

Everybody is dissecting the NSF proposals for their effects on retirement funding, on savings and on institutional capacity. Nobody is highlighting their possible political implications because they appear, for now, to be benign; that there will forever be a government seeking genuinely to find the best structures for retirement funding and social security, without motives for manipulation. But political implications are inescapable, for better or worse, depending on whose hand is at the state’s tiller. Once the structures are in place, they can be used or abused. National Treasury’s proposals should be seen as a warning shot across the bows of private-sector institutions in the business of retirement funding. Come up with something better – or suffer the consequences.

It was only from 1994 that the PIC, whose main client is the Government Employees Pension Fund (GEPF), was allowed to invest in JSE-listed companies. Since then, assets under its management have grown exponentially. From R221 billion in 2000, they reached R600 billion in 2006 when 46 percent of the portfolio was invested in equities, to R720 billion this year when 48,6 percent was invested in equities. It represented over nine percent of the JSE’s entire market capitalisation.

So far, so good. But what might happen if a fresh set of government nominees is appointed by a different government to the boards of the GEPF and the PIC? What if they were to insist on activism criteria coloured more by the approach of a Zuma than an Mbeki?

Constrained by limitations on offshore investment and prudential requirements to hold mainly liquid stocks, the PIC’s equity portfolio already comprises stakes of eight to 22 percent in 40 of South Africa’s major financial and industrial corporates. The percentages can only grow as fund members’ contributions grow. What applies to the PIC is potentially magnified in the case of the NSF.

On migration of members to the NSF, it could conceivably kick off with an asset base not too dissimilar in size to that of the PIC. Certainly, the migration in terms of member numbers will be huge. On the proposals as they stand, which do not allow for a contracting out to alternative schemes, employees’ membership of the NSF will be mandatory and their contributions will be at 15 percent of the first R60 000 in every person’s annual wage or salary.

Various estimates – by Elias Masilela to a Sanlam symposium, by Colin Southey and Howard Buck to an Actuarial Society conference – indicate that at this threshold some 60 percent of existing fund members would be switched into the NSF. However, since it would not be feasible or economic for earners of up to say R120 000 additionally to be members of existing funds, and so duplicate costs, they too will probably migrate to the NSF. This might take the NSF to 80 percent of existing funds’ members, significantly at the expense of industrial and union funds comprising lower-paid workers.

Almost certainly, this will herald:

  • Wholesale industry consolidation;
  • Massive shrinkage in assets under the control of private-sector institutions and proportionate accumulation into the NSF, tantamount to a large part of the financial sector being nationalised;
  • Intense competition among asset managers for slices of NSF business. Where portions of NSF assets are delegated to them for investment, at the NSF’s discretion, they’d have to satisfy criteria that the NSF will determine (for example, on proxy voting?) and report to it;
  • The demise of funds whose members officially earn under R60 000 and effectively under R120 000 a year. This will particularly affect trade unions, whose members and funds will be subsumed;
  • Dilution of member participation. Whereas the Pension Funds Act allows for half the members of funds’ boards to be elected trustees, the NSF cannot be aligned to similar democratic intent. In the nature of a centralised megafund, communication with fund members and accountability of trustees becomes remote and formalistic. For instance, with the closure of many schemes, shop stewards will be removed from their participative role in managing members’ benefits.

Imponderables are whether:

  • Earnings thresholds and percentage contributions can be increased only under particular circumstances, or by government at will;
  • Higher earnings thresholds will cause larger parts of the industry to be nationalised;
  • “Socially responsible investment”, and “minimum of investment risk”, might be defined to mean reintroduction of prescribed assets by another name.

Many of these concerns can be circumvented by contributors being permitted to opt rather for mandatory membership of an accredited retirement fund in preference to the defined-contribution NSF. National Treasury’s discussion paper is non-committal in raising the “opt-out” possibility:

“Several governments have sought to supplement or substitute such public provision with mandatory contributions to individual accounts in privately managed pension schemes. Some countries have adopted a system of auto-enrolment, whereby individuals are automatically enrolled in a retirement plan, but can still practice individual choice to ‘opt out’ if they have more pressing financial needs or wish to transfer to an alternate plan. A variation on this is to allow people to opt out of a public scheme, provided that the individual or employer participates
in a private pension plan that meets approved standards.

”Where free-market adherents in the financial sector and socialistically inclined leaders in the trades-union movement find common cause, the argument for opting out must resonate. Whether it will resonate sufficiently before the next presidential election, and before introduction of the NSF, is a race against time.


For context, recall the mindsets and power plays that pervaded South Africa in the early 1990s. To the then business establishment, the ANC looked threatening for the nationalisation tone of its Freedom Charter. There was deep concern over potent forces in the ANC alliance, outstanding brains in the Communist Party and organisational muscle in Cosatu, to force an ideological agenda.

Almost overnight, once in government, prominent left-wingers became paragons of privatisation; former trades-union leaders and United Democratic Front activists switched into icons of what they might previously have denigrated as capitalist collaborators. There were sufficient of them to have made a profound impact in their new roles. Whether in cabinet, on the boards of financial and industrial companies, and in their own businesses, names once viewed with trepidation as hostile to business metamorphosed to champions of business.

It could so easily have been otherwise. For the state to have taken over the economy’s “commanding heights” – a Marxist euphemism for large corporates – would then have been structurally simple without the bluntness of nationalisation.

During the apartheid era, sanctions prevented foreign money from coming into South Africa and exchange controls prevented domestic money from going out. Contractual savings were concentrated into limited pools of local stocks. The two biggest repositories of local savings were Old Mutual and Sanlam, both mutual societies respectively falling under their own acts of parliament. Few listed companies of substance did not have Old Mutual and Sanlam as significant shareholders.

On the argument that mutual societies were owned by “the people”, the new government need merely have amended their acts of parliament to provide that their boards comprise a majority of government nominees. It would then have been able to sit atop an agglomeration of holdings – in mining houses, financial conglomerates and industrial behemoths – that fell respectively into the Old Mutual and Sanlam camps.

Coincidentally or intentionally, demutualisation eliminated the prospect of government infiltrating the “commanding heights” by such stealth. Ownership of Old Mutual and Sanlam was converted from policyholders to shareholders, and the former listed in London. Their separate acts of parliament fell away. Simultaneously, abolition of sanctions and relaxation of exchange controls allowed them to reconfigure as pure financial service providers, no longer corporate controllers.

Today, in the manner of the former mutuals, the largest single investor on the JSE is the Government Employees Pension Fund (with half its board appointed by government) through the Public Investment Corporation (whose sole shareholder is government). The NSF (falling under government) will inevitably grow larger proportions in market share of assets owned.

There’s no need to worry, so long as those at the helm of government are “people like us”.