Dear [National Treasury deputy director general Ismail Momoniat]
Intellidex is a Johannesburg-headquartered financial services and capital markets research and consulting company. Our client base spans the whole range of offshore and onshore asset managers, banks and other financial institutions – including those affected by Regulation 28. The views here are our own, understanding this client base but also understanding the necessity and complexity of kickstarting an infrastructure investment drive in South Africa.
We welcome reform of Regulation 28 and broadly believe that elements of the proposed reforms are positive. We should also remember that this is an important step is countering the debate over prescribed assets. Regulation 28 offers an effective tool to shape investment discussions without compulsion and the shutting down of that debate will help sentiment.
However, we also believe that some of the proposed changes need clarification and some are likely to be negative from a public interest perspective.
Conceptual starting point
Regulation 28 is a prudential regulation. Its purpose is to ensure that prudence is applied in the management of pension funds. This is a conduct regulation that protects the members of pension funds from undue risk. It backs up the fiduciary responsibility that the managers and trustees of pension funds have to the members. This is clear in section 2 of the existing Regulation 28, particularly paragraph 3(c) that requires funds to “ensure that the fund’s assets are appropriate for its liabilities”.
This is important because it maintains public confidence in pension funds. While pension savings have clear tax incentives, were members to lose confidence that pension fund managers act in their best interests, there would be significant displacement from pension funds into other discretionary savings vehicles.
As it is, we detect anecdotal evidence of savers diverting pension contributions into post-tax savings vehicles in the belief that discretionary savings provide such significant return advantages over pension funds that they make up for the tax costs. We should not do anything that contributes to this (false) perception.
The lessons therefore are:
- In any assessment of Regulation 28, member interests must in fact be paramount
- We must be conscious of whether Regulation 28 will be perceived as serving member interests first and foremost
With this starting point in mind, we comment on the two clear changes the draft regulation proposes.
Separation of private equity and increasing of the ceiling to 15%
In our view this is an appropriate and welcome change. Over the last two decades, private markets have become an increasing source of capital, both debt and equity, while public markets have shrunk. This is a global phenomenon. As we show in figure 1, below, the number of listed companies on many major public capital markets has been shrinking, including on JSE. This creates concentration risk problems as it is more difficult for investors to diversify exposure.
Figure 1: Number of listed companies per market
Data Source: World Federation of Exchanges Statistical Portal*The LSE group was formed after the merger of London Stock Exchange with the Milan Stock Exchange (Borsa Italiana)
In contrast, private equity has been growing and is expected to grow at an estimated CAGR of 11% from 2020 to 2025. It is therefore important that investors have greater access to this increasingly large component of the investment universe to achieve the objectives of maximising returns while minimising risk.
It is also important to separate private equity from hedge funds and other investments in the current investment bucket. Private equity is a clear asset class on its own with unique risk and return features.
In developing an investment policy statement and in tactical investing decisions, funds should be able to decide independently about private equity exposures without having to consider their hedge fund and other asset exposures as part of the same decision.
There is also an appropriate catalyst into infrastructure investing in the change, as many pension funds and insurance companies obtain their exposure to infrastructure through private equity funds.
Our one caution regarding the change is the increase in the ceiling from 10% to 15% in one step. While the vast majority of funds are far from the existing ceilings as it is, there is some minor risk that funds will rapidly expand exposure before the supply side has been able to scale up to provide appropriate investment opportunities, leading funds to invest in low quality assets.
However, given that a gradual approach would be administratively burdensome and this risk is relatively small, we think the single step increase to 15% is acceptable.
Creation of the infrastructure overlay
Conceptually the addition of the infrastructure overlay is an interesting approach to the debate over whether or not infrastructure is an asset class. By treating it as an overlay rather than a specific line item in the Regulation 28 table, the amendments sidestep that debate to some extent. However, in doing so, there is some loss in the coherence of the prudential framework.
Our concern is that it may appear, and may in fact be, an approach designed to drive infrastructure investment not in the best interest of the pension fund member, but in the development interests of the country.
While we believe that a well-functioning pension fund market is critical to development, “well-functioning” should mean that the market protects the fiduciary interests of members, with development as a consequence, rather than the object of regulation. We are concerned that the approach confuses objectives and consequences in this way and risks undermining public confidence in the prudential function of Regulation 28.
This concern is exacerbated by the definition of infrastructure in the draft amendment. By referring to the definitions in the Infrastructure Development Act, the regulation potentially abdicates its prudential responsibility. In investment terms, an asset class should be defined by its risk return characteristics. Doing so allows one to determine correlations with other asset classes in developing an optimised portfolio to meet investor objectives. Therefore, it is critical that the draft has a definition that serves a prudential function. The definitions of the Infrastructure Development Act were determined for completely different purposes.
Moreover, when looking at the Act, the definitions are highly ambiguous. The Act states that “infrastructure means installations, structures, facilities, systems, services or processes relating to the matters specified in Schedule 1 and which are part of the national infrastructure plan.”
We believe that the definition as it exists is not clear.
One interpretation is that the “part of the national infrastructure plan” implies only the Strategic Integrated Projects (SIPs) can count as “infrastructure”. This would be wholly inappropriate definition for a prudential guideline as the SIPs are not a sufficient or necessary definition of the risk/return features of an asset
Globally, infrastructure is recognised as having specific risk/return features –a long time horizon and a risk profile that starts high during construction phase before settling into a low-risk, long run phase of operation. It is seen as uncorrelated to equity and debt markets and non-cyclical.
Infrastructure is also increasingly recognised as an important environmental, social, and governance asset class given that much infrastructure supports renewable energy transition and social impact objectives such as the provision of basic services. . It is these features that should define the asset class, rather than any legislated strategic infrastructure objectives. It is a naturally attractive asset class because of these features.
Schedule 1 of the Act lists all the usual kind of infrastructure you might expect (dams, airports, etc) that Strategic Integrated Projects’ (SIPs) can be formed in relation to, and it might be the list that is the intended definition of infrastructure. But the start of the schedule frames the list as that which “may be designated as strategic infrastructure projects”. Again, this wholly fails as a definition to suit a prudential framework where the objective should be to capture the risk/return features of a class of assets. The risk/return features are consistent whether the infrastructure is private (like broadband networks) or public infrastructure (like power stations) or whether it is domestic or foreign.
Therefore, a definition should allow for private, public, domestic and foreign assets. Indeed, the assets held by South Africa’s existing stock of specialist infrastructure funds do reflect this diversity.
The World Bank’s definition of infrastructure is: “Hard infrastructure often refers to the transport system (such as roads, airports, port facilities, and rail), public utilities (such as energy, water supply and sewer, and irrigation), communication network (such as telecommunication and broadband), and social infrastructure (such as schools and hospitals).” Again, the term “often” here means it is not definitive.
Different subcategories of infrastructure can be important to a definition.
- Economic infrastructure like transportation and utilities that usually generate user fees, vs social infrastructure that provides public services like education and healthcare that can often be free to the users.
- Type of financing mechanism used, like project finance or publicprivate partnerships and setting these financing structure as the definition of infrastructure.
- The risk stage of the infrastructure, separating the early greenfield high-risk stage from the long run brownfield low risk stage.
We believe a definition should give a fund clear guidance though these different considerations. We would urge that Regulation 28 include its own purpose-built definition of infrastructure.Our suggestion is:
“Infrastructure” refers to transportation and, communications networks, utilities (such as energy and water), housing and office accommodation and social infrastructure that provides the land and buildings for the provision of social services like health and education. It includes both greenfield and brownfield assets financed through project finance, public-private partnerships, or other means by the public or private sector.
This would be a definition better aligned with the objectives of a prudential regulation. Positive, open ended enough to be flexible but still clear.
We are also not convinced by the overlay approach.
The vast majority of funds will not be able to categorise the asset class lines into infrastructure and non-infrastructure. While the approach will have the positive effect of catalysing better disclosure at the asset level, it will always be the case that infrastructure forms a relatively small component of the underlying asset classes. We also believe that the existing lines of Table 1 must already be seen as overlapping.
For example, an asset can be simultaneously a foreign exposure and a domestic debt instrument (such a foreign index-linked note issued by a South African bank) and should be counted under both categories (though we find significant variation in practice and confusion in the market on this). We therefore think it should be straightforward to place infrastructure into the rows of the table rather than the columns. Infrastructure assets can be simultaneously debt or equity without any breakdown in the coherence of the table.
Where it is placed into rows, we would advise a lower ceiling than 45%. We think some research would need to be done on historic performance of infrastructure and the value at risk that would be acceptable for pension fund regulations. We would expect this would result in a figure more like 20%. This may become like an infrastructure taxonomy, but one should be careful not to add more red tape on such categorisation.
We welcome further engagement and more discussion on this matter, as we believe that this regulation has the potential to usher South Africa into a new era of higher, sustainable, potential growth.
Peter Attard Montalto, Head of Capital Markets Research, Dr Stuart Theobald, Chairman, Intellidex