It is my pleasure to welcome you to this Financial Markets Department (FMD) cocktail function. We last hosted this function in May 2019 – just a day before the national elections. As COVID would have it, it has taken five years to meet again – coincidentally, just before another national election. Hopefully we will not have to wait until 2029 for our next engagement. 

This event serves a couple of important purposes. One is to acknowledge the important relationship we have with you, as practitioners. When I became responsible for financial markets, one of the things that most impressed me was the level of trust, expertise and commitment people were bringing to our engagements. It is really a model for a flourishing industry. Many of those people are here tonight. Thank you.

The other goal of this function is to update you on what the Financial Markets Department has been up to.  The role of a financial markets department in a central bank is often not well understood, in part because we perform such a complex variety of functions. Some are business as usual, such as managing foreign reserves or facilitating government bond auctions. Others are more strategic, and so we undertake a variety of projects and initiatives to innovate and improve, within the limits of what a central bank can do in the greater financial markets landscape. This cocktail is our opportunity to share some of our perspectives and work. 

A few of our projects are well known to you all. I won’t dwell on the benchmark reference rate reform, for example, which is well advanced. The main takeaway from the Market Practitioners Group conference held last year was that the market is gearing up for a new benchmark interest rate. From our perspective, we are making good progress, and we look forward to the market saying goodbye to Jibar [Johannesburg Interbank Average Rate] and switching to ZARONIA [South African Rand Overnight Index Average]. This will go a long way in further enhancing trust in the financial sector. 

Similarly, we are making progress with our ongoing triparty collateral management initiative. We look forward to a proof of concept this year, and hopefully this system will be up and running in 2026.

Another project we have worked hard on over the past year, which has had much less of a public profile, is the gradual refinement of our market dysfunction framework. Central banks have served as lenders of last resort to banks for centuries. What is newer is an emphasis on markets, in addition to banks, as the institutions that can fail, and in which central banks can achieve better equilibria by timely and well-judged interventions.  Many of our peer central banks refer to this as market maker of last resort.  The emphasis is on last resort.  

One of the reasons we have been developing this framework is because the market is deeply complex. It requires a great deal of thinking to work out what tools you need for what markets.

For a start, what we have done with our new framework is identify core markets, defined as those which are big, connected and for which there are no ready substitutes. We see three such markets: the government bond market, the foreign exchange market, and the money market. There are obviously many other markets for financial products out there, but our assessment is that these are the three markets which the South African Reserve Bank (SARB) must keep going.

Of course, we do not want to get trigger-happy, blundering into these markets at the first sign of trouble, generating moral hazard problems and wasting precious resources. 

Accordingly, we have invested considerable effort working out intervention strategies and indicators that will help us to be effective and minimise risks. 

Just because a market is showing a price you do not like, it does not mean there is dysfunction. 

Market dysfunction is more properly understood as the inability of markets to form prices. If trading is breaking down, if bid-offer spreads blow out, if small trades move markets by large amounts, very likely there is dysfunction – especially in markets that are normally deep and liquid, such as the core markets I cited. 

In these contexts, there is no contradiction between the central bank being committed to a market-determined price and the central bank intervening to restore market functioning. 

A good example of this is our new FIMA[1] on-lending facility, which allows us to provide dollar liquidity directly from the New York Federal Reserve (Fed) to local banks. Simply put, the US Fed is willing to lend us dollars, as South Africa’s central bank, through their FIMA window. We can then make these dollars available to local banks, through our on-lending facility.  This is not an everyday source of dollar financing, because the rate is set above normal market rates. But it allows us to provide emergency liquidity in dollars, the most widely accepted foreign currency, and it will be immune to market dysfunction because our counterparty is the Fed. The ultimate aim is to make sure that the foreign exchange market remains liquid and functional, even during stress episodes.

This FIMA system is a good example of a tool that works through secured lending. We like these tools because they are less risky, and we can set pricing and then let the market uptake tell us if the intervention demand is there. But we cannot always intervene by lending. Our most important market dysfunction operation in recent history was the 2020 bond market intervention, where we purchased bonds outright. We look back at that as a success. That said, we have also learnt some lessons from the episode, mostly about the indicators we rely on and the circumstances which warrant intervention. 

As you know, we have also consolidated our new monetary policy implementation framework (MPIF), which has replaced the old shortage system with a surplus system. To our knowledge, this is the first such system in an emerging market – although these kind of floor systems are common in advanced economies. 

The new framework gives us more flexibility than we used to have when we relied on rationing liquidity to transmit interest rate decisions. Hopefully, with more liquidity in the system, the chances of shortages and money market dysfunction are low. But if we have to intervene, we feel we can now provide whatever liquidity the system needs to stabilise. So this is also a powerful tool.

This brings me to the hot topic of the day ‒ the GFECRA reform.

There has been a great deal of discussion in the public domain in the past few weeks on the Gold and Foreign Exchange Contingency Reserve Account (GFECRA).  I do not want to venture into a long discussion here on its features, as well as the details of the current transfer to the government as seen in the 2024 Budget Review.  Instead, I would like to make a few comments on what the operational aspects of a distribution means for financial markets, as that is my main focus this evening. 

From time to time, a central bank must implement policy that puts strain on its balance sheet. This happens, for example, if it has to accumulate foreign exchange reserves, buy sovereign bonds in response to market dysfunction or when it has to make transfers to government. The strains come from the need to sterilise the effects of purchases on the asset side of our balance sheet to neutralise the effect of monetary expansion. To remain solvent, central banks then need adequate capital buffers. This is an important underlying principle of our agreement with government. 

What the recent discussion of GFECRA has done is put renewed focus on how finance experts, accountants and auditors need to work rigorously together to determine our capital or equity needs that arise from a particular policy decision.  Failure to do so would have implications for financial markets and the ability of the SARB to pursue its mandates in future. 

That is why the settlement of unrealised GFECRA balances will follow a waterfall approach, which requires that we first build sufficient buffers to absorb large exchange rate shocks and sterilisation costs associated with the remuneration of excess reserves, before distributing any funds to the government. 

By moving a portion of GFECRA into our contingency reserve, we end up looking more like other central banks, where valuation effects are reflected in central bank equity, and foreign exchange reserves are funded more by central bank debt and equity – instead of large amounts of government equity, which has become the case in South Africa. The implication for government can reduce its borrowing.

The new framework also implies that we will be expanding the money market surplus. We have committed to being transparent about this and sharing the transition plan so that banks can plan for the change. The expansions of liquidity during the 2022 MPIF transition and the 2023 NTSDA[2] drawdown both went smoothly. I think part of that was our good planning, but a lot of it was skillful liquidity management by banks. We want to see this completed without market disruptions and without material changes to financial conditions. I look forward to close cooperation with you to achieve these goals. 

In future, I expect GFECRA distributions to be smaller, because they will reflect balances accumulated over one year rather than two decades. However, consistent with our goal of avoiding any distortions of financial conditions, National Treasury and the SARB have committed to stagger any disbursements of funds if the quantities are large enough to create disturbances. Again, we are committed to maintaining effective monetary policy transmission.

Let me conclude by thanking you all again for joining us this evening, and for our productive relationship. I will spare you the standard joke about enjoying the liquidity on offer.