Assessing South African credit markets
By Ryan van Breda, Portfolio Manager – Credit
The credit curve we created
In our 2026 Mid-Year Investment Outlook, we argued that the margin for safety in South African credit continues to narrow. Credit spreads have tightened despite a more complicated macro environment, while the introduction of FLAC¹ (First Loss After Capital) and the transition from JIBAR² (Johannesburg Interbank Average Rate) to ZARONIA³ (South African Rand Overnight Index Average) had failed to produce the increase in compensation that many investors initially expected. Liquidity continues to remain the dominant force. That remains our view, however, focusing only on the level of spreads risks overlooking a more interesting question who actually controls the credit curve. Bank funding provides a useful place to explore this because much of the bank credit market is established through primary market auctions. Banks determine:
- When to issue
- How much to raise
- The instrument
- The maturity
Arrangers provide initial price guidance but ultimately investors determine the clearing spread through the prices at which they are prepared to bid. In other words, the bank credit curve is not something imposed on investors. Collectively, investors create it.
This is an important distinction. It is easy to observe that spreads are tight, that the term premium is low or that FLAC should arguably command more compensation for structural subordination. Yet, if an auction is comfortably covered and investors continue to bid at progressively tighter levels, the auction reveals something different from what investors may say about valuation. The uncomfortable conclusion is that, to a large extent, we are the curve.
Auctions reveal what investors are willing to pay
Primary auctions are effectively exercises in revealed preference. An investor may believe that a bond should clear wider, but the price that ultimately matters is the price submitted into the book.
This creates an interesting dynamic. Most investors would probably prefer the entire credit curve to be wider, yet no individual investor controls where it clears. A portfolio manager who consistently bids materially wider than the market simply risks receiving little or no allocation for the bond on issue, particularly in an environment where liquidity and cash remain abundant with demand for high-quality floating-rate assets exceeding available supply.
The price that ultimately matters is the price submitted into the book.
The rational decision for an individual portfolio can therefore contribute to an outcome that the investor base collectively dislikes. This is particularly visible in South African bank funding. Bank balance sheets remain fundamentally strong, floating-rate instruments provide attractive contractual income with limited base-rate duration, and secondary-market volatility remains relatively low. Against this backdrop, the cost of missing an auction can appear greater than the cost of accepting less spread [Figure1].
Figure 1: The self-reinforcing cycle in bank credit
A transaction clearing tighter becomes the new comparable and the secondary curves adjust towards that level. The next auction references the previous clearing spread, and investors assess relative value against an increasingly compressed starting point. This does not require a material improvement in credit fundamentals – only sufficient competition for the available bonds.
This effect does not stop with banks. Bank funding curves are an important anchor for corporate credit, meaning tighter bank auctions have dragged broader corporate funding expectations lower too. One aggressive clearing level can therefore reset relative value well beyond the bond being issued.
The uncomfortable conclusion is that, to a large extent, we are the curve.
The Credit Adjustment Spread was not lost, it was bid away
The transition from JIBAR to ZARONIA provides perhaps the clearest example of the market accepting less compensation than the mechanics of the transition initially implied. As ZARONIA is an overnight reference rate and sits below 3-month JIBAR, a Credit Adjustment Spread (CAS) was introduced to bridge that difference and preserve economic equivalence through the transition. Conceptually, if a JIBAR-linked senior bank bond offered JIBAR plus 100 basis points and the applicable CAS was approximately 16 basis points, an equivalent ZARONIA-linked spread would need to be approximately ZARONIA plus 116 basis points. That is not what occurred.
As banks began issuing directly over ZARONIA during 2025, the difference became visible in actual clearing levels. Three-year OpCo senior unsecured bonds initially cleared around ZARONIA plus 99 to 102 basis points, with subsequent transactions tightening further. This was materially below the spread that would have resulted from simply taking comparable JIBAR-linked issuance and adding the CAS to preserve economic equivalence.
In other words, the change in benchmark did not translate into an equivalent increase in the quoted spread over ZARONIA (despite ZARONIA being a structurally lower base rate vs JIBAR). Strong investor demand absorbed the difference, and much of the compensation represented by the CAS was effectively bid away through the auction process.
It would be easy to describe this as a mechanical consequence of benchmark reform. It was not. The benchmark changed mechanically, the amount of compensation investors required was determined through the auction. The CAS was, in effect, auctioned away [Figure 2].
Figure 2: 3-year bank issusance spreads through the benchmark
FLAC ‘notes’ changed the risk, investors determined the premium
FLAC ‘notes’ materially changed the creditor position of South African bank investors. Senior investors historically had a direct claim against the regulated operating bank. FLAC sits at the holding-company level, is designed to absorb losses in resolution and depends more heavily on cash being upstreamed from operating subsidiaries. That is a meaningful change in risk.
Yet the market has charged surprisingly little for it. Three-year FLAC has recently cleared around ZARONIA +90 to +95 basis points, broadly similar to where three-year OpCo senior ZARONIA paper initially cleared during 2025 at around +95 to +102 basis points. The comparison is not perfectly like-for-like because the market tightened over the period, but that is precisely the point; the market has allowed the broader tightening to absorb much of the premium one might otherwise have expected for moving from OpCo senior to HoldCo FLAC and from JIBAR to ZARONIA.
The market has allowed the broader tightening to absorb much of the expected FLAC premium.
The term premium is also our decision
The same logic becomes particularly interesting when considering maturity. We already know that the South African bank credit curve is relatively flat. Once investors have accepted FLAC risk, the more useful question is what investors are actually demanding for each additional year of exposure.
Using same-bank, same-auction FLAC transactions, the current curve tells a clear story. The median three-to-five-year premium is now approximately 17 basis points. Extending from five to seven years adds only another 8 basis points. Extending from seven to ten years has recently added around 9 basis points, although the ten-year sample remains smaller. This means that the first two years of extension from three to five years still earn a recognisable premium. Beyond five years, however, the additional compensation declines sharply.
Figure 3: Evolution of term premiums from OpCo/JIBAR to FLAC/ZARONIA
Source: Bank auction data at 07 September 2026, Perpetua Research
The important point is not simply that the curve is flat. It is that every point on that curve reflects an investor decision. If five-year FLAC clears at 107 basis points and seven-year FLAC clears at 116 basis points, investors have collectively decided that an additional two years of credit, liquidity, regulatory and structural exposure requires only another 9 basis points of compensation. This changes the portfolio-management question. It is no longer: does ZARONIA plus 116 basis points look attractive in absolute terms? It is; is another 9 basis points sufficient to justify extending the risk for two additional years? Those are very different questions.
The reason rational investors continue to tighten the curve
None of this implies that investors are behaving irrationally. The incentives are understandable. Cash needs to be invested. The major banks remain high-quality credits. Floating-rate coupons provide attractive all-in income. Supply of comparable high-quality assets remains limited. Primary allocations can be difficult to obtain. In a buy-and-hold market characterised by limited visible spread volatility, the immediate cost of sacrificing several basis points can appear small. This creates a collective-action problem.
An individual investor can demand materially wider spreads and preserve valuation discipline, but if others are willing to bid tighter, the likely result is simply a smaller allocation. That investor then faces the same decision at the next auction against a new, tighter market reference.
In this environment, the most aggressive capital can become the marginal price setter for the broader market.
The most aggressive capital can become the marginal price setter for the broader market.
This helps explain why spreads can remain tighter for longer than traditional macro relationships might suggest. It also explains why higher issuance volumes have not necessarily produced wider funding spreads. The decisive variable is not simply how much banks issue – it is how much capital is competing to own those bonds. Liquidity does not merely influence the curve. Liquidity expresses itself through investor bids and creates the curve.
The discipline is knowing when not to set the price
There is an important irony in a market where almost everyone agrees that spreads are tight and auctions continue to clear tighter because investors continue to submit the bids that allow them to do so. That does not mean investors should collectively stop buying bank credit. Nor is there a single “correct” spread that every portfolio should demand.
Different portfolios have different liabilities, liquidity requirements, benchmarks and return objectives. But active portfolio management requires recognising that participation itself influences price.
At some point, competition for allocation becomes a competition to accept less.
At some point, competition for allocation stops being a process of demanding more compensation for risk and becomes a competition to accept less. That is where valuation discipline matters most.
Recent bank auctions have provided an important reminder that this process can also reverse. Standard Bank’s latest Tier 2 auction sought R1.5 billion but attracted only R1.369 billion of bids, forcing the bank to take the entire book at the top end of pricing guidance. For once, the market did not provide enough demand to validate the existing curve. That is significant as it shows that when investors collectively step back, the clearing price has to adjust upwards.
For us, the opportunity in bank credit remains to collect resilient floating-rate income while retaining flexibility. We remain comfortable owning high-quality bank risk, however, are more cautious where relatively small increments of spread require materially longer exposure or a weaker creditor position. The bank credit curve may be exceptionally tight, but it is not outside our control. Every auction is a vote on what risk is worth. Sometimes the most important vote is the decision not to bid. Collectively, investors determine the result.
³ZARONIA is a financial benchmark used in South Africa to reflect the interest rate at which commercial banks obtain overnight, rand denominated wholesale funds.
²JIBAR (Johannesburg Interbank Average Rate) is the historical term benchmark rate used in South Africa, calculated from quotes representing the average interest rates at which major banks lend unsecured funds to one another for specified durations (such as 3 months)
¹ FLAC refers to a type of financial instrument designed to absorb losses and recapitalise a bank if it experiences severe financial distress