FLAC instruments raise investment, risk and capital questions for South African insurers at both the technical and systemic risk levels. The standard formula offers choices with materially different results. Insurers should examine the alternatives and, where exposures are material, assess the appropriateness of the standard formula as part of the Own-Risk Solvency Assessment (ORSA).

Insurers, actuaries and risk managers should pay attention to their accumulation of banking sector exposures.

Introduction: How South Africa’s First Loss After Capital instruments work

South Africa’s initial First Loss After Capital (FLAC) instruments were issued in recent weeks. Absa issued approximately R3 billion and Standard Bank R2 billion, with combined bids exceeding R18 billion.

FLAC instruments are designed to absorb losses when a systemically important bank enters resolution. They can be written down or converted to equity under the resolution framework of the South African Reserve Bank (SARB), ensuring that losses fall on FLAC holders rather than taxpayers or depositors. Six South African banks are designated as systemically important financial institutions (SIFIs) and are required to build up FLAC buffers on a phased basis to 2030.

From a banking and public policy perspective, the framework is sound, and Moody’s has described it as credit-positive for senior creditors and depositors, consistent with FLAC's subordinated position in the loss hierarchy.

A material proportion of the institutional demand for these instruments will be coming from, or flowing through to, insurance company balance sheets, whether through direct investment, fixed-income mandates or pooled funds. In this article, I consider two questions:

• How should FLAC instruments be treated under the Solvency Assessment and Management (SAM) standard formula?
• From a systemic risk perspective, is it appropriate for insurers to be absorbing bank losses in a financial sector crisis?

SAM classification of South African FLAC instruments

Spread risk as the appropriate module

The SAM market risk module captures credit risk through two main channels: spread risk and default risk. The distinction, broadly, is as follows:

• Spread risk applies to instruments for which value is exposed to credit spread movements. The likely path to loss is through mark-to-market as spreads widen. The holder can trade out, but at a loss.
• Default risk applies when the holder is less likely to trade out and the loss crystallises through actual non-payment.

This broadly aligns with Solvency II. It should be noted, however, that this interpretation is not universally agreed upon and must be read into the Financial Soundness Standards for Insurers (FSI) framework’s mix of principles, rules, and examples.

FLAC instruments are listed, tradeable, with defined maturities and coupons. For an insurer holding these, the initial path to loss is most likely through spread widening as banking stress develops well before any resolution trigger is pulled. On this basis, spread risk is the appropriate classification for capturing pre-resolution mark-to-market risk, noting that the ultimate risk is principal loss at resolution.

Why it’s vital to get the SCR inputs right to reflect FLAC risk characteristics

The classification is straightforward. The harder question is whether the inputs to the spread risk formula adequately reflect the risk characteristics of FLAC.

The spread risk formula in FSI 4.1 includes a Loss Given Default (LGD) adjustment factor for cases for which the credit quality step (CQS) is based on an issuer rating rather than an instrument-specific rating. This is a departure from Solvency II, which has no equivalent mechanism. This adjustment also applies only to spread risk, not default risk.

The standard is internally inconsistent on this point. Paragraph 9.10 requires that the CQS should reflect the credit quality of the instrument, not the issuing entity. Paragraph 9.12 then provides detailed guidance for when an issuer rating is used instead. If 9.10 were always followed, 9.12 would not need to exist.

For most instruments, this does not matter. For FLAC, it does. At the moment, South African FLAC instruments do not have an instrument-level issuer rating, although there may be a difference between bank rating and holding company (Holdco) rating due to the structural subordination of Holdco obligations.

LGDs and the impact of subordination: CGS adjustment vs. LGD adjustment

FLAC instruments are designed to absorb losses before senior unsecured creditors are affected. In resolution, FLAC is written down or converted to equity. The LGD cannot therefore be the 45% typically assumed for senior unsecured bank debt. For an instrument specifically designed for full write-down, a meaningfully higher LGD is appropriate, possibly in the 80% to 100% range.

The FSI 4.1 default risk section assigns a 100% LGD ratio to “equity, junior debt, mezzanine debt or preference shares exposures, and structurally subordinated exposures.” FLAC is typically issued by the bank Holdco (e.g., Absa Group Limited), not the operating bank (the “Opco,” e.g., Absa Bank Limited). The Holdco’s claims on Opco assets rank behind Opco’s own creditors. This is structural subordination. In addition, FLAC carries a statutory bail-in power under South Africa’s bank resolution framework, meaning the loss mechanism is not merely a function of creditor ranking in a winding-up but a discretionary resolution tool. An LGD ratio of 100% appears most appropriate under FSI 4.1, giving an LGD adjustment factor of 2.22. FLAC is senior to Tier 2 capital, and a 100% LGD may appear conservative on that basis. However, this appears to be the only option consistent with FSI 4.1.

FSI 4.1 prescribes LGD ratios by nature of exposure in the default risk section,¹ and corresponding adjustment factors for spread risk.² The adjustment factors are simply the prescribed LGD ratio divided by 45% (the senior unsecured baseline). Table 1 sets out the full mapping.

If an insurer uses an instrument-level CQS (as paragraph 9.10 requires), the rating agency might notch FLAC down 1–2 CQSs from the issuer rating for subordination and bail-in risk, and LGD_adj is set to 1.

When only an issuer rating is available and the LGD adjustment is applied, the prescribed factor of 2.22 more than doubles the capital charge. These should be two ways of capturing the same risk, but they produce quite different results.

Table 2 shows the ratio of spread risk factors at various CQS notch downgrades relative to the starting CQSs, using values from Part A of Attachment 4. Where the ratio matches 2.22, the two approaches produce the same capital charge. The table uses starting CQS values of 10 to 12 as a relevant subset of the full range.

The LGD adjustment of 2.22× implies a capital charge broadly equivalent to a 3- to 4-CQS downgrade at the longer durations typical of FLAC instruments, and higher at shorter durations. A 1- to 2-notch downgrade, which is what rating agencies often apply, falls well short of this. The two routes through the standard produce materially different capital charges for the same instrument.

There is a further inconsistency. If the same FLAC instrument were classified under default risk rather than spread risk, the prescribed LGD would be 100% with no discretion. The spread risk framework, by relying on instrument-level ratings that may not fully reflect loss severity, can produce an effective LGD substantially below 100% for the same instrument.

Capital and risk implications for South African insurers investing in FLAC bonds

Growth of FLAC in portfolios

FLAC offers a yield pickup over senior unsecured debt from the same issuer. In a low-yield environment, the pressure to capture an additional 20–50 basis points is real, and asset managers benchmarked on yield or tracking error have a direct incentive to accumulate these positions. FLAC will appear in fixed-income and money market portfolios, whether or not the insurer has made a deliberate decision to accept the risk.

Care required around interpretation of CQSs, LGDs, and LGD_adj

If the SCR calculation uses issuer-level ratings with LGD_adj set to 1, which is the default from many data vendors, the additional risk is invisible: same rating bucket, same duration, higher yield, same capital charge.

FSI 4.1 requires look-through for collective investment schemes, but in practice, this is imperfect. Multi-manager mandates, fund-of-funds structures, and balanced funds will inevitably hold FLAC instruments. In many cases, the insurer’s risk team may not be aware of the exposure, as has been the case with credit-linked notes and other structured exposures.

LGD_adj treatment potentially conservative

Given the lack of instrument-level ratings in South Africa, the LGD_adj treatment may be the only available route, making the return-on-capital equation look unattractive. Insurers should form their own view of the appropriate LGD for FLAC for internal risk and capital assessments. Return on economic capital should be considered alongside regulatory capital implications for decision-making.

Diversification and concentration by issuer and risk factor

Diversification across six South African bank names should not be confused with genuine diversification. These institutions are all exposed to the same sovereign, the same macroeconomic environment, and largely the same depositor base. The standard formula’s concentration risk sub-module will not capture this correlated exposure.

Practical recommendations

Those involved with investments, credit risk, capital management or ALM should:

• Establish what rating basis and LGD_adj value are being applied to FLAC in the SCR calculation. Compare the result under both approaches. When the issuer rating is used with an LGD_adj of 1, the capital charge will be understated.
• Ensure FLAC exposures are visible, deliberate, and addressed in investment mandates.
• When FLAC exposures are material, assess the appropriateness of the standard formula as part of the ORSA. The standard formula offers choices with materially different results; insurers should satisfy themselves that the chosen approach is adequate.

Systemic risk: Who should hold bail-in bonds?

The questions above concern the standalone capital treatment. There is a broader question about whether it is appropriate for insurers to be a significant funding source for bank bail-in capacity.

Wrong-way risk

FLAC instruments absorb losses when a bank enters resolution. The SARB would trigger the resolution of a SIFI only during severe financial stress. During such periods, insurers face simultaneous pressures, which might include:

• Equity market declines
• Credit rating deterioration and credit spread widening
• Changing yields, putting pressure on asset-liability matching
positions, including the impact of convexity
• Pressure on liquidity, including repo rollover
• Elevated lapses and surrenders
• Reduced new business volumes, adding unit expense pressure

The insurer therefore takes a significant loss on its FLAC holdings at precisely the moment its own solvency position is most strained. This is a textbook example of wrong-way risk. The liquidity dimension compounds the problem: in a systemic banking stress, FLAC instruments are likely to become illiquid, and repo eligibility and haircuts may tighten materially, particularly given the instrument’s bail-in role.

The Credit Suisse episode in 2023 illustrates the gap between resolution theory and practice. When Credit Suisse faced a crisis of confidence, the Swiss Financial Market Supervisory Authority (FINMA) ordered the complete write-down of CHF 16.5 billion in Additional Tier 1 (AT1) capital instruments, a class of loss-absorbing instruments comparable in purpose to FLAC. The write-down inverted the expected creditor hierarchy: AT1 holders were wiped out entirely while equity holders received value through the emergency merger with UBS. The decision was subsequently ruled unlawful by the Swiss Federal Administrative Court in October 2025 on the grounds that it lacked sufficient legal basis⁴. The episode illustrates that when resolution mechanisms are triggered, the outcome can be messy, politically driven, and legally contested. Holders of loss-absorbing instruments bear not only credit risk but also substantial legal and political uncertainty.

Contagion through cross-holdings

Cross-holdings of loss-absorbing instruments transform idiosyncratic risk into systemic risk. If Bank A holds Bank B’s FLAC and Bank B enters resolution, Bank A’s capital is impaired. European regulators have discouraged cross-holdings of Total Loss-Absorbing Capacity (TLAC) instruments and Minimum Requirement for own funds and Eligible Liabilities (MREL) instruments within the banking sector for this reason. TLAC is the Financial Stability Board’s global standard requiring systemically important banks to maintain sufficient loss-absorbing capacity; MREL is the EU’s implementation of the same principle.

The same logic applies to insurers. If the insurance sector becomes a major holder of bank FLAC, banking stress transmits directly to insurance balance sheets. The two sectors become more tightly coupled at precisely the point when independence between them is most valuable.

Beyond FLAC: Banking sector exposure

If the concern is wrong-way risk between bank instruments and insurer stress, the logic extends beyond FLAC to bank Tier 2, AT1, senior unsecured, and arguably any material banking sector exposure. The difference is one of degree, not kind. South African insurers hold a significant proportion of their credit portfolios in bank paper because banks are among the few investment-grade ZAR issuers with regular supply and reasonable liquidity.

From taxpayers to policyholders

The FLAC framework prevents taxpayer bailouts by shifting losses to creditors. When those creditors include insurers, the loss has moved from the state to the insurance sector—and ultimately to the same public the framework aims to protect.

For most insurers, FLAC will represent a small fraction of invested assets. But FLAC losses crystallise at precisely the point when local equity portfolios may be falling, credit spreads are likely widening, and liquidity is under pressure. The standard formula’s aggregation of market risk sub-modules does not fully capture this wrong-way dimension, and the extent to which it is captured depends on which route through the spread risk formula is applied.

Conclusion: Considerations for South Africa’s insurers when adding FLAC instruments

FLAC instruments raise investment, risk, and capital questions for South African insurers at both the technical and systemic risk levels. The standard formula offers choices with materially different results. One likely default system-generated answer (issuer ratings with no LGD adjustment) understates the capital requirement. The most accessible, compliant option may overstate the risk. Insurers should examine the alternatives and, where exposures are material, assess the appropriateness of the standard formula as part of the ORSA.

On the systemic risk question, insurers, actuaries, and risk managers should pay attention to their accumulation of banking sector exposures. This concern is compounded if capital charges are understated. The insurance sector may accumulate more exposure than the framework accounts for, amplifying the same wrong-way risk it should be capturing.

---

¹ Paragraph 9.22 of Prudential Standard Financial Soundness Standards for Insurers (FSI) 4.1.

² Spread factors are contained in Part D of Attachment 4 of FSI 4.1.

³ FINMA, "FINMA provides information about the basis for writing down AT1 capital instruments", 23 March 2023. The AT1 write-down was part of the emergency measures facilitating UBS’s acquisition of Credit Suisse. https://www.finma.ch/en/news/2023/03/20230323-mm-at1-kapitalinstrumente/.

⁴ Swiss Federal Administrative Court, partial decision in Case B-2334/2023, published 14 October 2025. FINMA has appealed to the Federal Supreme Court; the appeal was pending at the time of writing.