Swap hedge accounting under strain from prepayments and a collateral call

What is being hedged. The swap converts floating BBSW funding into a fixed cost, offsetting the fixed yield on the loan book. The hedged item is the forecast interest payments on the rolling borrowings; the hedging instrument is the swap. Economically the relationship is sound - falling rates improve the lender's funding cost while the swap loses value, which is exactly the offsetting pattern a cash flow hedge is meant to capture.

Sources of ineffectiveness

1. Prepayment and amortisation. The loans amortise and about 15% prepay annually, so the borrowings they fund will shrink well before year five. A bullet A$800m notional designated against a declining funding requirement is an over-hedge from around year two: the designated notional will exceed expected future principal, the hedge relationship will fail prospectively, and the whole swap could be reclassified to fair value through profit or loss.
2. Hedge ratio. A$800m against a A$1.2bn book is a 67% hedge. That can be designated, but the ratio must be documented and rebalanced as the book changes.
3. Credit valuation adjustment. A A$95m out-of-the-money position means the dealer's credit standing affects the swap's fair value. AASB 9 permits the CVA to be excluded from the designated hedging instrument and treated as a cost of hedging, which prevents dealer spread moves from generating accounting ineffectiveness.
4. Basis. The borrowings price at BBSW plus a credit spread; only BBSW is hedged, so the spread is a deliberately unhedged component and should be documented as such.

What preserves the relationship: designate a proportion of the swap (an amortising or layered notional aligned to expected borrowings), separate the CVA as a cost of hedging, document the economic relationship and rebalance annually. AASB 9 has no 80-125% bright line - the tests are economic relationship, credit not dominating, and a hedge ratio consistent with risk management.

The arithmetic. The swap is A$95m out of the money to Pacific Agri. With a A$20m threshold and a A$1m minimum transfer amount (already exceeded), the collateral to be posted is A$95m less A$20m = A$75m in cash, varying daily with rates. A A$6m valuation dispute is unresolved, so in practice part of that obligation is not being met.

Liquidity consequences

1. HQLA conversion. Posting A$75m of cash converts liquid assets into an encumbered claim on a dealer. For LCR purposes the cash leaves the HQLA pool, while the resulting receivable is not a liquid asset - a direct hit to the liquidity buffer.
2. Stress outflow. Under the liquidity standard, collateral obligations arising from a ratings downgrade or a deterioration in the bank's own credit standing must be modelled as an outflow in the stress scenario. The ratings-based trigger in this CSA makes that outflow concrete: a downgrade could take the threshold towards zero and require a further A$20m or more at the worst possible moment.
3. Funding cost. The A$75m must be funded. At, say, 5%, that is roughly A$3.75m a year of carry against a hedge that is protecting a margin benefit of about 180bp on A$800m. The hedge is still economically right, but the funding and CVA cost must be priced into the loan margin - this is the practical lesson of the case.
4. Operational and default risk. The unposted A$6m dispute is the most urgent issue. Under the CSA, failure to resolve a dispute through the agreed procedure can escalate to an event of default, and an event of default under the ISDA can cross-default other obligations.

Recommendation: resolve the valuation dispute on the contractual procedure (use the agreed mid or a third-party determination), document a collateral policy with daily monitoring and a funding line ring-fenced for margin, and stress the downgrade trigger.

Diagnose the concentration first. The credit exposure is not the headline A$95m: after A$75m of collateral it is the unsecured portion - the A$20m threshold, the disputed A$6m and potential future exposure on a five-year swap. That is manageable for an A-rated dealer. The real concentrations are valuation and liquidity: one curve, one dispute process, one CSA, one counterparty that can call collateral on its own terms.

Novation to a clearing house

• Benefits: daily variation margin against a central counterparty, removal of bilateral CVA and dealer-specific valuation disputes, and greater confidence in the mark.
• Costs and new risks: an initial margin requirement and default fund contribution, which are new cash outflows and a new source of liquidity stress - initial margin is not returned on a normal close-out. Members and clearing fees apply, and the bank may need a clearing member. The economic hedge and its ineffectiveness profile are unchanged, and continuation of hedge accounting must be documented on novation.

Splitting across two dealers

• Benefits: competing marks reduce dispute risk and give a fallback if one dealer withdraws; it caps single-dealer dependence.
• Costs: two sets of thresholds and minimum transfer amounts mean less netting and, in aggregate, more collateral posted, not less. There is no cross-CSA netting.

Advice. For a A$800m five-year hedge with a real prepayment mismatch, clear the hedge if the bank can bear initial margin, because the valuation and dispute problems are the ones actually causing harm. If clearing is not feasible in the near term, split the notional across two dealers on matched documentation and require both to quote the disputed mid.

What the bank keeps either way: basis risk (BBSW versus its own funding spread), prepayment risk on the hedged book, and the collateral funding cost. Those are structural, not counterparty problems, and must be priced into the loan margin rather than hedged away.