Covenant Waiver at Marchmont Engineering
Marchmont Engineering Ltd is a privately owned UK manufacturer with an £8m term loan and a £2m revolving credit facility, of which £1.5m is drawn. At origination the bank's model assigned a twelve-month PD of 0.8% and an LGD of 45%. A key customer has now been lost: EBITDA has fallen from £6.8m to £4.1m, leverage has risen from 2.9x to 5.8x against a 4.0x covenant, and interest cover is 1.1x against a 1.5x covenant. Both covenants will be breached at the next test in three weeks; no payment is yet overdue. The sponsor has conditionally agreed a £2m equity injection but will not complete it unless the bank waives the covenants and grants a twelve-month amortisation holiday. The relationship manager argues the bank should support the business because the alternative is administration with an estimated recovery of 50 pence in the pound. The bank's revised view is a twelve-month PD of 6% with LGD unchanged at 45%, and a 50% credit conversion factor on the undrawn commitment.
1. Calculate the twelve-month expected loss on the facility and explain what has driven the change since origination, including the sensitivity of the answer.
Read the worked answer
Exposure at default. Drawn term loan £8.0m + drawn RCF £1.5m = £9.5m. Undrawn commitment £0.5m × 50% CCF = £0.25m. EAD = £9.75m. (Using £10.0m by ignoring the CCF is a common error; the undrawn line is a real, if partial, exposure.)
Expected loss. EL = PD × LGD × EAD = 6% × 45% × £9.75m = £263,250, say £263k.
At origination. 0.8% × 45% × £9.75m ≈ £35k (or £36k on £10m). Expected loss has risen roughly sevenfold, driven almost entirely by PD.
What drove it. PD moved from 0.8% to 6% because the customer loss cut EBITDA by 40%, leverage nearly doubled to 5.8x against a 4.0x covenant, and interest cover fell to 1.1x - the business may not service debt from operating cash flow. Note that the LGD was held constant at 45%, which is questionable: a distressed manufacturer's enterprise value and collateral coverage typically deteriorate alongside earnings, and the 50p-in-the-pound recovery estimate implies an LGD nearer 50% and possibly higher after costs. Sensitivities worth stating: at LGD 50%, EL = £292k; at LGD 60%, EL = £351k; at PD 10% and LGD 50%, EL = £488k. The point for the committee is that £263k is a floor, not a central estimate, and the LGD assumption should be challenged.
2. Determine the appropriate IFRS 9 stage and prudential default status if the waiver and amortisation holiday are granted, giving your reasoning.
Read the worked answer
IFRS 9. Two questions must be separated: significant increase in credit risk (SICR), and credit impairment.
SICR is clearly present - the loss of a major customer, a 40% fall in EBITDA, a covenant breach and a sevenfold rise in PD are all qualitative and quantitative indicators. The facility therefore moves to Stage 2 at minimum, with lifetime expected credit loss rather than twelve-month ECL. That significantly increases the provision beyond the £263k computed on a twelve-month basis.
The harder question is whether granting the concession makes it Stage 3. The waiver and payment holiday are being given because of the borrower's financial difficulty - that is the definition of forbearance, and a distressed restructuring is an indicator of credit impairment. The fact that no payment is yet overdue is not decisive; IFRS 9 impairment turns on whether there is objective evidence of impairment, not on days past due alone. On these facts the balance of argument favours Stage 3, with interest income recognised on the net carrying amount.
Prudential. Distressed restructuring is a default trigger: the bank is granting a concession it would not otherwise give because the borrower cannot comply. Granting the waiver therefore likely produces a default classification even with payments current, and the exposure becomes non-performing for NPL and forbearance reporting. A probation period (typically 90 days of compliance, and longer for forbearance) applies before returning to performing status. The committee must decide this knowingly: approving the waiver is approving a default classification, with consequential effects on ECL, NPL ratios and regulatory reporting.
3. Recommend a credit decision with conditions, justifying the structure you propose.
Read the worked answer
Recommendation: approve a conditional, time-limited waiver - not an unconditional one.
Why support rather than enforce: enforcement yields an estimated 50p in the pound, roughly £4.9m on £9.75m EAD, against a going-concern exposure that a £2m injection and a deleveraging plan could restore. The expected loss on support is materially lower, and the bank retains the relationship. But support must buy control, not just time.
Conditions.
- The £2m equity injection completes before the waiver takes effect - conditionality must be the borrower's, not the bank's alone.
- A credible deleveraging plan with a named asset sale or cost programme, and a path back below 4.0x within twelve months.
- Revised covenants set at achievable but ratcheting levels (for example 5.0x stepping to 4.5x then 4.0x), tested quarterly, with tighter interest cover.
- Enhanced security or a sponsor guarantee, and a margin ratchet to price the added risk.
- Monthly management information, a thirteen-week cash flow forecast, and early-warning triggers that escalate automatically.
- No dividends, no shareholder distributions, no new debt ranking ahead.
- The RCF undrawn portion frozen or cancelled to prevent further drawdown.
Also required. Classify as forborne, place on a watchlist with a named relationship and credit owner, review at each covenant test, and ensure the provision reflects the Stage 2 or Stage 3 treatment. The relationship manager's recommendation is economically sound but must be re-expressed as a structured, monitored exposure rather than a concession granted on optimism.