NFRS 9: Impairment Models
A practical guide to calculating Expected Credit Losses (ECL) using the Provision Matrix for trade receivables and the general EAD x LGD x PD model.
This method involves categorizing receivables based on their aging (how overdue they are) and applying a historical or forward-looking loss rate to each category. This approach must be based on historical data, current conditions, and reasonable future forecasts.
Example: Building a Provision Matrix
This is a demonstrative example. Please do not use for actual Scenarios.
Using Central Bank’s Provision as a Baseline
| Pass Loans | 1% | upto 30 Days |
| Watchlist Loans | 5% | upto 3 Months |
| Substandard Loans | 25% | upto 6 Months |
| Doubtful Loans | 50% | Upto 12 Months |
| Loss Loans | 100% | After 12 Months |
The respective rates are then multiplied with an adjusting factor.
Where adjusting factor = Original Rate / k * Original Rate^p
Where, k is a scaling factor and p is a power to control adjustment intensity.
Example Calculation
K is calculated as: The Total Asset of Largest Commercial Bank for which the bank’s Provision Apply / Total Asset of The concerned Entity. Let’s just Assume, the value is 10.
P (Adjustment Intensity) is calculated as: Total Asset of the Concerned Entity / Total Asset of Lowest class of Financial Institutions for which the Provisions Applies. Let’s Assume that as 0.08.
The Adjusted rates come as:
| Classification | Allowance Rate |
|---|---|
| Prime Receivables | 0.05% |
| Standard Receivables | 0.26% |
| Watchlist receivables | 1.49% |
| Concern receivables | 3.15% |
| Doubtful receivables | 6.67% |
Write off any receivables that have been concluded as being Non-Receivables.
Allowance Calculation
| Current | Prime Receivables | Standard Receivables | Watchlist receivables | Concern receivables | Doubtful receivables |
|---|---|---|---|---|---|
| Allowance rate | 0.05% | 0.26% | 1.49% | 3.15% | 6.67% |
| Portfolio of Receivables | 22,500,000.00 | 14,500,000.00 | 2,500,000.00 | 3,500,000.00 | 7,000,000.00 |
| Allowance Amount | 10,273.47 | 37,784.96 | 37,179.82 | 110,206.20 | 466,666.67 |
Again, This is just an Example, Please don’t try to use it.
EAD x LGD x PD
Exposure at Default x Loss Given Default x Probability of Default
A. Exposure at Default (EAD)
This is the total value a company is exposed to when a debtor defaults. For loans, it's typically the outstanding principal and accrued interest at the time of default.
For our example, we'll use a total loan portfolio with an EAD of 51,133,800.
B. Loss Given Default (LGD)
This is the portion of the EAD that is expected to be lost if a default occurs. It's calculated after considering any collateral, guarantees, or other credit enhancements.
LGD = EAD - (Recoveries from Collateral + Recoveries from Guarantees/Insurance)
Example LGD Calculation:
| Loan Portfolio | EAD | FV of Collateral | Insurance Coverage | Loss Given Default |
|---|---|---|---|---|
| Portfolio A | 5,225,000 | 3,396,250 | 261,250 | 1,567,500 |
| Portfolio B | 9,856,000 | 8,574,720 | 492,800 | 788,480 |
| ... | ... | ... | ... | ... |
| Total Loss Given Default for Portfolio | 6,404,018 | |||
C. Probability of Default (PD)
This is the likelihood that a borrower will default over a specific time horizon (e.g., 12 months for Stage 1 assets, lifetime for Stage 2/3). This is often the most complex part, derived from historical data, macroeconomic models, and credit rating transitions.
Example PD Rates (illustrative):
| Risk Category / Stage | 12-Month PD |
|---|---|
| Good Loans (Stage 1) | 3.45% |
| Sub-Standard (Stage 2) | 8.90% |
| Risky Loans (Stage 3) | 34.04% |
| Loss Loans (Stage 3) | 100.00% |
| Risk Category | EAD | LGD % (LGD/EAD) | PD | ECL (EAD x LGD% x PD) |
|---|---|---|---|---|
| Good Loans | 155,083,844 | 12.5% | 3.45% | 669,459 |
| Sub-Standard | 69,038,883 | 12.5% | 8.90% | 768,435 |
| Risky Loans | 18,673,321 | 12.5% | 34.04% | 794,547 |
| Loss Loans | 7,203,952 | 12.5% | 100.00% | 900,494 |
| Total Impairment Loss to Recognize | 3,132,935 | |||
Note: LGD % is simplified here for calculation. In practice, LGD would be calculated for each specific asset or sub-portfolio.
