Exercises — IFRS 9 Deep Practice
In Lessons 3 through 5, you learned IFRS 9 staging, the ECL formula, PD/LGD/EAD components, and macroeconomic scenario weighting. In Lesson 11, you saw how IFRS 9 interacts with Basel capital. Now you apply all of that knowledge to two extended exercises that mirror real-world portfolio assessment work: a UK retail mortgage stage assessment and a GCC corporate ECL model.
These exercises require you to make judgments — not just calculations. Staging decisions involve qualitative assessment of borrower circumstances. ECL models involve parameter choices that change the answer materially. The banking plugin's ifrs9-staging and ifrs9-ecl skills can assist with computation, but you drive the classification decisions and parameter selections.
Exercise 1: IFRS 9 Stage Assessment — Retail Mortgage Portfolio
Jurisdiction: United Kingdom (PRA regulated)
Duration: 35 minutes
Skills used: ifrs9-staging, ifrs9-ecl
Portfolio Data
You are the credit risk analyst for a UK building society. The following 8 mortgage facilities require stage assessment at the reporting date. For each facility, you have the current data and the relevant history.
| Facility | Balance | DPD | Rating at Origination | Current Rating | LTV | Additional Information |
|---|---|---|---|---|---|---|
| A001 | GBP 245K | 0 | A | A | 68% | Performing, no change in circumstances |
| A002 | GBP 380K | 0 | A | BBB | 71% | 2-notch downgrade since origination |
| A003 | GBP 190K | 35 | BBB | BB | 82% | 35 days past due, rating deteriorated |
| A004 | GBP 520K | 0 | A | A | 55% | Borrower recently notified bank of job loss |
| A005 | GBP 155K | 95 | BBB | CC | 88% | 95 days past due, severe deterioration |
| A006 | GBP 290K | 0 | BBB | BBB | 75% | Property value has fallen 15% since origination |
| A007 | GBP 410K | 0 | A | A | 61% | Performing, rate rise stress test shows affordability strain |
| A008 | GBP 175K | 62 | BB | B | 79% | 62 DPD, borrower has notified hardship |
ECL Parameters
| Parameter | Stage 1 | Stage 2 | Stage 3 |
|---|---|---|---|
| PD (12-month) | 0.8% | — | — |
| PD (lifetime) | — | 12% | 55% |
| LGD (LTV up to 80%) | 25% | 25% | 25% |
| LGD (LTV above 80%) | 40% | 40% | 40% |
Your Tasks
Step 1: Stage each facility. For each of the 8 facilities, determine whether it belongs in Stage 1, Stage 2, or Stage 3. Document your reasoning — which SICR triggers are present? Consider both quantitative indicators (DPD, PD/rating movement) and qualitative indicators (employment status, property values, hardship notification).
Quantitative triggers for Stage 2:
- Days past due of 30 or more (rebuttable presumption)
- Significant increase in PD since origination (typically 2+ notch downgrade or PD doubling)
Qualitative triggers for Stage 2:
- Significant adverse change in borrower circumstances (job loss, business failure)
- Material decline in collateral value affecting the credit decision
- Forbearance or modification granted
Stage 3 triggers (credit-impaired):
- Days past due of 90 or more
- Borrower in default (per the bank's default definition)
- Significant financial difficulty of the borrower
Step 2: Calculate ECL for each facility. Apply the appropriate ECL measurement:
- Stage 1: 12-month ECL = Balance x PD(12m) x LGD
- Stage 2: Lifetime ECL = Balance x PD(lifetime) x LGD
- Stage 3: Lifetime ECL = Balance x PD(lifetime) x LGD
Select the correct LGD based on the facility's LTV (25% for LTV up to 80%, 40% for LTV above 80%).
Step 3: Sensitivity analysis. If A006's property value falls a further 10% (LTV rises from 75% to approximately 88%), does the staging change? What is the ECL impact?
Step 4: Produce a provision movement summary. Create a stage distribution table showing:
| Stage | Number of Facilities | Total Balance | Total ECL |
|---|---|---|---|
| Stage 1 | ? | ? | ? |
| Stage 2 | ? | ? | ? |
| Stage 3 | ? | ? | ? |
| Total | 8 | ? | ? |
Step 5: Draft an IFRS 7 note. Write a one-paragraph summary suitable for the financial statements disclosure, describing the key credit risk characteristics of the portfolio and the staging distribution.
Exercise 2: Corporate ECL Model — GCC
Jurisdiction: Gulf Cooperation Council (multi-jurisdiction: KSA, UAE, Bahrain)
Duration: 55 minutes
Skills used: ifrs9-ecl, ifrs9-scenarios
Portfolio Data
You are building an ECL model for a GCC-based bank's corporate loan book. The bank operates across Saudi Arabia, UAE, and Bahrain. Six borrowers represent the key segments.
| Borrower | Country | TTC Rating | Drawn (USD M) | Undrawn (USD M) | Collateral | Sector |
|---|---|---|---|---|---|---|
| Al-Jazira Steel | KSA | BB+ | 85 | 25 | 40% plant and machinery | Manufacturing |
| Dubai Towers | UAE | BBB | 165 | 0 | Property LTV 72% | Real Estate |
| Gulf Energy | Bahrain | A- | 210 | 50 | Cash deposit + govt guarantee | Energy |
| Riyadh Retail | KSA | BB | 45 | 30 | Stock and receivables | Retail |
| Emirates Logistics | UAE | BBB- | 120 | 40 | Fleet assets 60% of drawn | Transport |
| Manama Finance | Bahrain | BB- | 60 | 20 | Nil collateral | Financial Services |
Through-the-Cycle PD Table
| Rating | TTC PD |
|---|---|
| A- | 0.4% |
| BBB | 0.9% |
| BBB- | 1.3% |
| BB+ | 1.8% |
| BB | 2.8% |
| BB- | 4.2% |
GCC Macro Adjustment
The GCC region's credit risk is sensitive to oil prices and GDP growth. Apply a base PD multiplier of 1.15 to convert TTC PDs to point-in-time PDs reflecting current GCC conditions.
Credit Conversion Factor (CCF): Apply 75% to all undrawn commitments.
EAD Calculation: EAD = Drawn + (Undrawn x CCF)
Macroeconomic Scenarios
| Scenario | Oil Price | GDP Growth | PD Multiplier | Probability Weight |
|---|---|---|---|---|
| Upside | $95/barrel | 3.5% | 0.85 | 20% |
| Base | $75/barrel | 2.2% | 1.15 | 50% |
| Adverse | $50/barrel | 0.5% | 1.65 | 30% |
LGD Assumptions
| Collateral Type | LGD |
|---|---|
| Cash deposit / govt guarantee | 15% |
| Property (LTV < 80%) | 30% |
| Plant and machinery | 45% |
| Stock and receivables | 55% |
| Fleet assets | 40% |
| Nil collateral | 70% |
Your Tasks
Step 1: Calculate EAD for each borrower. Apply the 75% CCF to undrawn commitments and add to drawn balances.
Step 2: Determine point-in-time PD under each scenario. For each borrower:
- Start with the TTC PD from the rating table
- Apply the scenario-specific PD multiplier
For example, Al-Jazira Steel (BB+, TTC PD 1.8%):
- Upside PD: 1.8% x 0.85 = 1.53%
- Base PD: 1.8% x 1.15 = 2.07%
- Adverse PD: 1.8% x 1.65 = 2.97%
Step 3: Calculate ECL under each scenario. For each borrower under each scenario:
- ECL = EAD x PD(scenario) x LGD
Step 4: Calculate probability-weighted ECL. For each borrower:
- Weighted ECL = (Upside ECL x 20%) + (Base ECL x 50%) + (Adverse ECL x 30%)
Step 5: Aggregate and analyze. Produce a summary table:
| Borrower | EAD | Upside ECL | Base ECL | Adverse ECL | Weighted ECL |
|---|---|---|---|---|---|
| Al-Jazira Steel | ? | ? | ? | ? | ? |
| ... | ... | ... | ... | ... | ... |
| Total | ? | ? | ? | ? | ? |
Step 6: Management discussion questions.
- Which borrower contributes the most ECL? Why?
- How sensitive is the total ECL to the oil price assumption?
- If you were to apply a Post-Model Adjustment (PMA), which sector would you target and why?
Try With AI
Use these prompts in Claude or your preferred AI assistant to work through these exercises.
Prompt 1: Staging the Mortgage Portfolio
I have 8 UK mortgage facilities to stage under IFRS 9. Here is
the data for each:
A001: GBP 245K, 0 DPD, A to A, LTV 68%, performing
A002: GBP 380K, 0 DPD, A to BBB, LTV 71%, 2-notch downgrade
A003: GBP 190K, 35 DPD, BBB to BB, LTV 82%, 35 days past due
A004: GBP 520K, 0 DPD, A to A, LTV 55%, borrower lost job
A005: GBP 155K, 95 DPD, BBB to CC, LTV 88%, severe deterioration
A006: GBP 290K, 0 DPD, BBB to BBB, LTV 75%, property down 15%
A007: GBP 410K, 0 DPD, A to A, LTV 61%, rate rise stress
A008: GBP 175K, 62 DPD, BB to B, LTV 79%, hardship notified
For each facility, tell me:
1. Which stage (1, 2, or 3) and why
2. Which SICR triggers are present (quantitative and qualitative)
3. The ECL using: Stage 1 PD 0.8%, Stage 2 lifetime PD 12%,
Stage 3 lifetime PD 55%, LGD 25% (LTV<=80%) or 40% (LTV>80%)
Then produce a summary table showing total ECL by stage.
What you are learning: Staging is not mechanical — facilities like A004 (no DPD, no rating change, but job loss) and A006 (no DPD, no rating change, but property value decline) require qualitative judgment. The AI can assist with computation, but the staging decision reflects your professional judgment about whether a significant increase in credit risk has occurred.
Prompt 2: Building the GCC Corporate ECL Model
Build an ECL model for these 6 GCC corporate borrowers:
Al-Jazira Steel: KSA, BB+ (TTC PD 1.8%), drawn $85M, undrawn
$25M, collateral: plant 45% LGD
Dubai Towers: UAE, BBB (0.9%), drawn $165M, no undrawn,
property LTV 72% -> 30% LGD
Gulf Energy: Bahrain, A- (0.4%), drawn $210M, undrawn $50M,
cash+guarantee -> 15% LGD
Riyadh Retail: KSA, BB (2.8%), drawn $45M, undrawn $30M,
stock/receivables -> 55% LGD
Emirates Logistics: UAE, BBB- (1.3%), drawn $120M, undrawn $40M,
fleet -> 40% LGD
Manama Finance: Bahrain, BB- (4.2%), drawn $60M, undrawn $20M,
nil -> 70% LGD
CCF for undrawn: 75%
Three scenarios: Upside (PD x 0.85, 20%), Base (PD x 1.15, 50%),
Adverse (PD x 1.65, 30%)
Calculate EAD, ECL under each scenario, and probability-weighted
ECL for each borrower. Show a summary table.
What you are learning: The GCC macro adjustment (PD multiplier) demonstrates how regional economic conditions translate into credit risk parameters. The probability-weighted ECL is always higher than the base case ECL because of the non-linearity effect — the adverse scenario's higher weight relative to upside pulls the weighted average above the base case. This is the same principle from Lesson 5, now applied to a real multi-borrower portfolio.
Prompt 3: Sensitivity and PMA Design
Using the GCC corporate ECL model results, answer:
1. Which borrower contributes the most to total ECL and why?
2. If oil drops to $35/barrel (PD multiplier 2.20), what is the
new adverse scenario ECL? How much does the total weighted
ECL increase?
3. Design a Post-Model Adjustment (PMA) for the real estate
sector (Dubai Towers). The model uses a 30% LGD based on
current property values, but GCC real estate has shown
40% price declines in past downturns. What PMA amount
would you recommend and what governance documentation
is required?
What you are learning: Sensitivity analysis reveals concentration risk — which borrower, sector, or scenario drives the most ECL volatility. PMA design bridges the gap between what the model captures and what professional judgment identifies as missing. The governance documentation requirement (approval, time-limiting, back-testing) ensures PMAs are not arbitrary adjustments but disciplined overlays.