Banking Law and RegulationUnit 813 min read
Loan Classification, Provisioning & Risk Management: Rules, Collateral, Fraud & Recovery
Unit 8 of Banking Law and Regulation explores how banks classify loans, set aside provisions for bad debts, manage risks (credit, market, operational), and recover loans legally in Nepal—with real-world examples from Ncell, Daraz, and NRB regulations.
TAKEAWAYS:
- Loan classification in Nepal follows NRB’s 5-tier system (Standard, Sub-Standard, Doubtful, Loss, and Restructured), triggering provisioning requirements from 0% to 100% of the outstanding amount.
- Collateral valuation caps loans at 50% of liquid assets (e.g., gold, shares) or 75% for immovable property—but fraudsters exploit loopholes (e.g., fake land titles).
- Risk management tools include credit scoring models (used by banks like NMB), stress testing (for NEPSE-listed firms), and derivatives (e.g., NMB’s foreign currency hedging).
- Loan recovery in Nepal follows civil/criminal procedures: attachment of assets (Section 13 of BAFIA 2073), auction sales, or NRB’s "Debt Recovery Tribunal" for high-value defaults.
- Cyber-frauds (e.g., Khalti phishing scams) and money laundering (via eSewa’s fake merchant accounts) are combated by AML Act 2074 and BFIs’ suspicious transaction reporting.
- Provisioning shortfalls (e.g., Global IME Bank’s 2019 crisis) led to NRB’s Basel III compliance push, requiring banks to hold 8% CAR and 12.5% Tier 1 capital.
1. Loan Classification: NRB’s 5-Tier System
Banks classify loans based on repayment performance and collateral quality. The Nepal Rastra Bank (NRB) Circular 2075 defines:
| Classification | Criteria | Provisioning Requirement | Example in Nepal |
|---|---|---|---|
| Standard | Fully performing, no past dues | 0% | Ncell’s retail loan (90% of portfolio) |
| Sub-Standard | Past due ≤90 days, but recoverable | 20% | Daraz seller’s working capital loan |
| Doubtful | Past due >90 days, recovery uncertain | 50% | Kathmandu hotel’s COVID-19 loan |
| Loss | Uncollectible (written off) | 100% | Global IME’s 2019 NPL write-offs |
| Restructured | Rescheduled due to financial distress (e.g., lower EMI, extended tenure) | 25–100% (based on risk) | NMB’s farmer loan restructuring (2022) |
Why it matters:
- Provisioning = Risk buffer. If a bank classifies 10% of loans as "Doubtful," it must set aside 5% of total assets (10% × 50%).
- Basel III rules (adopted by NRB) require banks to hold capital equal to 8.5% of risk-weighted assets (RWA).
2. Collateral and Loan-to-Value (LTV) Ratios
Banks lend only up to a % of collateral value, reducing risk. NRB’s LTV limits (2076):
| Collateral Type | Max LTV | Example | Fraud Risk |
|---|---|---|---|
| Gold (24K) | 75% | NMB’s gold loan (Rs 500K loan for Rs 666K gold) | Fake gold certificates (e.g., Pokhara gold shop scams) |
| Shares (NEPSE-listed) | 50% | Standard Chartered’s share pledge loan | Overvalued shares (e.g., Nepal Bank’s 2020 share fraud) |
| Immovable Property | 75% | Global IME’s home loan | Fake land titles (e.g., Kathmandu’s "ghost plots") |
| Fixed Deposit (FD) | 90% | NIC Asia’s FD-backed loan | FD forgery (e.g., 2018 NIC Asia fraud) |
Worked Example: Daraz Seller’s Working Capital Loan
- Collateral: Rs 1M worth of inventory (valued at cost price).
- LTV Limit: 50% (for movable assets).
- Max Loan: Rs 500,000.
- Provisioning: If the loan turns Sub-Standard, the bank must set aside 20% of Rs 500K = Rs 100K.
Mermaid Diagram: Collateral Valuation Process
flowchart TD
A["Loan Application"] --> B["Bank Valuates Collateral"]
B --> C{"Collateral Type?"}
C -->|"Gold/Shares"| D["LTV = 50-75%"]
C -->|"Property"| E["LTV = 75%"]
C -->|"FD"| F["LTV = 90%"]
D --> G["Loan Approved ≤ LTV Limit"]
E --> G
F --> G
G --> H["Provisioning Applied"]3. Risk Management Tools in Nepali Banks
Banks use quantitative and qualitative tools to mitigate risks. Examples:
| Tool | How It Works | Nepal Example | Visual |
|---|---|---|---|
| Credit Scoring | Statistical models (e.g., FICO-like) rate borrowers A–E based on income, debt, and history. | NMB’s "Credit Score Nepal" (2023) | IMAGE: "Credit scoring model flowchart" |
| Stress Testing | Simulates economic shocks (e.g., 25% inflation, NPR depreciation). | NRB’s 2022 financial stability report | IMAGE: "Nepal’s NPL ratio trend graph" |
| Derivatives | Hedges against foreign exchange (FX) risk or interest rate fluctuations. | NMB’s USD/NPR futures | IMAGE: "FX swap contract example" |
| AML Filters | Flags suspicious transactions (e.g., Khalti transfers >Rs 500K in one go). | Global IME’s 2019 AML violations | IMAGE: "Red flags for money laundering" |
Real-World Example: Ncell’s Prepaid Loan Scam (2022)
- Risk: Ncell offered zero-interest loans via its app, but 90% defaulted.
- Bank’s Response:
- Reclassified loans as Sub-Standard after 30 days.
- Suspended new loans until collateral checks were tightened.
- Reported to NRB under Section 13(2) of BAFIA 2073 for misleading advertising.
4. Loan Recovery: Legal Procedures in Nepal
If a borrower defaults, banks follow civil and criminal procedures:
Step-by-Step Recovery Process
flowchart LR
A["Default Occurs"] --> B["Bank Issues Notice (15-30 days)"]
B --> C{"Response?"}
C -->|"No Response"| D["File Civil Suit (District Court)"]
C -->|"Partial Payment"| E["Negotiate Settlement"]
D --> F["Attachment of Assets (Sec. 13, BAFIA)"]
F --> G["Auction Sale (via Court-appointed valuer)"]
G --> H["Distribute Proceeds to Bank"]
E --> HKey Legal Provisions
| Stage | Law/Section | Example |
|---|---|---|
| Notice Period | BAFIA 2073, Section 12 | NMB sends 30-day notice before auction |
| Asset Attachment | BAFIA 2073, Section 13 | Global IME attaches a Kathmandu apartment |
| Criminal Action | Nepali Penal Code, Section 406 (Cheating) | Prosecution of fake gold loan defaulters |
| Debt Recovery Tribunal | NRB Act 2058, Section 45A | NIC Asia’s 2018 NPL recovery case |
Worked Example: Kathmandu Retail Shop’s Loan Default
- Loan: Rs 2M for inventory (collateral: shop + Rs 1.5M FD).
- Default: Shop closed; FD forged.
- Recovery Steps:
- Bank files civil suit under BAFIA Sec. 13.
- Court freezes shop assets (valued at Rs 3M).
- Auction sale yields Rs 2.5M → Rs 2M repaid, Rs 500K loss.
- Fraudster prosecuted under Nepal Penal Code Sec. 406.
5. Frauds and Cyber-Crimes in Banking
Top 3 Frauds in Nepal (2020–2023):
| Fraud Type | How It Works | Nepal Example | Prevention by Banks |
|---|---|---|---|
| Phishing (Khalti/eSewa) | Fake login pages steal credentials. | 2022 Rs 200M Khalti scam | OTP-based 2FA, NRB’s cybersecurity guidelines |
| Fake Collateral | Borrower pledges non-existent assets (e.g., fake land titles). | 2019 Global IME’s Rs 10B fraud | Third-party valuation, NRB’s collateral audit |
| Insider Trading | Bank employees leak loan approvals for kickbacks. | NIC Asia’s 2018 insider fraud | Whistleblower policies, NRB inspections |
AML Act 2074 Key Provisions:
- Reporting Threshold: Rs 500K+ transactions must be reported to FIU-Nepal.
- Customer Due Diligence (CDD): Banks must verify source of wealth (e.g., Ncell’s high-net-worth clients).
- Penalty: Up to Rs 10M fine or 5-year jail for money laundering (e.g., Daraz’s fake merchant accounts).
6. Provisioning and Capital Adequacy
Why Banks Fail Without Proper Provisioning:
- Global IME Bank (2019): NPL ratio = 30% → Rs 12B loss → NRB takeover.
- NIC Asia Bank (2018): Underprovisioned by Rs 8B → Liquidation.
NRB’s Basel III Requirements (2023):
| Ratio | NRB’s Minimum | Global Standard | Nepal’s 2022 Avg. |
|---|---|---|---|
| Capital Adequacy (CAR) | 8% | 8.5% | 10.2% |
| Tier 1 Capital | 6% | 6% | 8.7% |
| NPL Ratio | ≤3% | ≤5% | 2.8% (2023) |
Worked Example: NMB’s Provisioning Calculation
- Total Loans: Rs 500B
- NPL Breakdown:
- Sub-Standard: Rs 50B (20% provision)
- Doubtful: Rs 30B (50% provision)
- Total Provision Needed:
(50B × 20%) + (30B × 50%) = 10B + 15B = Rs 25B - CAR Check: If NMB has Rs 40B capital, its CAR = (40B / 500B) × 100 = 8% (meets NRB’s 8% minimum).
In the Real World
Ncell’s Loan Defaults (2022)
- Idea Used: Loan classification + provisioning.
- How: Ncell’s zero-interest prepaid loans had a 90% default rate. Banks reclassified them as "Sub-Standard" and set aside 20% provisions, costing lenders Rs 1.5B.
Daraz Seller’s Working Capital Fraud
- Idea Used: Collateral valuation + LTV limits.
- How: A Pokhara Daraz seller took a Rs 5M loan against Rs 6M inventory (LTV = 83% → illegal). When inventory was sold at 30% loss, the bank froze his account and reclassified the loan as "Doubtful".
Nepal Rastra Bank’s Stress Test (2023)
- Idea Used: Risk management + Basel III.
- How: NRB simulated a 25% NPR depreciation and found 4 banks would breach CAR <8%. They were ordered to raise capital or reduce risky loans.
Exam Tip
- Memorize NRB’s 5-tier classification and provisioning %—exams often ask for calculations (e.g., "If a Rs 1M loan is Doubtful, how much provision is needed?").
- Compare LTV limits for different collaterals (gold vs. property). Common exam trap: Asking if FD can be 100% collateral (answer: no, max 90%).
- Loan recovery process is high-yield: Expect 3–5 marks on BAFIA sections (e.g., "Describe the steps a bank takes to attach a defaulter’s property").
- Fraud examples: Link real cases (e.g., Global IME, NIC Asia) to legal provisions (e.g., BAFIA Sec. 13, AML Act 2074).
- Basel III vs. NRB’s rules: Know CAR, Tier 1 capital, and NPL limits—exams test numerical applications (e.g., "A bank has Rs 100B loans, Rs 8B NPL (all Doubtful). Calculate provisioning").
- Shortcut for provisioning:
- Sub-Standard (≤90 days): 20% of outstanding.
- Doubtful (>90 days): 50% of outstanding.
- Loss: 100% (already written off).
Final Mermaid: The Accounting Cycle for Loan Provisioning
Based on the TU BBA syllabus for Banking Law and Regulation (BNK205), unit 8.
Discussion
Loading…