StockEducation
NRB Key Ratios

Chapter 5 · Part 5 — NPL, Net NPL and Provision Coverage

NPL, Net NPL and Provision Coverage

The distance between two columns is the closest thing the KFI table has to a measure of accounting honesty.

5 of 7 · 10 min

Two columns, and the distance between them is the most revealing number on the whole sheet. NPL tells you how much of the loan book has gone wrong. Net NPL tells you how much of that the bank has already admitted to. The gap tells you how honestly it is being carried.

The two formulas

Straight from the KFI footnotes:

┌────────────────────────────────────────────────────────────┐
│                                                            │
│                     Non Performing Loan                    │
│      NPL %     =  ──────────────────────────  × 100        │
│                        Total Loan                          │
│                                                            │
│                                                            │
│                   NPL − Specific Provision                 │
│      Net NPL % =  ──────────────────────────  × 100        │
│                        Total Loan                          │
│                                                            │
│      Same denominator. The numerator of the second is      │
│      the first, less what has already been provided for.   │
│                                                            │
└────────────────────────────────────────────────────────────┘

Note the Net NPL formula carefully: the specific provision is subtracted from the numerator, but the denominator stays Total Loan — not "total loan less provision". This is NRB's definition and it is what the sheet prints.

What makes a loan non-performing

A loan is graded by how far behind it is. Performing loans sit in the top grade; everything below is non-performing.

THE CLASSIFICATION LADDER                              [R]
(structure is stable; the day-counts and the provision
 percentages are set by directive and change)

┌────────────────┬──────────────────────┬────────────────────┐
│  GRADE         │  ROUGHLY             │  PROVISION         │
├────────────────┼──────────────────────┼────────────────────┤
│  PASS          │  current, or only    │  a small general   │
│                │  slightly overdue    │  provision         │
│  ─────────────────── PERFORMING ────────────────────────── │
│  ─────────────────── NON-PERFORMING ────────────────────── │
│  SUBSTANDARD   │  overdue beyond the  │  larger            │
│                │  first threshold     │                    │
│  DOUBTFUL      │  longer overdue      │  larger still      │
│  LOSS          │  considered          │  the largest —     │
│                │  unrecoverable       │  typically full    │
└────────────────┴──────────────────────┴────────────────────┘

NPL = Substandard + Doubtful + Loss

Two kinds of provision matter here, and conflating them is a common error:

  • Specific provision is held against identified bad loans — the ones in the three non-performing grades. This is what Net NPL subtracts.
  • General provision is held against the performing book, for losses that exist statistically but have not been identified yet. This is not subtracted in Net NPL, and (subject to limits) it is the item that can count toward Tier 2 capital, which is why it turned up on Day 2.

The gap is the coverage

Since both ratios share a denominator, subtract them:

NPL % − Net NPL %  =  Specific Provision ÷ Total Loan × 100

... and as a proportion of the bad loans themselves:

                         NPL % − Net NPL %
Provision coverage  =  ─────────────────────  × 100
                              NPL %

That second ratio is the one to compute. It answers: of every rupee of bad loan, how many paisa has the bank already set aside?

THE GAP, DRAWN

NPL 8.85%  ├████████████████████████████████████████┤
                                                    │
already provided for (specific provision)           │
           ├██████████████████████████████████┤     │
                                           8.36 pp  │
                                                    │
still exposed (Net NPL)                             │
                                       ├───┤ 0.49 pp

coverage = 8.36 ÷ 8.85 = 94.5%

The banks, ranked by what the gap reveals

ASSET QUALITY — commercial banks, Chaitra end 2082
sorted by NPL

bank                    NPL %   Net NPL %   gap    coverage
─────────────────────────────────────────────────────────────
NIC Asia                 8.85       0.49    8.36     94.5%
Prabhu Bank              8.84       1.71    7.13     80.7%
NIMB                     8.41       2.28    6.13     72.9%
Himalayan                7.98       1.48    6.50     81.5%
Kumari                   6.94       2.69    4.25     61.2%
Prime Commercial         6.39       1.94    4.45     69.6%
Citizens                 6.82       2.37    4.45     65.2%
Laxmi Sunrise            5.29       1.39    3.90     73.7%
Nepal Bank               4.96       1.32    3.64     73.4%
NMB                      4.95       1.77    3.18     64.2%
ADBL                     5.06       1.31    3.75     74.1%
RBB                      4.48       1.35    3.13     69.9%
Nabil                    4.37       0.98    3.39     77.6%
Machhapuchhre            4.01       1.22    2.79     69.6%
Sanima                   3.99       1.49    2.50     62.7%
Siddhartha               3.71       1.18    2.53     68.2%
Nepal SBI                2.53       0.23    2.30     90.9%
Standard Chartered       1.81       0.37    1.44     79.6%
Everest                  0.61       0.22    0.39     63.9%
─────────────────────────────────────────────────────────────
System                   5.41       1.36    4.05     74.9%

Read this table twice.

First pass — the NPL column alone. NIC Asia (8.85%) looks like the worst book in the system; Everest (0.61%) the best. A fourteen-fold difference.

Second pass — with coverage. NIC Asia has provided for 94.5% of its bad loans; its Net NPL is 0.49%, the third-lowest on the sheet. Kumari's NPL is 6.94% — noticeably lower than NIC Asia's — but coverage is only 61.2%, so its Net NPL is 2.69%, the highest of the twenty.

THE RANKING FLIPS

By NPL (gross)              By Net NPL (after provision)
worst first                 worst first
──────────────              ────────────────────────────
1. NIC Asia      8.85       1. Kumari          2.69
2. Prabhu        8.84       2. Citizens        2.37
3. NIMB          8.41       3. NIMB            2.28
4. Himalayan     7.98       4. Prime           1.94
5. Citizens      6.82       5. NMB             1.77
...                         ...
17. Nepal SBI    2.53       15. NIC Asia       0.49  ← was 1st
18. SCB          1.81       18. SCB            0.37
19. Everest      0.61       19. Everest        0.22

NIC Asia moves from worst to fifteenth. Kumari moves from
fifth to worst. Same banks, same date, different question.

Which ranking is right? Both — they answer different questions.

  • Gross NPL measures how much lending went wrong. It is a judgement on

credit underwriting.

  • Net NPL measures how much of that is still exposed. It is a judgement on

how much loss is still to come through the P&L.

  • Coverage measures how much the bank has already recognised. It is the closest thing on this sheet to a judgement on accounting conservatism.

Worked example — converting the gap into rupees

Illustrative arithmetic on published figures.

NIC ASIA BANK — Chaitra end 2082

Total loan                Rs 220,181 m
NPL                             8.85%
Net NPL                         0.49%

Gross bad loans   =  220,181 × 0.0885   =  Rs 19,486 m
Still exposed     =  220,181 × 0.0049   =  Rs  1,079 m
                                           ────────────
Specific provision already held          =  Rs 18,407 m

Coverage  =  18,407 ÷ 19,486  =  94.5%

Put that provision next to the capital:
Core capital                             =  Rs 15,458 m

The bank has provided MORE against bad loans (18,407) than
it holds in core capital (15,458). That provision has
already passed through the P&L in earlier periods — which is
a large part of why core capital is only 6.48% of RWE.

Now the same arithmetic for a bank with weaker coverage:

KUMARI BANK — Chaitra end 2082

Total loan                Rs 305,287 m
NPL                             6.94%
Net NPL                         2.69%

Gross bad loans   =  305,287 × 0.0694   =  Rs 21,187 m
Still exposed     =  305,287 × 0.0269   =  Rs  8,212 m
Specific provision                       =  Rs 12,975 m
Coverage                                 =      61.2%

WHAT IF coverage had to rise to the system's 74.9%?

Required provision  =  21,187 × 0.749   =  Rs 15,869 m
Currently held                           =  Rs 12,975 m
                                           ────────────
Additional charge                        =  Rs  2,894 m

Core capital                             =  Rs 33,495 m
That charge is 8.6% of core capital, and it would come
straight out of profit.

This is exactly the calculation a supervisor or an auditor runs, and it is available to any investor with the KFI sheet and a calculator. It is also the single most useful thing in this course for anyone holding bank shares.

The finance companies

NPL — Class "C", Chaitra end 2082, worst first

Janaki Finance              59.12%   ████████████████████████████
Pokhara Finance             34.34%   ████████████████
Gorkhas Finance             17.71%   ████████
Nepal Finance               14.32%   ███████
Central Finance             14.17%   ███████
Best Finance                13.08%   ██████
Reliance Finance            10.34%   █████
─────── Class C total 9.81% ──────────────────
Goodwill Finance             9.69%   ████▊
Samriddhi Finance            8.09%   ████
Multipurpose Finance         7.44%   ███▋
Guheshwori Merchant          6.37%   ███
Progressive Finance          5.58%   ██▋
Shree Investment             4.02%   ██
Manjushree Finance           3.99%   ██
ICFC Finance                 2.68%   █▎
─────────────────────────────────────────────
Commercial banks   5.41%  for comparison

The Class "C" average NPL of 9.81% is roughly 1.8 times the commercial banks' 5.41%, and the dispersion is far wider — from ICFC's 2.68% to Janaki's 59.12%, a twenty-two-fold range. That dispersion is the reason the two sheets cannot be read with the same instincts.

Note also what the finance-company sheet does not have: a Net NPL column. You cannot compute coverage for a Class "C" institution from this release. The gross number is all you get, which means the softest number on the sheet comes with no cross-check.

The endgame, in one table

The two "Problematic" institutions complete the picture that Day 3 started.

┌───────────────────────────────────────────────────────────────┐
│  WHEN NPL APPROACHES 100%                                     │
│                                                               │
│                        Nepal Share      Capital Merchant      │
│                        Markets & Fin    Banking & Finance     │
│  ───────────────────────────────────────────────────────────  │
│  NPL                        98.48%           100.00%          │
│  Total loan            Rs 2,192,253 k    Rs 663,530 k         │
│  Total deposit         Rs   255,340 k    Rs  27,190 k         │
│  Net liquidity              43.51%           604.89%          │
│  CCAR                       42.96%            69.25%          │
│                                                               │
│  100.00% NPL means every rupee lent is non-performing.        │
│  Not most of it. All of it.                                   │
│                                                               │
│  Deposits have gone: Rs 27 million remains against Rs 935     │
│  million of paid-up capital. The loan book generates          │
│  nothing. The capital ratio is high because RWE collapsed.    │
│  The liquidity ratio is high because deposits collapsed.      │
│                                                               │
│  Every "good-looking" ratio here is good-looking for a        │
│  bad reason.                                                  │
└───────────────────────────────────────────────────────────────┘

This is the clearest illustration available of the discipline this course is really teaching: read the columns together, never one at a time. Any single ratio can be flattered. The combination cannot.

Formulas from today

NPL %          =  Non Performing Loan ÷ Total Loan × 100

Net NPL %      =  (NPL − Specific Provision) ÷ Total Loan × 100

Gap (pp)       =  NPL % − Net NPL %
               =  Specific Provision ÷ Total Loan × 100

Coverage %     =  (NPL % − Net NPL %) ÷ NPL % × 100

Bad loans (Rs) =  Total Loan × NPL %
Provision (Rs) =  Total Loan × (NPL % − Net NPL %)
Still exposed  =  Total Loan × Net NPL %

Top-up to peer coverage
               =  Bad loans × (peer coverage − own coverage)
                  ... then compare to core capital

What you should be able to do now

  • State both formulas and explain why the denominator is the same.
  • Explain the difference between specific and general provision, and which one Net NPL subtracts.
  • Compute provision coverage and explain what it says that NPL alone does not.
  • Show how the gross and net rankings can invert, and say which question each answers.
  • Convert the gap into rupees, size a coverage top-up, and compare it to core capital.
  • Explain why a 100% NPL institution can post the best capital and liquidity ratios in its class.

Tomorrow: base rate and spread — what a bank charges, what it keeps, and the cap that sits over the difference.

Saved in this browser only — there is no account to create. Clearing your browser data clears your progress.