Chapter 6 · Part 6 — NRB Ratios and Performance Ratios
Interest spread, Tier 1, CET 1, ROE and ROA
Why spread is not margin, why CET 1 is the only capital number that matters, and how to decompose a return.
Average Interest Spread
Simple definition. The gap between what the bank charges borrowers and what it pays depositors.
Technical definition. The difference between the weighted average lending rate and the weighted average deposit rate, calculated as per NRB Directives and measured on the last month of the quarter, as the report's own label states.
Formula
Average Interest Spread = Weighted Average Lending Rate
− Weighted Average Deposit RateSpread vs Net Interest Margin — do not confuse them.
INTEREST SPREAD NET INTEREST MARGIN (NIM)
─────────────── ─────────────────────────
Lending rate − Deposit rate Net interest income
÷ Average earning assets
Compares two RATES Measures return on the ASSET BASE
Ignores the benefit of funding CAPTURES the benefit of free
assets with FREE money funding (current accounts, equity)
→ NIM is normally HIGHER than spread.
The gap between them is the value of zero-cost funding.Reading NMB's numbers — the most under-appreciated line in the table.
FY2082/83 FY2081/82 Change Both Group & Bank 3.43% 3.82% −0.39pp
The spread FELL by 39 basis points — while net interest income ROSE 17.5%. How?
Deposit rate fell a lot (cost of funds −132bp)
Lending rate fell EVEN MORE
│
▼
Spread compressed from 3.82% to 3.43%
│
▼
But NII still rose, because:
(a) the loan BOOK grew ~10% — more volume at a thinner spread
(b) the timing of repricing favoured the bank during the year
(deposits repriced down before the full effect of lower
lending rates flowed through interest income)This is the crucial forward-looking signal in the whole ratio table. Volume growth and repricing timing masked genuine spread compression. As the repricing effect exhausts and if loan growth slows, NII growth will not merely slow — it can reverse.
Interpretation.
WIDER spread → more profitable lending
→ BUT may reflect weak competition, or
lending to riskier borrowers at higher rates
→ NRB may view an excessive spread as
exploitative and cap it [R]
NARROWER spread → competitive pressure
→ the bank must grow VOLUME to hold NII
→ volume growth at thin spreads is how
credit quality deterioratesThe volume-vs-spread trap.
Year 1: Book NPR 200bn × spread 3.82% → NII proxy 7.64bn
Year 2: Book NPR 220bn × spread 3.43% → NII proxy 7.55bn
─────────────────
10% book growth did NOT offset a 39bp spread fall.
To hold NII flat at 3.43%, the book must grow to:
7.64 ÷ 3.43% = NPR 222.7bn → 11.4% growth required
→ Each year of spread compression demands MORE growth just
to stand still. And growth at thin spreads is usually
bought with weaker underwriting — which is exactly what
rising NPL suggests is happening.Regulatory significance. NRB monitors spread as a measure of whether banks are passing rate benefits to borrowers and depositors, and has used spread caps as a policy tool [R].
Investor/analyst significance. The clearest measure of pricing power. A bank that holds its spread while competitors' compress has a genuine franchise. NMB's compression is system-wide, not bank-specific, but it constrains everyone.
Limitations.
- Not NIM. Does not capture free funding.
- Last-month-of-quarter measurement is a snapshot, not a period average.
- NRB methodology-dependent [R] and not reproducible from published statements.
- Says nothing about the risk embedded in the lending rate — a high spread may be compensation for high expected losses.
Related terms. Part 2.3 Net interest income · Cost of Funds · Base Rate · Part 17
ADDITIONAL INFORMATION
Tier 1 Capital to RWA
Simple definition. The bank's highest-quality, permanent capital as a share of its risk-weighted assets.
Technical definition. Core capital — comprising Common Equity Tier 1 and Additional Tier 1 instruments — expressed as a percentage of total risk-weighted assets under NRB's Capital Adequacy Framework.
Formula
Tier 1 Capital (CET1 + AT1)
Tier 1 Capital to RWA = ────────────────────────────────── × 100
Risk Weighted AssetsWhat qualifies as Tier 1.
TIER 1 = CET1 + AT1
CET1 — the purest loss absorber
• Paid-up ordinary share capital
• Share premium
• Statutory / general reserve
• Eligible retained earnings and other reserves
• LESS goodwill, DTA and other deductions [R]
AT1 — loss-absorbing but not common equity
• Perpetual instruments with discretionary,
non-cumulative distributions
• NMB's PNCPS: NPR 3 billion, 8.25%, perpetual,
non-cumulative → qualifies [R]Why Tier 1 matters more than total capital. Tier 1 absorbs losses while the bank is still operating (going-concern capital). Tier 2 largely absorbs losses only in liquidation (gone-concern capital). A regulator protecting depositors cares most about going-concern capital.
Reading NMB's numbers.
FY2082/83 FY2081/82 Change
Bank 9.90% 9.05% +0.85pp
Group 9.90% 9.09% +0.81pp
Decomposition:
Tier 1 9.90%
CET 1 8.99%
─────
AT1 0.91% ← the PNCPS contributionInvestor/analyst significance. The 85bp improvement is entirely the AT1 issue. Strip it out and Tier 1 would have been ~8.99% — essentially flat. The bank did not earn its way to a stronger capital position; it issued its way there.
Limitations. Same as CAR (6.1): prescribed risk weights, point-in-time measurement, unverifiable RWA.
Related terms. Capital fund to RWA · CET 1 · Part 1.30 Share capital · Part 12 Tier 1, AT1
CET 1 Capital to RWA
Simple definition. The very best kind of capital — ordinary shareholders' money — as a share of risk-weighted assets.
Technical definition. Common Equity Tier 1 capital, being ordinary paid-up capital, share premium, eligible reserves and retained earnings less prescribed regulatory deductions, expressed as a percentage of risk-weighted assets.
Formula
CET 1 Capital
CET 1 Ratio = ──────────────────────────────── × 100
Risk Weighted AssetsWhy CET1 is the number that matters most.
┌────────────────────────────────────────────────────────────────┐ │ LOSS-ABSORPTION HIERARCHY │ │ │ │ Losses hit in this order: │ │ │ │ 1. CET1 ─── ordinary shareholders │ │ ↓ absorb FIRST and FULLY, while the bank operates │ │ │ │ 2. AT1 ─── perpetual preference holders │ │ ↓ distributions can be cancelled; may be written │ │ down or converted [R] │ │ │ │ 3. Tier 2 ── subordinated debt holders │ │ ↓ absorb mainly in liquidation │ │ │ │ 4. Depositors ─ protected by all of the above │ │ │ │ → CET1 is the ONLY layer that absorbs losses fully, │ │ immediately, without any contractual trigger. │ └────────────────────────────────────────────────────────────────┘
Reading NMB's numbers — the finding that reframes the whole table.
FY2082/83 FY2081/82 Change Bank 8.99% 9.05% −0.06pp ▼ Group 9.02% 9.09% −0.07pp ▼
CET1 fell. While total capital rose 81bp and Tier 1 rose 85bp, the purest capital layer declined.
Why:
RWA grew (loan book +10%, and NPL formation raises risk weights)
│
CET1 grew more slowly because:
• 55% of profit was appropriated to reserves (still CET1, so
neutral) but…
• NPR 918m left as cash dividend → CET1 down
• Goodwill and DTA deductions continue → CET1 down [R]
• The NPR 3bn raised was AT1, NOT CET1 → no CET1 benefit
│
▼
Net: CET1 grew slower than RWA → ratio fell.What this means going forward. The bank has used its AT1 headroom. To keep growing its loan book it must now either (a) retain more profit — meaning lower dividends, (b) issue ordinary equity — diluting shareholders, or (c) slow lending growth. This is the single most important forward-looking conclusion available from the ratio table.
Required level. NRB prescribes a minimum CET1 ratio plus buffer requirements. Verify.
Investor/analyst significance. Always compare CET1 across banks, never total CAR. A bank can dress up total capital with AT1 and Tier 2; CET1 cannot be dressed up.
Limitations. As for CAR.
Related terms. Tier 1 · Capital fund to RWA · Part 1.14 Goodwill · Part 1.15 Deferred tax assets · Part 12 CET 1
Return on Equity (Annualized)
Simple definition. How much profit the bank makes for each rupee of shareholders' money.
Technical definition. Annualised profit attributable to ordinary equity holders, expressed as a percentage of average shareholders' equity, computed per the bank's stated methodology which — per the source report — deducts the PNCPS dividend and considers only ordinary shares.
Formula
Annualised profit attributable to ordinary equity holders
ROE = ───────────────────────────────────────────────────────────────────── × 100
Average ordinary shareholders' equityComponent explanation.
- Numerator — profit for the period, annualised, less the PNCPS dividend (the report is explicit that this deduction is made for ROE as well as EPS)
- Denominator — shareholders' equity attributable to the parent. Practice varies between average and closing equity [R] — the report does not state which; assume NRB's prescribed method.
- Annualised — at Q4 the factor is 1
Example calculation. Illustrative — HCBL.
Profit for the period NPR 4,013,671,000
Less PNCPS dividend NPR 122,055,000
─────────────────
Profit to ordinary shareholders NPR 3,891,616,000
Equity attributable to equity holders:
Opening NPR 32,174,608,000
Closing NPR 38,059,771,000
Less PNCPS (closing) NPR 3,000,000,000
Ordinary equity (closing) NPR 35,059,771,000
Average ordinary equity ≈ (32,174,608 + 35,059,771) ÷ 2
= NPR 33,617,190,000
ROE = 3,891,616,000 ÷ 33,617,190,000 × 100 = 11.58%This reproduces the reported Bank ROE of 11.58% — confirming the methodology.
Reading NMB's numbers.
FY2082/83 FY2081/82 Change Bank 11.58% 9.34% +2.24pp Group 15.30% 9.64% +5.66pp
Why is Group ROE so much higher? Because the subsidiaries earn a higher return on the equity employed in them. NMB Capital (merchant banking) and N.M.B. Securities (broking) are fee businesses with very little capital tied up — high ROE by construction. The microfinance subsidiary also runs at higher spreads than the bank.
Bank ROE 11.58% ← capital-intensive lending
Group ROE 15.30% ← blended with capital-light fee businesses
and higher-spread microfinanceThis is a genuine strategic insight: the non-bank businesses are the more capital-efficient part of this group.
The DuPont decomposition — where the ROE improvement came from.
ROE = ROA × Equity Multiplier
Bank:
FY2082/83: 1.04% × (Assets ÷ Equity)
Assets 397,381,306 ÷ Equity 38,059,771 = 10.44×
1.04% × 10.44 = 10.86% (approximates 11.58%; the
difference is the PNCPS adjustment and averaging)
FY2081/82: 0.88% × (351,592,507 ÷ 32,174,608 = 10.93×)
0.88% × 10.93 = 9.62% (approximates 9.34%)
→ Leverage FELL (10.93× → 10.44×) — the bank became LESS levered
→ So the entire ROE improvement came from ROA, i.e. from
PROFITABILITY, not from gearing up.Interpretation.
HIGHER ROE → efficient use of shareholders' capital
→ BUT can be achieved by excessive LEVERAGE,
which is risk, not skill
→ always decompose: ROE = ROA × leverage
LOWER ROE → over-capitalised, or poor profitability
→ a bank with a very high CAR will mechanically
show a lower ROEThe capital–ROE tension.
┌──────────────────────────────────────────────────────────────┐ │ REGULATOR wants HIGH capital → mechanically LOWER ROE │ │ SHAREHOLDER wants HIGH ROE → prefers LOWER capital │ │ │ │ This tension is permanent and structural. A bank that │ │ reports both a very high ROE and a very high CAR is either │ │ exceptionally profitable or something needs checking. │ └──────────────────────────────────────────────────────────────┘
Limitations.
- Rewards leverage. A more levered bank shows higher ROE for the same ROA. Always decompose.
- Denominator convention (average vs closing) materially changes the result [R].
- Ignores risk. Two banks with identical ROE can have wholly different risk profiles.
- Distorted by OCI. ROE uses profit, but equity includes OCI. NMB's equity was reduced by NPR 210 million of OCI losses that never touched the profit numerator.
- Annualisation at Q1–Q3 is unreliable (see Part 3.16).
Related terms. Part 2.24 Profit for the period · Part 1.34 Total equity attributable · ROA · Part 3.15 Basic EPS
Return on Assets (Annualized)
Simple definition. How much profit the bank makes for each rupee of assets it manages.
Technical definition. Annualised profit attributable to ordinary equity holders expressed as a percentage of average total assets.
Formula
Annualised profit attributable to ordinary equity holders
ROA = ────────────────────────────────────────────────────────────────── × 100
Average Total AssetsWhy ROA is the purer measure. ROA strips out leverage. Two banks with the same ROA have the same operating efficiency; the one with higher ROE is simply more levered. ROA measures skill; ROE measures skill plus gearing.
Example calculation. Illustrative — HCBL.
Profit to ordinary shareholders NPR 3,891,616,000
Average total assets = (397,381,306 + 351,592,507) ÷ 2
= NPR 374,486,907,000
ROA = 3,891,616,000 ÷ 374,486,907,000 × 100 = 1.04%Reproduces the reported Bank ROA of 1.04%.
Reading NMB's numbers.
FY2082/83 FY2081/82 Change Bank 1.04% 0.88% +0.16pp (+18% relative) Group 1.42% 0.89% +0.53pp (+60% relative)
Decomposing ROA for a bank.
Net interest income Non-interest income Impairment Operating expense Tax
ROA = ─────────────────── + ─────────────────── − ────────── − ───────────────── − ───
Average assets Average assets Avg assets Average assets Avg
Bank, using average assets of NPR 374,486,907 thousand:
Net interest income 9,426,240 ÷ 374,486,907 = 2.52%
Non-interest income 3,597,828 ÷ 374,486,907 = 0.96%
Impairment (2,345,056) ÷ 374,486,907 = (0.63%)
Operating expenses (4,561,504) ÷ 374,486,907 = (1.22%)
Non-operating net (369,480) ÷ 374,486,907 = (0.10%)
Tax (1,734,355) ÷ 374,486,907 = (0.46%)
───────
1.07%
(approximates the reported 1.04% before the PNCPS
adjustment and rounding)Interpretation.
HIGHER ROA → the bank earns more from each rupee of assets
→ reflects better margins, better fee income,
lower credit costs or lower operating costs
→ cannot be manufactured by leverage
LOWER ROA → thin margins, high credit costs, or bloated
cost baseLimitations.
- Ignores off-balance-sheet exposure — guarantees, LCs and undrawn commitments generate fee income (numerator) but sit outside total assets (denominator), inflating ROA for trade-finance-heavy banks.
- Ignores risk weighting. A bank full of government securities and one full of unsecured lending may have the same ROA and wholly different risk. Consider return on RWA as a complement.
- Average vs closing assets convention [R].
- Annualisation issues at interim dates.
Related terms. ROE · Part 1.17 Total Assets · Part 2.24
PE Ratio (Annualized)
Simple definition. How many years of current earnings an investor is paying for the share.
Technical definition. Market price per share divided by annualised earnings per share.
Saved in this browser only — there is no account to create. Clearing your browser data clears your progress.