Asset Quality, Capital, Solvency & Efficiency Ratios

First published 23 Aug 2026 · Last verified 29 Aug 2026

A bank with 22 percent return on equity can still be a value trap if its loan book is quietly rotting. A hydropower company with a 9 percent tariff escalation clause can still default on its term loan if its interest coverage has slipped below one. An insurer can report a rising premium book and still be unable to pay a large claim if its solvency margin has eroded. None of these facts show up in the price-earnings ratio. They show up in a separate family of ratios — asset quality, capital, solvency, debt-servicing, and efficiency — that most retail investors in Nepal skip because they look like regulatory paperwork rather than investment analysis. This chapter closes that gap and, with it, closes out Part VII of this book. By the end of Lesson 41.6 you will have every ratio category needed to underwrite a NEPSE-listed company the way a credit officer or an institutional analyst would — not just the way a headline-chasing retail trader does.

Lesson 41.1 — Asset Quality: The NPL Ratio and Provisioning Coverage Ratio

Every rupee a bank or development bank lends is a rupee that might not come back. Asset quality ratios measure how much of the loan book has already gone bad, and how well the bank has cushioned itself against the part that has.

What counts as non-performing Nepal Rastra Bank classifies loans into five categories: Pass, Watchlist, Substandard, Doubtful, and Loss. A loan is "non-performing" once it falls into Substandard, Doubtful, or Loss — broadly, once principal or interest is overdue beyond the regulatory grace period (90 days past due for most categories, with sub-classifications carrying escalating provisioning requirements). The Non-Performing Loan (NPL) ratio is:

NPL Ratio = Non-Performing Loans ÷ Total Loans and Advances

This single number tells you what fraction of the bank's earning assets are no longer reliably earning anything.

The sector picture has moved fast. NRB's Key Financial Indicators as of Asadh-end 2082 (mid-July 2025) put the system-wide NPL ratio at roughly 3.66 percent. By the time NRB's Financial Stability Report covered the third quarter of FY 2025/26, the sector-wide figure had jumped to 5.60 percent — a sharp deterioration within a matter of quarters, driven by a loan quality review that pushed several banks to reclassify restructured and rolled-over loans more conservatively. This is exactly the kind of move an investor tracking only the P/E ratio would miss entirely.

Cross-bank dispersion is the real story, though — the sector average hides enormous variation between individual banks:

Provisioning coverage: the second half of the story A rising NPL ratio is a red flag; whether it is a survivable one depends on the Provisioning Coverage Ratio (PCR):

PCR = Loan Loss Provisions ÷ Gross Non-Performing Loans

NRB requires escalating provisioning by category — a Watchlist loan might require 5 percent provisioning, Substandard 25 percent, Doubtful 50 percent, and Loss 100 percent. A bank whose NPL book is concentrated in Loss-category loans but which has provisioned adequately (PCR near or above 100 percent) has already absorbed the pain in its books; a bank with the same NPL ratio but a PCR of 40–50 percent is still sitting on unrecognized losses that will hit future quarters' profit and, ultimately, its capital.

Worked example: Suppose a mid-sized commercial bank reports gross NPLs of Rs 8.40 billion against total loans of Rs 190 billion (NPL ratio = 4.42 percent) and has booked loan loss provisions of Rs 5.30 billion.

PCR = 5.30 ÷ 8.40 = 63.1 percent

That leaves Rs 3.10 billion of the NPL book unprovisioned. If even half of that eventually needs writing off, the bank absorbs a one-time hit of roughly Rs 1.55 billion — for a bank with, say, Rs 24 billion of core capital, that is a manageable 6.5 percent capital hit, uncomfortable but not solvency-threatening. Run the same arithmetic with a PCR of 35 percent instead of 63 percent and the unprovisioned exposure more than doubles, which is the kind of stress test every investor should run mentally before buying a bank purely on trailing P/E.

Provisioning figures reported in rupee terms alone (not ratio terms) can look reassuring purely because a bank is larger. NIC Asia Bank's reported loan loss provisioning of roughly Rs 22.32 billion sounds enormous in isolation — but scale it against its own NPL base and total loan book before concluding the bank is well-cushioned. Always convert provisioning to a ratio (PCR) before comparing across banks of different sizes.

Watchlist loans: the leading indicator most investors skip Below the NPL line sits a category most quarterly disclosures report but few retail investors read: Watchlist loans — performing loans, technically, but flagged for early-warning signs such as repeated restructuring requests, sectoral stress (tourism, real estate, hospitality), or a single large borrower breaching a loan covenant. A bank can hold a stable or even improving NPL ratio for two or three quarters while its Watchlist book quietly balloons — and Watchlist loans are the pipeline that becomes tomorrow's Substandard and Doubtful loans. When a bank's Watchlist-to-total-loans figure moves from, say, 2 percent to 5 percent in a single reporting period, treat it as an early warning on the NPL ratio one to two quarters ahead of when the number itself moves, not as a footnote to skip past on the way to the profit figure.

Lesson 41.2 — Capital Adequacy Ratio Recap and Its Role in Valuation Screening

Chapter 39 introduced the mechanics of the Capital Adequacy Ratio (CAR) in the context of bank profitability constraints. Here we recap the formula briefly and then focus on the part most retail investors never do: using CAR as a valuation screening filter, not just a compliance checkbox.

CAR = (Tier 1 Capital + Tier 2 Capital) ÷ Risk-Weighted Assets (RWA)

Under NRB's Basel III-aligned Capital Adequacy Framework, Class "A" commercial banks must maintain a minimum Common Equity Tier 1 (CET1) ratio of 4.5 percent, a minimum Tier 1 ratio of 6 percent, a minimum total capital ratio of 8.5 percent, and — critically — an additional Capital Conservation Buffer (CCB) of 2.5 percent, bringing the effective minimum total CAR to 11 percent of risk-weighted assets. Fall below that 11 percent line and NRB's dividend directive kicks in: banks are restricted from distributing cash dividends until the buffer is rebuilt.

As of Asadh-end 2082 (mid-July 2025), the sector-wide average CAR stood near 12.78 percent (core capital adequacy near 10.03 percent) — comfortably above the 11 percent floor, but with real dispersion underneath:

Himalayan Bank's 0.16 percentage-point cushion is the number that should stop a dividend-focused investor cold — a single adverse quarter of credit-loss recognition (which, given Himalayan's NPL ratio of 7.68 percent noted in Lesson 41.1, is a live risk) could push it below the regulatory floor and trigger a dividend freeze regardless of how the P&L otherwise looks.

CAR as a valuation screen, not just a compliance metric Here is the cross-reference this chapter promised: CAR headroom directly caps a bank's sustainable growth rate, and therefore its justified Price-to-Book multiple (covered in Chapter 38's P/BV framework). A bank cannot grow its risk-weighted loan book faster than its capital base allows without either raising fresh equity (diluting existing shareholders) or breaching CAR. The self-sustainable growth ceiling is approximately:

Sustainable Loan Growth ≈ (Retained Earnings ÷ Risk-Weighted Assets) ÷ Minimum CAR Requirement

A bank with high ROE but a CAR buffer of only 0.16 percentage points above the floor cannot compound its book value the way its ROE alone suggests — it will be forced into a rights issue, a slower loan growth path, or a dividend cut. Conversely, Standard Chartered Nepal's 6.82 percentage-point buffer means it can grow its RWA book substantially before needing new capital, which is one reason foreign-parented banks with high CAR often trade at a valuation premium despite modest headline growth — the market is pricing in capital-funded compounding capacity, not just this year's earnings.

CAR also feeds into a bank's cost of funds, and therefore into its own profitability ratios from earlier in this Part. A bank with a comfortable CAR buffer can raise subordinated (Tier 2) debentures more cheaply, because debenture holders and rating agencies read the buffer as a going-concern cushion; a bank hugging the regulatory minimum pays a higher coupon on the same instrument, or may be unable to issue Tier 2 debentures at all until it rebuilds capital through retained earnings or a rights issue. This is one reason CAR and Net Interest Margin move together over a multi-year horizon far more than a single quarter's numbers suggest — a bank forced to slow loan growth to protect CAR also slows the compounding of net interest income that ROE ultimately depends on.

Lesson 41.3 — Solvency Ratios for Insurance Companies

Insurance companies don't carry an NPL ratio or a CAR — they carry a structurally different risk: the possibility that claims exceed the assets set aside to pay them. The regulator here is the Nepal Insurance Authority (formerly the Insurance Board), and the key ratio is the Solvency Ratio:

Solvency Ratio = Available Solvency Margin ÷ Required Solvency Margin

The Required Solvency Margin is actuarially derived from the insurer's outstanding liabilities (technical reserves, unearned premium reserves, outstanding claims) and its risk exposure; the Available Solvency Margin is the excess of admitted assets over liabilities — economically, the insurer's true net worth cushion. The Nepal Insurance Authority sets a minimum solvency ratio of 1.5 for non-life (general) insurers and 1.3 for life insurers — meaning an insurer must hold at least 1.5x (or 1.3x) the actuarially required cushion before it is considered adequately capitalised to meet claims obligations.

A solvency ratio of 1.0 would mean an insurer holds exactly the actuarially required cushion — zero margin for error. A ratio of 2.67 (the recent non-life sector average) means the sector, on average, holds nearly two-and-a-half times the mandated minimum cushion. Read the ratio as a multiple of the regulatory floor, not as a percentage.

The most recent non-life insurance sector disclosure put the industry average solvency ratio at 2.67, up from 2.46 in the prior period — comfortably above the regulatory floor of 1.5, but again with wide dispersion:

On the life insurance side, the Nepal Insurance Authority's most recent disclosure across twelve life insurers showed an industry average solvency ratio of roughly 4.44 against a minimum requirement of 1.3 — an even larger average cushion, reflecting the longer-duration, more conservatively reserved nature of life insurance liabilities.

A merged insurance company's solvency ratio in the first one to two reporting periods after amalgamation should be read with extra skepticism for the same reason. Merging two actuarial reserve bases, two claims-reserving philosophies, and two capital structures can produce solvency ratios that swing by a full point or more in either direction purely from methodology harmonization — not from any real change in the underlying claims-paying capacity. Wait for at least two clean post-merger quarters before trusting the number.

Why this matters for an equity investor, not just a policyholder: an insurer operating close to the 1.5 (or 1.3) floor faces the same dividend and growth constraints a thinly-capitalised bank faces under CAR — the regulator can restrict dividend distribution, new product underwriting, or require a capital call, all of which directly affect the equity holder's return, independent of how strong the reported premium growth or combined ratio looks.

Lesson 41.4 — Debt-to-Equity and Interest Coverage: Hydropower and Manufacturing

Banks and insurers are regulated on capital and solvency because they are financial intermediaries. Hydropower and manufacturing companies are not subject to CAR or solvency rules — but they carry their own version of the same underlying question: can this company service its debt through a full business cycle? That is what the Debt-to-Equity (D/E) ratio and Interest Coverage Ratio (ICR) are built to answer.

D/E = Total Interest-Bearing Debt ÷ Total Shareholders' Equity

ICR = Earnings Before Interest and Tax (EBIT) ÷ Interest Expense

Hydropower: a sector built on debt by design Nepali hydropower projects are, by construction, among the most leveraged listed businesses on NEPSE. Project financing for run-of-river IPPs is conventionally structured at 70:30 or even 80:20 debt-to-equity during the construction phase — developers bring in the minimum promoter and public equity required to satisfy the Department of Electricity Development and lender covenants, then fund the bulk of construction cost through syndicated term loans from commercial banks and development banks. This is not a red flag by itself — it is how the entire sector is built, and it is exactly why interest coverage, not just D/E, is the ratio that separates a resilient hydropower stock from a fragile one.

Worked example (illustrative): Consider "Kalpataru Hydropower Ltd." (a hypothetical mid-sized run-of-river plant used here purely for illustration), a 24 MW project with total project cost of Rs 4.80 billion, financed 75:25 debt-to-equity:

• Debt: Rs 3.60 billion at an average lending rate of 10.5 percent → annual interest expense ≈ Rs 378 million
• Equity: Rs 1.20 billion
• D/E = 3.60 ÷ 1.20 = 3.0x

Once commissioned and generating at a plant load factor of roughly 45 percent (typical for run-of-river design energy), annual energy sales at the PPA tariff might generate EBIT of roughly Rs 520 million.

ICR = 520 ÷ 378 = 1.38x

An ICR of 1.38x is thin — a single dry-season shortfall in river flow, an unplanned outage, or a PPA payment delay from the offtaker can push coverage below 1.0x, at which point the company is servicing debt out of reserves rather than operations. Compare this to the same project ten years into its 25-30 year loan amortisation schedule, where a declining principal balance and (if the loan is fixed-rate) unchanged tariff escalation typically push ICR up toward 2.5–3.5x as the debt burden shrinks relative to a stable or growing revenue base. This is precisely why mature, largely-delevered hydropower names on NEPSE (companies well past their loan amortisation midpoint) trade at different multiples than newly-commissioned or still-under-construction IPPs carrying peak leverage — the market is pricing the interest coverage trajectory, not just current-year EPS.

Manufacturing: lower leverage, but still watch the cycle Nepali listed manufacturers — consumer-facing names like Bottlers Nepal, Unilever Nepal, and Nepal Lube Oil among them — typically run structurally lower D/E than hydropower or even banks, because their capex cycles are shorter, working capital (not fixed asset construction) drives most of their borrowing needs, and many are majority-owned by multinational parents that prefer equity-funded expansion. A D/E of 0.3x–0.8x is common in this group, versus 2x–4x for a hydropower company mid-construction or a bank (which is leveraged by design as a deposit-taking institution). Capital-intensive manufacturers — cement, steel rerolling, and cable manufacturing — sit in between, often in the 1.0x–2.0x range, reflecting genuine plant and machinery financing.

When comparing D/E across sectors, never benchmark a hydropower company against a manufacturing company directly — benchmark it against its own PPA-implied revenue schedule and loan amortisation curve instead. A hydropower stock's "correct" D/E in year 3 of a 25-year loan is supposed to look nothing like its correct D/E in year 20. The ratio only means something in the context of where the project sits on its financing timeline.

The same discipline applies to manufacturers going through a capacity expansion cycle. A cement or cable manufacturer financing a new production line will temporarily push its D/E from, say, 1.0x toward 2.0x for the two to three years construction and ramp-up take, while interest coverage compresses as new debt service begins before the expanded capacity is fully utilised. This is not automatically a red flag — a manufacturer expanding into genuine demand growth, with a credible utilisation ramp-up plan, is doing exactly what a hydropower developer does at a smaller scale. The distinguishing question is the same one that matters for hydropower: does interest coverage recover toward its pre-expansion level within the timeframe management has guided to, or does a second or third round of debt-funded expansion arrive before the first has been digested? A manufacturer that layers expansion on expansion without interest coverage ever recovering above 2x is exhibiting the same warning sign as a hydropower company whose PPA revenue never catches up to its debt service schedule.

Lesson 41.5 — Efficiency Ratios: Cost-to-Income (Banks) and Asset Turnover (Manufacturers)

Asset quality, capital, and solvency ratios tell you whether a company can survive stress. Efficiency ratios tell you how well it converts its resources — staff, branches, or fixed assets — into revenue during normal times. They are the ratios that separate a well-run company from a merely solvent one.

Cost-to-Income Ratio (banks and financial institutions) Cost-to-Income Ratio = Total Operating Expenses ÷ Total Operating Income

A lower ratio means the bank spends less to generate each rupee of income — greater operating leverage, and typically a sign of disciplined branch expansion, digital channel adoption, and staff productivity.

The mechanics matter for a dividend-focused investor: a bank with a high cost-to-income ratio has less operating profit cushion to absorb a bad-loan-provisioning shock in a downturn — it is simultaneously less efficient in good years and less resilient in bad ones. When you screen banks on ROE alone, you can be fooled by a bank that generates decent ROE today through aggressive loan growth while carrying an inefficient cost base that will compress margins the moment loan growth slows.

Asset Turnover Ratio (manufacturers and trading companies) Asset Turnover Ratio = Net Sales ÷ Total Assets

This measures how many rupees of revenue a company generates per rupee of assets deployed — the manufacturing-sector analogue to a bank's cost-to-income ratio, and directly relevant to the DuPont ROE decomposition introduced earlier in Part VII (ROE = Net Margin × Asset Turnover × Equity Multiplier).

Worked example: Consider a hypothetical FMCG bottling manufacturer with net sales of Rs 6.2 billion and total assets of Rs 3.1 billion.

Asset Turnover = 6.2 ÷ 3.1 = 2.0x

Compare this to a hypothetical cement manufacturer with net sales of Rs 8.0 billion but total assets of Rs 16.0 billion (heavy kiln, quarry, and plant investment):

Asset Turnover = 8.0 ÷ 16.0 = 0.5x

Neither ratio is "better" in isolation — the FMCG bottler's asset-light model naturally produces a higher turnover figure than a capital-intensive cement plant, exactly as the D/E table in Lesson 41.4 predicted for the same two archetypes. What matters for stock selection is whether a given company's asset turnover is improving or deteriorating relative to its own history and its direct sector peers, and whether declining turnover is being offset by rising net margin (pricing power) or simply compounding into a falling ROE.

Efficiency ratios and solvency/capital ratios answer different questions and must never substitute for one another. A bank can have an excellent cost-to-income ratio and a dangerously thin CAR buffer at the same time — operational efficiency does not protect against a capital shortfall. Always read efficiency ratios alongside, never instead of, the asset-quality and capital ratios from Lessons 41.1 and 41.2.

Lesson 41.6 — Building a Cross-Sector Ratio Scorecard

You now have every ratio family this book covers: valuation multiples (Part VI), profitability and per-share metrics (earlier in Part VII), and — from this chapter — asset quality, capital, solvency, leverage, coverage, and efficiency ratios. The final skill is operational: turning this into a spreadsheet you actually maintain every quarter, across every sector you hold.

Structure the scorecard as one row per company and organise columns into three blocks, so the same sheet works whether the row is a bank, an insurer, a hydropower company, or a manufacturer:

Block 1 — Universal columns (every company): Sector, Latest Quarter, Market Price, EPS (TTM), P/E, P/BV, ROE.

Block 2 — Sector-specific stability columns: for banks, NPL Ratio, PCR, CAR, CAR headroom above 11%; for insurers, Solvency Ratio and headroom above 1.5 / 1.3; for hydropower and manufacturers, D/E and Interest Coverage Ratio.

Block 3 — Sector-specific efficiency columns: Cost-to-Income for banks, Asset Turnover for manufacturers, Combined Ratio for insurers.

A simplified snapshot of what four rows might look like at a single point in time, mixing sectors on one sheet:

Notice that this single table already reproduces the real dispersion you saw in Lessons 41.1 through 41.4 — Everest Bank's 0.61 percent NPL ratio, Himalayan Bank's 0.16 percentage-point CAR cushion, Oriental Insurance's 1.60 solvency ratio sitting just above its 1.5 floor, and a hydropower company's characteristically thin post-commissioning interest coverage. A scorecard built this way turns four completely different regulatory regimes into one comparable risk view, which is the entire point of building it yourself rather than relying on a broker's one-size-fits-all screener.

Weighting the flags into a single score A scorecard with four or five flag columns per company is useful for a single glance, but for a portfolio of fifteen or twenty holdings across multiple sectors, convert the flags into a simple weighted score so positions can be ranked rather than merely color-coded. A workable weighting for a bank might allocate 35 percent of the composite score to CAR headroom, 35 percent to the NPL ratio (inverted, so lower NPL scores higher), 15 percent to PCR, and 15 percent to cost-to-income — reflecting that capital and asset quality matter roughly twice as much to survivability as operating efficiency does. For an insurer, weight solvency-ratio headroom above the regulatory floor most heavily; for a hydropower or manufacturing name, weight interest coverage above D/E itself, since coverage is the metric that determines whether the debt load is actually serviceable this year rather than merely how large it is on paper. None of these weights need to be precise to be useful — the exercise of assigning them forces you to decide, once, which risk each sector poses most acutely, rather than re-litigating it every time a new quarterly result lands in your inbox.

Where to source the raw inputs each quarter: NRB's "Key Financial Indicators of Commercial Banks" and quarterly Financial Stability Report for bank NPL, PCR, and CAR figures; the Nepal Insurance Authority's quarterly solvency disclosures for insurers; individual company quarterly financial reports filed on NEPSE's corporate disclosure portal and each company's own investor-relations page for D/E, interest coverage, cost-to-income, and asset turnover. None of this requires a paid data terminal — every figure in this chapter's tables came from public regulatory and company disclosures freely available online.

A scorecard, finally, is only as current as its last update. NPL ratios, CAR buffers, and solvency ratios can move materially within a single quarter — as the jump from a 3.66 percent to a 5.60 percent sector NPL ratio within a few quarters in this chapter demonstrates. Set a standing calendar reminder for each quarter-end reporting window (mid-Ashwin, mid-Poush, mid-Chaitra, and mid-Ashadh) and refresh the scorecard within two weeks of each, rather than trusting a stale figure from two quarters ago.