A bank, an insurer, a hospital, a builder - each writes its accounts in a different language. Find a sector's five equivalents and you can read any annual report on the exchange.
The statements everyone first learns belong to a manufacturer: revenue, gross margin, inventory, a bit of debt. Then you open a bank and none of it fits - there is no gross margin, deposits are the raw material, and borrowing ten times your equity is the healthy design, not a red flag. Every sector below rewrites the accounts its own way. For each one, this page names the five things to look for - what replaces revenue, where the real margin lives, what the capital actually is, the metric that leads the financials, and what is simply absent - plus the numbers worth knowing and where the truth tends to hide. Search a metric (ARPOB, combined ratio, ALM, ARPU) to jump to its sector. A way of reading, never a stock tip.
The five equivalents
For any unfamiliar sector, find these five. Get them right and you can open any annual report on the exchange.
1The top line - what replaces revenue
2The real margin - where the economics live
3Capital consumed - what sits where inventory would
4The operating metric that leads the financials
5What's absent - what you'd normally lean on but can't
Interest earned on loans and investments; net interest income (the spread) is the real top line - no revenue-from-operations or gross margin exists
The real margin - where the economics live
Net interest margin - the spread between interest earned and interest paid, expressed against earning assets; pre-provision operating profit is the cleanest measure of what the business earned
Capital consumed - what sits where inventory would
Deposits are the raw material (cheap, sticky CASA is the good kind), advances are the earning asset; net worth is a thin sliver by design because ~10x leverage is normal - debt-to-equity is meaningless
The operating metric that leads the financials
Net interest income / net interest margin, then pre-provision operating profit before the provisioning judgement lands
What's absent - what you'd normally lean on but can't
Gross/operating margin and debt-to-equity - a big deposit liability is good not bad, and loans are an asset not debt owed
Metrics to know
CASA ratio
gross NPA
provision coverage ratio
net NPA
capital adequacy ratio
net interest income (NII)
net interest margin (NIM)
pre-provision operating profit (PPOP)
credit cost
cost-to-income
Where the truth hides · Provisions are the discretionary swing line and a low gross NPA can hide thin coverage, delayed recognition (evergreening/restructuring) and concentration - read PPOP and credit cost, not reported profit.
Interest income on loans, keeping a wider spread than a bank because it lends to riskier borrowers; no deposit franchise, funded by borrowings
The real margin - where the economics live
Net interest spread over cost of funds - wider than a bank's but bundled with higher credit and funding risk; return on assets, not gross margin
Capital consumed - what sits where inventory would
No deposits - funded by wholesale borrowings and short commercial paper that must be continuously refinanced; borrows short and lends long, so the near-term maturity gap is negative by design; held to a higher capital ratio than a bank
The operating metric that leads the financials
The asset-liability (ALM) maturity table - the near-term cumulative funding gap - read before the profit; cost of funds beside it
What's absent - what you'd normally lean on but can't
Cheap sticky CASA and the comfort that capital/profit signal safety - a well-capitalised, profitable NBFC with good loans can still fail on liquidity
Metrics to know
asset-liability (ALM) maturity gap
cumulative near-term gap
cost of funds
liquidity buffer / liquidity coverage ratio
gross stage-3 assets
ECL / stage-3 coverage
net interest income
spread
return on assets
capital adequacy
Where the truth hides · Solvency is not liquidity - the danger hides in a funding freeze plus optimistic behavioural-maturity assumptions and off-balance-sheet securitisation/co-lending commitments; access to funding matters more than its price.
Interest on tiny, unsecured loans to low-income borrowers - a high yield in good years on a large book, a thin margin
The real margin - where the economics live
A thin margin on a large book that high credit cost can wipe out in a stress year; profit is high in calm years and crashes in a shock
Capital consumed - what sits where inventory would
Unsecured lending via joint-liability groups (peer guarantee substitutes for collateral); no collateral base - the book's fragility is set by geographic concentration and growth quality (ticket size vs an already-indebted base)
The operating metric that leads the financials
Collection efficiency - the share of instalments actually collected - the leading indicator, read every quarter before profit
What's absent - what you'd normally lean on but can't
A timely, trustworthy reported profit and current NPA - both lag by a quarter or two, so they reassure exactly when the leading indicator is already flashing
Metrics to know
collection efficiency
portfolio-at-risk (PAR-30/60/90)
credit cost
geographic/state concentration
joint-liability-group structure
growth quality (ticket size vs borrower income)
restructured/refinanced book
gross NPA (lagging)
Where the truth hides · The collection number can be flattered by restructuring/evergreening/advance instalments, cannot forecast the exogenous shock, and misses political/regulatory risk (waivers, rate caps) - a suspiciously stable collection figure through a stressed period is itself the warning.
Value of new business (VNB) - the present value of future profit from this year's policies - replaces a revenue/profit top line; premiums sit in the policyholders' Revenue Account
The real margin - where the economics live
VNB margin and the operating return on embedded value - the real return on the book; reported profit after tax is a lagging, near-useless number because profit emerges over a policy's whole life
Capital consumed - what sits where inventory would
Two statements: a large policyholders' Revenue Account (premiums, investment income, reserves) where the business happens, and a small shareholders' P&L fed by an actuarially-determined surplus; new-business strain means fast growth suppresses reported profit
The operating metric that leads the financials
Value of new business (VNB) and its margin, embedded value and its operating return - not reported profit or P/E
What's absent - what you'd normally lean on but can't
A meaningful reported profit and the price-to-earnings multiple - ranking insurers on P/E rewards the slow grower and punishes the fast grower (a sign reversal)
Metrics to know
value of new business (VNB)
VNB margin
embedded value
operating return on embedded value (RoEV)
persistency (13/25/37/49/61 month)
new business premium
product mix (protection share)
price-to-embedded-value
price-to-VNB
new-business strain
Where the truth hides · Embedded value and VNB are actuarial projections only as honest as their assumptions - a rising VNB can be genuine mix shift or quietly optimistic re-assumption; check persistency and sensitivity tables, and how much came from the business versus the model.
Net earned premium, read through the combined ratio (claims + expenses over premium), plus investment income on the float - two engines, not one revenue line
The real margin - where the economics live
Underwriting result (100 minus the combined ratio) PLUS investment income on the float - the insurer can run an underwriting loss and still be highly profitable overall
Capital consumed - what sits where inventory would
The float - premiums collected today and held against claims paid months or years later - is a large, cost-free pool invested for the insurer's own account, often several times annual premium; that pool, not the underwriting margin, is where the compounding lives
The operating metric that leads the financials
The combined ratio (100 the dividing line), read against what the float earns
What's absent - what you'd normally lean on but can't
The idea that the bottom line or combined ratio alone tells the story - a combined ratio above 100 is an underwriting loss, not a losing business
Metrics to know
combined ratio
claims (loss) ratio
expense ratio
insurance float
float size relative to premium
float yield / investment income
reserve adequacy
reserve-development triangle
IBNR reserves
catastrophe exposure / reinsurance
Where the truth hides · The combined ratio is only as honest as the reserves behind it - under-reserving flatters the claims ratio now and forces a top-up later (only the reserve-development triangle certifies it); also decompose float income into durable interest versus one-off equity gains, and watch un-modelled catastrophe/tail risk.
Fee revenue = volume (AUM or transactions) times yield (the fee rate) - the AUM headline overstates the business because the two variables move in opposite directions
The real margin - where the economics live
Operating margin on a fee stream; revenue growth (pool times a compressing yield) is the signal, AUM growth is noise
Capital consumed - what sits where inventory would
Almost no capital employed - no plant, inventory or loan book, just people and software; return ratios are inflated by a near-zero denominator and can't be compared with a capital-heavy business; the real virtue is high cash generation and payout
The operating metric that leads the financials
Revenue decomposed into volume (AUM) times yield on assets (fee rate)
What's absent - what you'd normally lean on but can't
Meaningful return on equity (a 30-40% ROE is arithmetic of a tiny denominator, not quality) and the AUM headline as a growth rate - the two numbers the business leads with are the two most misleading
Metrics to know
assets under management (AUM)
yield on assets (fee rate, bps)
revenue growth
product mix (equity vs liquid/debt yield)
flow quality (net inflows vs market appreciation)
operating margin
return on equity (distrust)
cash generation / payout
equity-only yield
Where the truth hides · The numbers show yield compression but not why - benign mix shift toward low-fee products versus genuine fee erosion as clients flee look identical; the real risk is operational/reputational (mis-selling, outage, fund blow-up, key-manager departure), not the pristine capital-light balance sheet.
Same-store growth (occupancy and ARPOB at existing hospitals) plus new-bed additions, read apart
The real margin - where the economics live
Mature-unit EBITDA margin (~25-26%), not the expansion-depressed blend
Capital consumed - what sits where inventory would
Beds and buildings; a new unit's heavy fixed costs sit ahead of the patients, losing money for two to four years by design
The operating metric that leads the financials
Occupancy (share of beds filled) and ARPOB (average revenue per occupied bed per day)
What's absent - what you'd normally lean on but can't
The blended margin as a health check - it is near-useless while the chain is opening units
Metrics to know
occupancy
ARPOB
mature-unit margin
blended EBITDA margin
installed/operational beds
J-curve position of new units
same-store vs new-bed growth
case-mix
Where the truth hides · In the mature-versus-ramping split - a falling blended margin during expansion is usually the J-curve, not decline, and only occupancy and ARPOB confirm a ramp is real.
Net distributable cash flow (NDCF) - the cash available to unitholders, not reported profit
The real margin - where the economics live
Distribution coverage (NDCF divided by the distribution) and the distribution yield the cash supports
Capital consumed - what sits where inventory would
A portfolio of value-holding property/infra carrying a large non-cash depreciation charge; leverage measured as loan-to-value against valuations that can fall
The operating metric that leads the financials
NDCF and distribution coverage, underpinned by occupancy, WALE and LTV
What's absent - what you'd normally lean on but can't
Reported profit and the P/E - depreciation on property that holds its value makes earnings an artefact and the multiple uninformative
Metrics to know
net distributable cash flow (NDCF)
distribution coverage
distribution yield
occupancy
WALE (weighted-average lease expiry)
loan-to-value (LTV)
maintenance capex
re-leasing spread
Where the truth hides · In cash coverage and the portfolio beneath it - a yield is only as safe as its coverage, the honesty of maintenance capex, and the leverage against valuations that can be reset in a downturn.
Rate base (approved regulated assets) and the allowed return set by the regulator, not market revenue
The real margin - where the economics live
Profit is roughly regulated equity times the allowed return percentage - a formula, not a competitive margin
Capital consumed - what sits where inventory would
Grows only by growing the rate base through approved capex; each approved asset earns the allowed return going forward
The operating metric that leads the financials
Rate base and allowed return (profit ≈ regulated equity × allowed return)
What's absent - what you'd normally lean on but can't
Market share and pricing power - there is no competition; steady profit signals rate-base growth, not business quality
Metrics to know
rate base
allowed return %
regulated equity
regulatory assets
discom receivable days
approved capex
tariff-order status
one-off true-ups vs rate-base growth
Where the truth hides · In two balance-sheet lines under the bond-like profit - regulatory assets (costs booked on the promise of a future tariff order) and discom receivables (cash stuck with weak state buyers); and in the regulator's discretion, which the accounts cannot show.
Passenger revenue read per available seat-kilometre (RASK), earned across an enormous volume of seat-kilometres flown
The real margin - where the economics live
The razor-thin RASK-minus-CASK spread per seat-kilometre, multiplied across huge volume
Capital consumed - what sits where inventory would
Aircraft are largely leased - under Ind AS 116 they sit as right-of-use assets and lease liabilities that dominate the balance sheet
The operating metric that leads the financials
RASK, CASK and load factor (with CASK-ex-fuel for durable efficiency)
What's absent - what you'd normally lean on but can't
A meaningful reported net-debt / debt-to-equity figure - it excludes the aircraft leases, so it understates real leverage massively
Metrics to know
ASK
RASK
CASK
CASK-ex-fuel
load factor
yield
lease-inclusive net debt
fuel cost share
Where the truth hides · In lease-inclusive leverage and the RASK-minus-CASK spread - a full plane still loses money when a fuel spike pushes CASK above RASK.
Service revenue driven by ARPU (average revenue per user) across the subscriber base
The real margin - where the economics live
A fat EBITDA margin but a razor-thin profit after tax, once network depreciation, spectrum amortisation and interest are taken out
Capital consumed - what sits where inventory would
A largely fixed network (towers, fibre, spectrum, equipment) funded by debt, plus deferred spectrum obligations and tower/fibre leases
The operating metric that leads the financials
ARPU (the profit lever) and churn (the sustainability check)
What's absent - what you'd normally lean on but can't
A net margin and net-debt figure that mean what they say - the thin margin is a leveraged residual, and reported debt excludes spectrum and leases
Metrics to know
ARPU
churn
operating leverage
subscribers
spectrum obligations
lease liabilities
net-debt-incl-spectrum/EBITDA
EBITDA margin
Where the truth hides · In ARPU and churn, and in net debt including spectrum obligations and leases - the hidden debt that decides who survives a price war.
Pre-sales (flats booked this year) and collections (cash received), which lead revenue by years
The real margin - where the economics live
Realisation per square foot on pre-sales; reported margins are lumpy artefacts of completion timing
Capital consumed - what sits where inventory would
Land held via joint-development agreements (revenue-share, not owned) and project debt buried in per-project SPVs
The operating metric that leads the financials
Pre-sales and collections (with area sold and realisation per sq ft)
What's absent - what you'd normally lean on but can't
Reported revenue as a measure of the year - recognised only on completion, it can spike in a weak selling year and slump in a strong one
Metrics to know
pre-sales / bookings
collections
area sold
realisation per sq ft
cancellation rate
unsold inventory
SPV-level and consolidated net debt
JDA land pipeline
Where the truth hides · In pre-sales and collections for the business, and one layer down in SPV/JDA notes for the true leverage - the parent's standalone net debt understates the group.
Order book (contracted work not yet executed), read against revenue as book-to-bill
The real margin - where the economics live
Execution margin on the backlog, only as real as billing and cash confirm - the completion estimate is a lever
Capital consumed - what sits where inventory would
Working capital stretched by retention money withheld to completion; revenue recognised over time on percentage-of-completion
The operating metric that leads the financials
Order book and book-to-bill (order inflow vs revenue), plus unbilled-versus-billed revenue
What's absent - what you'd normally lean on but can't
Receivable days as an ordinary alarm - ~140 days is structural for a contractor because of retention, not a collection failure
Metrics to know
order book
book-to-bill
order inflow
unbilled revenue vs billed
retention money
receivable days (ex-retention)
percentage-of-completion estimate
claims under arbitration
Where the truth hides · In the order-book trend for durability and in unbilled-versus-billed for honesty - a widening unbilled gap is the percentage-of-completion lever, and the cash, slowed by retention, is the slow truth that catches up.
Segment revenues from three businesses - upstream oil/gas sales, refining product sales, marketing fuel sales - not one blended 'revenue from operations'
The real margin - where the economics live
Gross refining margin (product value minus crude cost) for refining; upstream realisation against a fairly fixed cost per barrel
Capital consumed - what sits where inventory would
A finite oil/gas reserve being depleted upstream, plus refinery-complexity assets; large net debt that must be read against cyclical profit
The operating metric that leads the financials
Gross refining margin (GRM) per barrel, alongside the crude price
What's absent - what you'd normally lean on but can't
A single blended profit that reads cleanly - the segments move in opposite directions with crude, so the group number hides more than it shows
Metrics to know
GRM per barrel
crude price
upstream EBIT
refining EBIT
marketing EBIT
under-recoveries
reserve-replacement ratio
refinery complexity
Where the truth hides · In the segment split, the crude-and-GRM cycle, and whether the under-recovery subsidy receivable is actually cash.
Commodity sales at a world-set price - the miner is a price-taker, so revenue swings with the commodity cycle
The real margin - where the economics live
Through-cycle EBITDA margin, and cost per tonne against the industry cost curve
Capital consumed - what sits where inventory would
A finite, depleting ore reserve; the miner's value is the profit stream over the life of that reserve
The operating metric that leads the financials
Reserve life (reserves ÷ annual production) and cost per tonne
What's absent - what you'd normally lean on but can't
Renewable capacity - there is no factory to rebuild; the reserve depletes and a single peak-year profit is a poor guide
Metrics to know
reserve life
stripping ratio
cost per tonne
cost-curve position
through-cycle margin
reserve-replacement ratio
commodity price
depletion
Where the truth hides · In reserve life and whether it is being replaced, the rising stripping cost, and the cost-curve position that decides trough survival.
Segment revenues - a cyclical sugar core plus a steadier ethanol-and-power by-product annuity
The real margin - where the economics live
Sugar-segment margin (a government-fixed cane cost against a floating sugar price) versus the steadier, policy-supported ethanol/power margins
Capital consumed - what sits where inventory would
Working capital choked by large subsidy receivables the government owes but has not yet paid
The operating metric that leads the financials
Segment EBIT split (sugar vs ethanol/power) and the ageing of the subsidy receivable
What's absent - what you'd normally lean on but can't
A stable single-year profit - a fixed cost against a floating price lurches feast-to-famine, and part of the reported profit is uncollected subsidy
Metrics to know
cane price (FRP/SAP)
sugar price
sugar-segment EBIT
ethanol EBIT
power EBIT
subsidy receivable
receivable ageing
through-cycle margin
Where the truth hides · In the segment split (cyclical sugar vs durable annuity) and the ageing of subsidy receivables the government has not paid in cash.
RevPAR - revenue per available room - the same-store core metric, split into occupancy and average room rate
The real margin - where the economics live
EBITDA margin amplified by operating leverage; it swings far more than revenue in both directions
Capital consumed - what sits where inventory would
Asset-heavy and largely fixed cost base (property, staff, upkeep); often leveraged, sometimes through leases
The operating metric that leads the financials
RevPAR (occupancy times average room rate)
What's absent - what you'd normally lean on but can't
A cushioning variable cost base - costs do not fall when guests stop coming, so margin stability is absent
Metrics to know
RevPAR
Occupancy
Average room rate (ADR)
EBITDA margin
Operating leverage / incremental margin
Revenue
Where the truth hides · In the occupancy-versus-rate mix a single RevPAR conceals (real demand vs discount-bought volume), and in a single peak or trough year read out of cycle; the balance sheet decides survival through a downturn.
Revenue = shipments × revenue per shipment; read the split, because the headline top line hides which is driving growth
The real margin - where the economics live
Margin only means something beside asset turnover - a thin margin on high turnover (asset-light) versus a fatter margin on low turnover (asset-heavy)
Capital consumed - what sits where inventory would
A tiny, low-debt balance sheet coordinating third-party carriers (asset-light) versus a large debt-funded fleet and warehouses (asset-heavy)
The operating metric that leads the financials
Return on capital (operating margin × asset turnover) and revenue per shipment
What's absent - what you'd normally lean on but can't
A margin that can be judged on its own - a thin margin is strength in asset-light and would-be weakness in asset-heavy, so the margin alone tells you nothing
Metrics to know
asset turnover
operating margin
return on capital (ROCE)
revenue per shipment
shipment volume
net-debt-to-equity
utilisation
Where the truth hides · In the model (asset-light vs asset-heavy), the margin read against asset turnover and capital, and revenue-per-shipment for the quality of growth.
GMV is the headline but is scale not revenue; real revenue is GMV times the take-rate (the slice the platform keeps)
The real margin - where the economics live
Contribution margin - revenue minus the variable cost of serving one order; its sign decides whether the model can work
Capital consumed - what sits where inventory would
Capital is investor cash funding a deliberate loss; survival is read as cash burn against runway, not assets
The operating metric that leads the financials
Contribution margin per order (and take-rate); burn versus runway for survival
What's absent - what you'd normally lean on but can't
Reported profit as a verdict - the loss is a deliberate choice, so earnings and profit margin are meaningless here
Metrics to know
GMV
Take-rate
Revenue
Contribution margin
EBITDA
Cash burn
Runway
Organic vs discount-driven growth
Where the truth hides · In the contribution-margin definition (what is classed variable vs fixed) and in whether GMV growth is organic or discount-bought; the two headline numbers, GMV and the loss, are the least useful.
Billings (cash collected upfront) lead; recognised revenue lags because fees earn out over the course
The real margin - where the economics live
Margin depends on cohort economics - lifetime value against acquisition cost - not on a single reported margin line
Capital consumed - what sits where inventory would
Deferred revenue is a healthy, cash-backed liability (future revenue banked); cash arrives before revenue is earned
The operating metric that leads the financials
Cohort retention and lifetime-value-to-acquisition-cost
What's absent - what you'd normally lean on but can't
A receivable/cash-lags-revenue pattern - the normal instinct that a growing liability is a debt is absent/inverted here
Metrics to know
Deferred revenue
Billings
Recognised revenue
Cohort retention
Lifetime value / acquisition cost
Active students
Revenue per student
Where the truth hides · In blended cohort metrics that mask poor retention in newer, discounted cohorts, and in the lagging recognised-revenue line; enrolment growth can be a leaky bucket of churning, unprofitable customers.
Segment mix - eroding US generics vs fat, stable branded home-market vs rest-of-world - not the blended revenue
The real margin - where the economics live
Segment margins read apart (US generics thin and eroding, branded high and durable); the blend flatters the weak part
Capital consumed - what sits where inventory would
R&D is expensed as incurred, so the pipeline - the real asset - sits off the balance sheet and inflates ROCE
The operating metric that leads the financials
USFDA pipeline: pending ANDAs (future launches) and plant inspection status (the binary risk)
What's absent - what you'd normally lean on but can't
A meaningful ROCE and a coherent single margin - the return ratio is an accounting artefact and the blend fuses opposite businesses
Metrics to know
segment mix (US generics / branded / RoW)
segment margins
R&D expensed
ROCE (flattered)
US price erosion rate
launch cadence
pending ANDAs
plant inspection status / warning letters
Where the truth hides · Outside the financial statements - the most valuable asset (the R&D pipeline) and the defining risk (a plant-level USFDA import alert) both surface in the accounts only after the fact.
Revenue billed in foreign currency, read as constant-currency growth (currency stripped out), not reported growth
The real margin - where the economics live
EBIT margin, driven by utilisation (share of billable people actually deployed) as the direct margin lever
Capital consumed - what sits where inventory would
Almost no capital employed; the real asset - people - sits off the balance sheet, which holds only cash and receivables
The operating metric that leads the financials
Utilisation and attrition - the people-metrics that lead the financials
What's absent - what you'd normally lean on but can't
A meaningful balance sheet and physical assets - inventory, plant and financial leverage are essentially absent
Metrics to know
Constant-currency growth
Utilisation
Attrition
TCV (large-deal)
Revenue per employee
Client concentration
ARR / net revenue retention
EBIT margin
Where the truth hides · In TCV that is not comparable across firms because each defines it differently, and in currency-flattered headline growth; the real risks (attrition, concentration, deal pricing, technology shifts) are all off-statement.
Content revenue, but the real story is the capitalised content library - the largest asset - and what it earns back
The real margin - where the economics live
Reported profit resting on the content-amortisation charge (the largest expense), whose assumed life is a profit lever
Capital consumed - what sits where inventory would
Content is capitalised on the balance sheet and amortised over a judged life; a flop must be written down as an impairment
The operating metric that leads the financials
Per-title economics - what each title costs and earns - since content is hit-driven
What's absent - what you'd normally lean on but can't
A predictable, factory-like asset - content can stop earning suddenly, so a stable depreciable-asset base is absent
Metrics to know
Content library value
Amortisation charge
Amortisation life
Impairments
Per-title economics
Content spend
Library ageing
Where the truth hides · In a stretched amortisation life (library grows while the charge falls) and an un-impaired, ageing library carried above what it can earn - deferred write-downs dressed as an asset.
Net revenue (the markup kept), not gross revenue - most of the headline is pass-through worker salaries
The real margin - where the economics live
Margin on the net markup (around 37%), not the tiny margin on gross revenue (under 3%) which is thin by construction
Capital consumed - what sits where inventory would
Low-capital trade; no meaningful asset base - the business is sourcing and placing workers
The operating metric that leads the financials
Net markup and its margin, driven by the placement mix (specialised/permanent vs commodity temp)
What's absent - what you'd normally lean on but can't
A revenue figure that means what it says - the gross top line is mostly other people's wages passing through
Metrics to know
Gross revenue
Pass-through salary
Net revenue (markup)
Margin-on-markup
Placement mix
EBIT margin on gross
Where the truth hides · In the gap between gross and net - gross growth can swell on low-value commodity volume while the net markup stays flat and its margin slides; the placement mix that decides durability is often only half-disclosed.
Sum-of-the-parts value of the stakes owned (listed at market, unlisted estimated), not the standalone top line
The real margin - where the economics live
No meaningful margin - the parent's own P&L is mostly upstreamed dividends and tells you almost nothing
Capital consumed - what sits where inventory would
Capital is the portfolio of stakes; value is stakes less parent debt, owned at one remove so it cannot be freely realised
The operating metric that leads the financials
The discount to sum-of-the-parts, read against whether a catalyst will close it
What's absent - what you'd normally lean on but can't
Operating earnings you can value the company on - the blended/standalone accounts are near-meaningless
Metrics to know
Sum-of-the-parts (NAV)
Holding-company discount
Listed stake value
Unlisted stake value
Net cash/debt
Catalyst (demerger/buyback/monetisation)
Where the truth hides · In the discount's cause and direction (capital allocation, governance, promoter intent) and in soft unlisted-stake values; a wide discount closes only on a catalyst, and without one is a value trap, not a bargain.
Common questions
Why don't normal financial ratios work for banks and insurers?
Because their statements are built differently. A bank has no gross margin and deposits are its raw material, so borrowing about ten times equity is the healthy design, not a red flag. An insurer's real engine is the float it invests. Reading either with a manufacturer's ratios gives the wrong answer.
How do I read a bank's financial statements?
Start with net interest income and net interest margin (the spread between interest earned and paid), then pre-provision operating profit, then provisioning and the gross and net NPAs. CASA ratio and capital adequacy matter; ordinary gross margin and debt-to-equity do not apply.
What do ARPOB, combined ratio and ARPU mean?
They are the metrics each sector leads with. ARPOB is average revenue per occupied bed (hospitals). A combined ratio below 100 means an underwriting profit (general insurance). ARPU is average revenue per user (telecom). The anatomy above lists the headline metric for every sector.