A bank is not an FMCG

By the end of this lesson

Why generic ratios break for financials, and what to use instead
The two or three metrics that actually matter in each major Indian sector
How to judge a company against the right benchmark rather than a universal one

Picture this

Judging a doctor by a chef's criteria

You'd never rate a surgeon on how fast he chops onions. Yet applying debt-to-equity to a bank is exactly that mistake: a bank at 9× leverage isn't reckless, it's functioning. Deposits are its raw material.

Every sector has a handful of numbers that reveal how it really works — and the generic ratios you've learned are, in some sectors, actively misleading.

The one thing

Judge a company against its own industry's economics and its own history. A 5% net margin is dreadful for software and outstanding for a grocery chain.

Banks and NBFCs

The biggest special case. A bank borrows (deposits) and lends. Its balance sheet isn't a support structure for the business — it is the business. So debt ratios, EV-based multiples and capital-employed returns all lose their meaning.

What to look at instead

NIM — net interest margin
What it means: The spread between what it earns on loans and pays on deposits
Roughly healthy: 3–4% for a strong retail bank
GNPA / NNPA
What it means: Gross and net non-performing assets — loans not being repaid
Roughly healthy: GNPA under ~3%; NNPA under ~1%
PCR — provision coverage
What it means: How much of bad loans is already provided for
Roughly healthy: Above 70% is conservative
CASA ratio
What it means: Share of deposits in cheap current and savings accounts
Roughly healthy: Above 40% is a real funding advantage
CAR — capital adequacy
What it means: Capital cushion against risk-weighted assets
Roughly healthy: Comfortably above the regulatory minimum
ROA — return on assets
What it means: Profit per rupee of assets; the cleanest bank quality measure
Roughly healthy: Above 1.5% is strong
Cost-to-income
What it means: Operating expense as a share of income
Roughly healthy: Below 45% is efficient

Metric	What it means	Roughly healthy
NIM — net interest margin	The spread between what it earns on loans and pays on deposits	3–4% for a strong retail bank
GNPA / NNPA	Gross and net non-performing assets — loans not being repaid	GNPA under ~3%; NNPA under ~1%
PCR — provision coverage	How much of bad loans is already provided for	Above 70% is conservative
CASA ratio	Share of deposits in cheap current and savings accounts	Above 40% is a real funding advantage
CAR — capital adequacy	Capital cushion against risk-weighted assets	Comfortably above the regulatory minimum
ROA — return on assets	Profit per rupee of assets; the cleanest bank quality measure	Above 1.5% is strong
Cost-to-income	Operating expense as a share of income	Below 45% is efficient

For a bank, asset quality beats growth every time. A bank growing its loan book 30% a year is either taking share brilliantly or lending to people other banks refused — and you find out which about three years later, in the GNPA line.

Where people lose money

Fast loan growth is the warning, not the achievement

Lending is the one business where you can book revenue today and discover the cost of it years later. A loan made badly still looks like an asset — for a while.

So when a lender grows its book far faster than the system, ask what it is lending against, to whom, and at what rate. Unusually high yields mean unusually risky borrowers. The provisions arrive later, and they arrive together.

The other major sectors

What matters where

IT services
The metrics that actually reveal the business: Revenue per employee, utilisation, attrition, deal wins (TCV), dollar-revenue growth, client concentration
FMCG / consumer
The metrics that actually reveal the business: Volume growth separated from price growth, rural vs urban mix, distribution reach, ad spend as % of sales, gross margin
Pharma
The metrics that actually reveal the business: US vs domestic mix, R&D as % of sales, product approvals, regulatory observations at plants, price erosion in generics
Auto
The metrics that actually reveal the business: Monthly volumes (wholesale vs retail), capacity utilisation, discount levels, EV transition, exports mix
Cement
The metrics that actually reveal the business: Realisation per tonne, EBITDA per tonne, capacity utilisation, freight cost, regional mix
Metals / commodities
The metrics that actually reveal the business: Realisation per tonne, cost per tonne, cycle position, integration (own ore or bought)
Real estate
The metrics that actually reveal the business: Pre-sales, collections, inventory overhang in months, net debt, land bank quality
Retail
The metrics that actually reveal the business: Same-store sales growth (SSSG), sales per square foot, store additions, negative working capital
Power / infra
The metrics that actually reveal the business: Order book to revenue coverage, plant load factor, receivable days from state utilities, execution timelines
Insurance
The metrics that actually reveal the business: Persistency ratios, VNB margin, embedded value, claims ratio, solvency ratio

Sector	The metrics that actually reveal the business
IT services	Revenue per employee, utilisation, attrition, deal wins (TCV), dollar-revenue growth, client concentration
FMCG / consumer	Volume growth separated from price growth, rural vs urban mix, distribution reach, ad spend as % of sales, gross margin
Pharma	US vs domestic mix, R&D as % of sales, product approvals, regulatory observations at plants, price erosion in generics
Auto	Monthly volumes (wholesale vs retail), capacity utilisation, discount levels, EV transition, exports mix
Cement	Realisation per tonne, EBITDA per tonne, capacity utilisation, freight cost, regional mix
Metals / commodities	Realisation per tonne, cost per tonne, cycle position, integration (own ore or bought)
Real estate	Pre-sales, collections, inventory overhang in months, net debt, land bank quality
Retail	Same-store sales growth (SSSG), sales per square foot, store additions, negative working capital
Power / infra	Order book to revenue coverage, plant load factor, receivable days from state utilities, execution timelines
Insurance	Persistency ratios, VNB margin, embedded value, claims ratio, solvency ratio

Worked exampleThe same 20% revenue growth, read through three sector lenses

FMCG company: 20% revenue growth
— but volume growth: 2%the rest was price
IT services company: 20% revenue growth
— but headcount growth: 24%no productivity gain
Bank: 20% loan growth
— but GNPA: 2.1% → 3.4%quality deteriorating

FMCG: growth came from price rises, not more households buying → limited runway

IT: revenue grew slower than headcount → margins under pressure ahead

Bank: the book grew by lending to weaker borrowers → provisions coming

So → Identical headline growth, three different problems — and none of them is visible in the generic ratios. Every one required the sector's own metric.

Cyclical vs secular: the distinction that changes everything

Examples
Cyclical: Metals, cement, sugar, chemicals, shipping, autos
Secular growth: Consumer staples, healthcare, IT services, insurance
What drives earnings
Cyclical: The commodity price or the economic cycle
Secular growth: Structural demand growth and share gains
Value it on
Cyclical: Mid-cycle earnings, or price-to-book
Secular growth: Earnings and cash flow, with growth
Low P/E means
Cyclical: Probably peak earnings — a warning
Secular growth: Possibly an opportunity — investigate
Where the risk is
Cyclical: Debt taken on during good years
Secular growth: Multiple contraction if growth slows

	Cyclical	Secular growth
Examples	Metals, cement, sugar, chemicals, shipping, autos	Consumer staples, healthcare, IT services, insurance
What drives earnings	The commodity price or the economic cycle	Structural demand growth and share gains
Value it on	Mid-cycle earnings, or price-to-book	Earnings and cash flow, with growth
Low P/E means	Probably peak earnings — a warning	Possibly an opportunity — investigate
Where the risk is	Debt taken on during good years	Multiple contraction if growth slows

Before analysing any company

☐Which sector is this, and is it cyclical or secular?
☐What are the two or three metrics this industry actually runs on?
☐What is the sector median for the ratios I'm about to use?
☐Is this business capital-light or capital-heavy? That changes what good returns look like.
☐Am I comparing it against genuine peers, or against the whole market?

See it live · Our engine judges each company through one of eight sector frameworks — a bank is scored as a bank, an FMCG as an FMCG — rather than one universal checklist.

See sector frameworks →

In one screen

1Generic ratios break for banks: use NIM, GNPA, PCR, CASA, ROA and cost-to-income.
2Every sector has two or three metrics that reveal how it really works.
3Cyclical and secular businesses need different valuation approaches entirely.
4Always benchmark against the sector median and the company's own history.

Check yourself

Nothing is scored or stored — the explanation after each answer is the point.

1/2

A bank grows its loan book 35% while the banking system grows 12%. What should you check first?

2/2

An FMCG company reports 18% revenue growth with 1% volume growth. What does that mean?

Next lesson →

Your 45-minute research routine

← Previous

What the price is already assuming

← All lessons