How to Analyse a Stock in India: A Professional Research Checklist

Stock analysis is not the act of finding enough positive facts to justify a purchase. It is a process for answering a harder question:

What must be true for this company to create value from today’s price, and what evidence would show that the thesis is wrong?

That question forces business quality, numbers, expectations and risk into the same frame. It also prevents a familiar mistake: completing a detailed company study without deciding what should be monitored after the investment is made.

This is a practical research sequence for Indian listed companies. It is designed to work whether the starting point is a screener result, a quarterly announcement, an analyst idea or a company already in the portfolio.

The 12-step stock-analysis framework

Step	Question	Evidence to preserve
1	What does the company sell, to whom and why?	Annual report, segment note, customer and product mix
2	What makes demand grow or shrink?	Volumes, prices, capacity, market share and industry data
3	Where does the company make money?	Segment revenue, margins and capital employed
4	Are reported profits converting into cash?	Cash flow, receivables, inventory and payables
5	How fragile is the balance sheet?	Debt, cash, maturities, guarantees and contingent liabilities
6	Does the company reinvest well?	ROCE, incremental returns, capex and acquisitions
7	How has management allocated capital?	Dividends, buybacks, dilution, M&A and related-party transactions
8	Does management do what it says?	Dated guidance versus subsequent outcomes
9	What could be misstated or misunderstood?	Auditor notes, exceptional items and forensic signals
10	What is the market already expecting?	Valuation history, scenarios and available estimates
11	How does the position change the portfolio?	Exposure, concentration, common drivers and liquidity
12	What evidence should trigger a review?	Thesis conditions, thresholds and monitoring alerts

The sequence matters. Valuation comes after the operating model, not before it.

1. Explain the business without using management adjectives

Write one paragraph that names the customer, product, route to market, revenue mechanism and major cost. Avoid words such as leading, premium and innovative unless they are supported by a measurable fact.

For a consumer-appliance manufacturer, for example, the research map may include units sold, average selling price, product mix, dealer reach, localisation, commodities, currency and warranty costs. For a bank it would instead include deposits, loan mix, spreads, credit costs and capital.

If the company cannot be explained clearly, the financial model will probably contain drivers that are labels rather than causes.

2. Identify the operating equation

Revenue is usually the product of a few underlying variables.

Examples include:

• manufacturer: volume × realisation
• retailer: stores × sales per store
• hotel: available rooms × occupancy × room rate
• lender: earning assets × spread, adjusted for credit cost
• platform: customers × frequency × value per transaction × monetisation
• project company: opening order book + inflow − execution

Build the equation before forecasting growth. A 20% revenue target based on capacity, utilisation and realisation is testable. “The industry has a long runway” is not.

3. Read the segments before the consolidated total

Consolidated revenue can hide a growing low-margin business, a shrinking high-return franchise or a cyclical segment at peak profitability. Compare each meaningful segment on four dimensions:

1. growth;
2. operating margin;
3. capital intensity; and
4. cash requirement.

The segment with the most revenue is not necessarily the segment that creates the most value. The useful question is how much incremental operating profit and cash a segment produces for the capital it consumes.

4. Reconcile profit with cash

Profit after tax is assembled under accounting rules. Cash flow records when money moved. Over a sensible period, compare cash flow from operations with attributable profit and then explain the bridge.

Look specifically at:

• receivables growing faster than sales;
• inventory building ahead of demand;
• payables funding the business temporarily;
• capitalised costs;
• repeated exceptional items;
• taxes paid versus the income statement; and
• capital expenditure required merely to maintain operations.

One weak quarter may be timing. A persistent gap is a research question.

5. Stress the balance sheet

Net debt alone can be misleading. Record gross debt, cash quality, lease obligations, maturity dates, interest rates, guarantees and contingent liabilities. Then test a downside case in which EBITDA falls, working capital absorbs cash and refinancing becomes expensive at the same time.

For financial companies, replace industrial leverage ratios with the sector’s own solvency, funding, liquidity and asset-quality measures. Never force a generic ratio onto a sector where it is not economically meaningful.

6. Understand the two engines of ROCE

Return on capital employed can be decomposed conceptually into operating margin and capital turnover:

ROCE ≈ operating margin × capital turnover

A low-margin distributor can earn a strong return by turning a small capital base rapidly. A high-margin manufacturer can earn a weak return if plants, inventory and receivables absorb too much capital.

That is why a high ROCE is not a complete stock thesis. Ask whether the company can reinvest at a similar return and how large that opportunity is. Efficiency and reinvestment runway are separate questions.

See the worked low-margin ROCE example.

7. Audit capital allocation

Operating performance can be good while shareholder outcomes are poor if surplus cash is misallocated. Build a five-year cash-allocation table covering:

• organic capex;
• acquisitions and disposals;
• dividends and buybacks;
• debt reduction;
• equity issuance; and
• investments in subsidiaries or related parties.

Then compare what management said each project would achieve with what happened. A disciplined capital allocator closes the loop; an undisciplined one repeatedly resets the story.

8. Build a management-guidance ledger

Do not grade management on confidence or presentation quality. Preserve dated claims.

Date	Management statement	Metric and horizon	Later outcome	Status
Result call	Revenue-growth range	FY27	Future filing	Open
Annual report	Plant commissioning	Q3 FY27	Exchange filing	Open
Prior year	Margin objective	FY26	FY26 result	Met / missed / changed

The most useful signal is often not one miss. It is the pattern: conservative delivery, repeated delay, moving definitions or unexplained abandonment.

9. Separate anomalies from accusations

Forensic analysis should identify questions, not manufacture guilt. Review auditor qualifications, related-party transactions, pledged shares, unusual loans, contingent liabilities, working-capital divergence, frequent CFO turnover and changes in accounting policy.

Every flag needs context. Fast growth can legitimately absorb working capital. A corporate action can alter a time series. A missing number is not zero. Trace an anomaly to the filing and reporting basis before explaining it.

10. Value scenarios, not a single target

Start with normalised earnings or cash flow, not the most flattering reported period. Then write three cases with explicit drivers.

Case	Revenue and margin	Capital and cash	Valuation logic
Downside	Demand or pricing weakens	Working capital and leverage worsen	Lower earnings and multiple
Base	Operating plan broadly delivered	Normal reinvestment	Historically defensible range
Upside	Share, mix or utilisation improves	Returns remain healthy	Higher cash flow, not hope alone

A low P/E may be a peak-cycle illusion or include a one-time gain. A high P/E may still disappoint if the market already expects flawless execution. Valuation is the price of assumptions.

11. Add portfolio context

A company can be attractive in isolation and still be the wrong addition. Check whether the portfolio already has the same exposure through sector, commodity, currency, interest rates, promoter group or customer concentration.

Position size should reflect uncertainty, liquidity and downside—not only upside. The research memo and the portfolio decision are connected but not identical.

12. Convert the thesis into monitoring rules

End the memo with five items:

1. the core thesis in one paragraph;
2. the three assumptions that matter most;
3. evidence that would strengthen each assumption;
4. evidence that would weaken or falsify it; and
5. the next filing, result or operating KPI to review.

“Monitor quarterly results” is not a rule. “Review if two-quarter gross margin falls below the thesis range while inventory days rise” is closer to one.

Where AI helps—and where it does not

AI is useful for locating clauses, comparing documents, extracting management statements, drafting bridges and asking the same question across many companies. It becomes risky when a fluent answer is mistaken for evidence.

A dependable division of labour is:

• code calculates financial metrics and applies rules;
• source-linked retrieval supplies documents and dates;
• AI interprets language and organises evidence;
• the analyst chooses assumptions and position size.

On Altys, company research, guidance history, scorecards, models and monitoring sit in one source-linked workflow. Quantitative work can be exported to Excel so a second analyst can reproduce the rules and calculations independently. That verification step matters more as AI makes polished prose cheaper.

The short answer

To analyse an Indian stock well, move from business → operating drivers → financials → cash and balance sheet → management → valuation → portfolio → monitoring.

The goal is not a longer memo. It is a smaller set of dated, testable claims that can survive the next quarter.

Primary research starting points: NSE corporate filings, BSE corporate announcements, company investor-relations pages and audited annual reports.

Related reading:

• How to analyse quarterly results
• How to read an annual report
• How professional investors build a thesis
• How to monitor a portfolio of holdings

This article is educational and does not constitute investment advice.