Fundamental analysis estimates what a business is worth by examining what it earns, what it owns, what it owes, and how durable all of that is. If your estimate is materially above the market price, the stock is a candidate. If it is below, it is not.
The method is old and well documented. What is less well documented is which parts of the disclosure to trust, and which warning signs have historically preceded the failures. That is where most of this article goes, using Indian filings for the concrete examples, though the reasoning transfers to any market with comparable disclosure.
For the question of when to act on a view, see technical analysis. For turning any of this into a testable rule, quantitative analysis.
The three statements, and which one lies least
The income statement shows revenue, costs and profit over a period. It is the most quoted and the most manipulable, because profit depends on judgement calls: when revenue is recognised, how quickly assets are depreciated, what gets capitalised rather than expensed.
The balance sheet is a snapshot of assets, liabilities and equity at one instant. It tells you about leverage and solvency. Its weak point is that many assets are carried at values reflecting an accounting policy rather than what anyone would pay.
The cash flow statement tracks cash actually moving. It is the hardest to manipulate, because cash either arrived or it did not, and it is the statement experienced analysts read first.
The relationship between the first and the third is where most of the useful signal lives.
The mechanism behind that divergence is usually receivables or inventory. Sales are booked, profit is recorded, and the cash never arrives because customers are not paying or goods are not moving. Both show up on the balance sheet as growing assets, which looks like expansion until it is recognised as a write-off.
Ratios, grouped by the question they answer
Ratios are compression, not insight. Each answers one narrow question.
| Question | Ratios | What to watch for |
|---|---|---|
| Is it profitable? | Net margin, return on equity, return on capital employed | High ROE achieved through leverage rather than operating quality |
| Is it expensive? | Price to earnings, price to book, EV to EBITDA | Comparisons only mean something within the same industry |
| Is it solvent? | Debt to equity, interest coverage | Interest coverage is the more urgent of the two |
| Is it growing? | Revenue and earnings growth, PEG | Growth funded entirely by debt or equity issuance is not the same as growth |
Return on capital employed deserves more attention than it usually gets. Return on equity can be inflated simply by borrowing more, so a company with mediocre operations and heavy debt can post an impressive ROE. Return on capital employed looks at returns against all capital, debt included, and is much harder to flatter.
On valuation multiples, the only useful comparison is against the same company’s history and against genuine peers. A price-to-book of 1.2 means something entirely different for a bank than for a software company, and cross-industry screening on multiples mostly discovers which industries are asset-heavy.
Governance red flags
This is where generic guides are least useful, because the governance risks that matter most here barely feature in material written for other markets.
Promoter shareholding, and changes in it. Indian companies typically have a dominant promoter group. A high and stable promoter stake generally aligns interests. A steadily falling one, especially through quiet market sales, is worth understanding before you invest alongside it.
Promoter share pledging. This is the one to check first. Promoters pledging their shares as loan collateral creates a mechanism whereby a falling share price triggers lender sales, which push the price lower and trigger more sales. Several well-known Indian collapses followed exactly this sequence. Pledge data is disclosed and it is one of the highest-value checks available.
Related-party transactions. Money moving between the listed company and other entities the promoter controls deserves scrutiny. Some is legitimate operational reality. Some is value leaving the company you own for one you do not.
Auditor resignations and qualifications. An auditor resigning mid-term, or a qualified opinion, is a rare and serious signal. Auditors are not quick to resign, and when they do it is worth more attention than a quarter of earnings.
Contingent liabilities. These sit in the notes rather than on the balance sheet, and for Indian companies they frequently include substantial disputed tax demands. A contingent liability comparable to net worth is a material risk that never appears in any ratio.
Companies listed in India are required to disclose all of this under the exchange listing regulations, and the SEBI circulars listing is where the current disclosure requirements are defined. The filings are published on the exchange sites. Almost nobody reads the notes, which is exactly why reading them is worth something.
Durability, which is what you are actually buying
Every valuation is a claim about the future, so the central question is not what a company earns now but whether it will still be earning it in ten years. High returns attract competition, and competition erodes returns unless something prevents it.
The things that prevent it are worth naming, because they are checkable rather than vague.
Switching costs. How disruptive is it for a customer to leave? Enterprise software embedded in a client’s operations is difficult to displace. A commodity supplier is not.
Scale in a market that rewards it. Being large only helps where size lowers unit costs or improves reach in a way rivals cannot replicate. In many businesses scale brings no advantage at all.
Network effects. The product becomes more valuable as more people use it. Genuinely powerful, and claimed far more often than it exists.
Regulatory position or licensing. Common in Indian financial services, infrastructure and utilities. Durable while it lasts, and dependent on policy that can change.
Brand, but only where it supports pricing. A recognised name is not a moat. A name that lets you charge more than a functionally identical competitor is.
The evidence for durability shows up in the numbers over time rather than in a single year. Stable or rising gross margins through a competitive cycle suggest pricing power. Margins that compress whenever a competitor moves suggest there was never much of an advantage. This is why several years of statements are worth more than the most recent one, however detailed.
Be sceptical of moat narratives that arrive only after strong performance. It is easy to explain why a company that has done well was always protected. The test is whether you could have identified the protection beforehand, from the disclosure.
Valuation, and its dependence on assumptions
Discounted cash flow is the theoretically correct approach: project the cash a business will generate, discount it back at a rate reflecting risk, and sum it.
The practical difficulty is that the answer is dominated by the assumptions you are least able to justify. A large share of a typical DCF value sits in the terminal value, which depends on a perpetual growth rate you are guessing, and the whole result is highly sensitive to a discount rate you are also guessing. Small changes in either produce very different answers.
That does not make DCF useless. It makes it most useful in reverse. Rather than computing a value and comparing it to the price, take the current price and solve for what the market must be assuming about growth and margins. Then ask whether those assumptions are plausible. That reframing turns an exercise in false precision into a genuine question you can reason about.
Where fundamental analysis falls short
It says nothing about timing. A stock can stay mispriced for years, and being early is indistinguishable from being wrong for as long as it lasts.
Reported numbers can be false. Every technique here assumes the statements are broadly accurate. When they are not, ratios computed from them are precise measurements of fiction, and the governance checks above are the main defence.
It works poorly for businesses whose value is not on the balance sheet. Companies whose principal assets are intangible show book values that describe very little.
The easy analysis is already priced. Anything computable from a screener is known to everyone with a screener, and has been since the moment the filing was published. Whatever edge exists lives in judgement about durability, management quality and the parts of the disclosure most people skip, which is inconvenient precisely because those are the parts that cannot be automated or checked quickly.
Where to go next
Combining a view on individual companies into a portfolio is portfolio construction, and sizing it so a single mistake is survivable is risk management. For systematising fundamental signals across many companies at once, see machine learning for stock selection, which is largely about avoiding the ways that process fools you.
If you take one procedural habit from this, make it the reading order. Most people open an annual report at the chairman’s letter and the headline numbers, which are the two sections written to be persuasive. Start instead with the cash flow statement, then the notes to the accounts, then the related-party and contingent-liability disclosures, and only then the narrative. By the time you reach the letter you will know whether it is describing the same company the numbers describe, and that comparison is frequently the most informative thing in the document.
Read the cash flow statement first, read the notes second, and treat the income statement as the company’s argument rather than as the facts.