Learn the tools of the trade.

Economy analysis

The economy sets the ceiling on what any company inside it can earn. A well-run business in a slowing economy will still find its customers spending less, its borrowing costing more and its shares valued against a more attractive bond. Hence we read the macro indicators before the company, and we read them as a chain rather than a checklist: production rises because demand exists, demand becomes revenue, revenue becomes profit, and the RBI's rate decisions decide how much of that profit the market is willing to pay for.

Growth indicators

These three measure the same thing, i.e. whether economic activity is increasing, at different speeds and from different angles. PMI arrives first, IIP second and GDP last, and hence they are read together rather than separately.

Monetary policy

The RBI controls the amount of money in the system through four instruments. Each one either restricts how much banks can lend or changes the price at which they borrow, and hence each one reaches every company in the economy through its cost of capital.

How RBI decisions reach share prices

Structural indicators

These move more slowly than the growth and policy indicators, but they set the backdrop against which the others are read.

Where to get the data

Industry: Banking

The ratios in the ratio toolkit do not work on banks. A bank's raw material is money, and hence debt is its inventory rather than a burden, which makes debt to equity meaningless. There is no revenue line in the normal sense, no gross margin and no inventory turnover, and free cash flow does not behave the way a standard DCF assumes, which is why the Valuation page already has a separate bank/NBFC mode that values net profit directly. A bank is a spread business wearing a balance sheet, and it needs its own set of numbers. The eleven below are organised as four questions.

1. Does it earn a good spread?

A bank borrows money from depositors at one rate and lends it to borrowers at a higher rate. The difference between the two is where all of its profit comes from, and hence the first question is how wide that difference is and how cheaply the bank is funded.

2. Is the loan book clean?

A bank's profit is only as good as the loans behind it. A wide spread earned on loans that are not repaid is not profit at all, and hence asset quality is checked before the spread is trusted.

3. Is it growing safely?

Growth in a bank is growth in lending, and lending is only good if the bank has the capital to absorb the loans that go wrong. Hence growth and safety are read together, never separately.

4. Is it run efficiently?

The last question is how much it costs the bank to run its branches, staff and systems relative to what it earns. This is where private banks have historically separated from public sector banks.

The order to read them in

Cash flow and earnings quality

Profit growth can be managed through accounting choices. Operating cash flow and free cash flow are much harder to manage, as they show what actually came into the company's bank account. This chapter covers the six checks we use to compare what a company reports as profit with the cash it actually generated.

1. Managerial efficiency — CFO ÷ PAT

Cash from operations divided by profit after tax (shows how much of the profit reported actually comes into the business as cash). Higher is better, and it should stay above 1 over any period of several years. Profit includes sales that have not yet been collected and excludes costs that have not yet been paid, whereas cash from operations is what cleared the bank. Comparing the two is the fastest quality-of-earnings check available. BLS International, for example, has averaged a CFO to net profit ratio of 1.4 over the last five years, showing that its revenue is turning into actual payments from customers. A company that reports rising profit while CFO stays flat is showing, in its own numbers, that the profit exists only on paper.

2. The cash flow ratios

3. Free cash flow

FCF = CFO − net capex. Free cash flow is the cash from operations less what the company had to spend on plant and equipment in order to maintain and grow the business. This is the cash that is genuinely free: it can be paid out as dividends, used to repay debt, or reinvested. Dividing it by the number of shares outstanding gives FCF per share, which can then be compared directly with the share price.

4. The FCF yield ladder

FCF per share divided by the share price gives the cash return an investor is buying at today's price. This is the same idea as an earnings yield, but uses the figure that is harder to manage.

5. Working capital and the cash conversion cycle

Working capital = inventory + receivables − trade payables. It is the money tied up in running the business from day to day: stock on the shelf and invoices not yet collected, less the credit that suppliers have extended. Most investors assume that more working capital is safer. In practice the opposite is usually true.

Negative working capital is a good thing, as it means the business is being financed by its suppliers. A negative cash conversion cycle means the company collects from customers before it has to pay its suppliers. It receives the cash for the sale, holds it for a few weeks, and only then settles the invoice for the goods. Such businesses usually carry little debt, for the simple reason that they do not need any. BLS collects its processing fees upfront before providing the service, and hence has no receivables to speak of and no need to borrow. The balance sheet in Chapter 1 is the shape this produces: receivables shrinking, cash building, and borrowings falling to zero.

6. The receivables warning

Check how fast receivables are growing compared with revenue. If receivables are outpacing revenue, the company is booking sales it has not yet been paid for. This usually means it is selling on credit in order to keep the growth number up, and it is carrying a collection risk that the P&L does not show. This is the check that catches the most common way in which profit is managed, and it requires only two line items and a division.

What to invest in

Stable businesses with regular cash flows. We avoid a business whose future cash flow cannot be estimated, or whose cash flow has been negative or irregular. This is not because such businesses never succeed, but because there is no way to value them, and a price that cannot be checked cannot be defended. Even with stable cash flows, a margin of safety should be applied. A good business is not worth buying at any price.

Capital allocation and the checklist

This is the qualitative half of the analysis. Ratios show what has already happened. This chapter covers how we judge whether management will keep it happening, starting with what they do with the cash the business generates. That decision, repeated over a decade, determines most of what a shareholder ends up owning.

Four ways to check capital allocation

1. Capital expenditure. Is capex only sustaining the business, or is it expanding it? Sustaining-only capex is a warning sign, as the company is spending simply to stay where it is. Large capital invested at a very high rate of return is the best wealth-creating combination there is, and large capital invested at a poor return is the worst. Most companies sit somewhere in between, and the return measures below show where.
2. Share buybacks. Is the company using cash to buy back its own shares? This is good for shareholders, as both earnings per share and book value per share rise, provided the shares are bought below what they are worth.
3. Mergers and acquisitions. Is the company buying businesses in order to prop up operations that are slowing down on their own? Management's track record on past acquisitions needs to be examined: what was paid, what was promised, and what was actually delivered.
4. Cash. How much is simply being held. Idle cash earns a bond return and pulls down every return measure in this chapter. Sometimes this is the honest choice, and sometimes it shows that management has nowhere to deploy it.

The return measures

Three measures that sound alike but answer different questions, followed by the formula that ties them together. This is the section most worth getting right.

The difference between the last two is the important one. A business with an excellent ROIC but no room to reinvest will continue to generate cash, but it has stopped compounding. A business with an ordinary ROIC and a very high ROIIC is improving with every rupee it deploys. Only the incremental figure shows which of the two the investor is looking at.

ROIIC is calculated over a chosen window of three, five, seven or nine years, in five steps:

The compounding formula, stated plainly: intrinsic value compounding rate = ROIIC × reinvestment rate. A business earning 20% on its reinvested capital while reinvesting 80% of its earnings compounds its intrinsic value at roughly 16% a year. Both terms matter. A very high ROIIC with nothing to reinvest in compounds slowly, and heavy reinvestment at a poor return destroys value faster than doing nothing at all.

Worked example — BLS International, ROIIC across four windows

BLS's invested capital (equity plus debt less cash) moves from ₹191 Cr in Mar-17 to ₹1,712 Cr in Mar-26. Each window below runs backwards from Mar-26, so B is the difference between ₹1,712 Cr and the invested capital three, five, seven or nine years earlier, and A is the net profit added up over the same years.

This is a clean example for three reasons. The reinvestment rate sits between 69% and 87%, showing that BLS is funding its growth out of its own earnings rather than outside capital, which is the situation the formula is designed for. ROIIC is steady at 38–44% across every window, which shows durability: the same answer comes out whichever start date is chosen, rather than one exceptional year. Lastly, the value compounding rate stays at 30–35% regardless of the window, and hence the investor gets a stable estimate rather than one that swings with the arithmetic.

The last row carries the main lesson. Market-cap CAGR is 13%, 59%, 35% and 21% against a compounding rate that barely moves. This is the same business with the same economics, and the four very different price outcomes depend only on which year the count started from. The five-year figure is high because Mar-21 was a lockdown low, and the three-year figure is low because Mar-26 comes after a 41% fall from the Mar-25 peak. Neither says anything about how the business performed in those years. This is the caveat above shown in practice: the compounding rate shows how fast the business is building value, and the price shows what the market was willing to pay on two particular days.

Warren Buffett's ₹1 test

The test in one line: for every ₹1 of earnings the business retains, it should create at least ₹1 of market value. If it does not, shareholders would have been better off receiving the money as dividends.

Change in market cap over the period ÷ total retained earnings over the same period, where retained earnings = net profit − dividends paid. Higher is better, and a figure below 1.0 means that retaining earnings has destroyed value. Buffett set out the principle in his 1984 letter: unrestricted earnings should be retained only where there is a reasonable prospect, backed by historical evidence or a thoughtful analysis of the future, that for every dollar retained at least a dollar of market value will be created for owners.

Worked example — BLS International, Mar-17 to Mar-26

3.8 is a good result and a realistic one. It is comfortably above the 1.0 threshold without being the kind of outlier that sets a false benchmark for the next company the test is run on.

The end-date problem, which this example shows almost by accident. BLS's market cap peaked at ₹16,412 Cr in Mar-25 and fell to ₹9,702 Cr by Mar-26, a drawdown of 41%. Running the same test with Mar-25 as the end date gives approximately 10.7 instead of 3.8. This is the same business, the same retained earnings and the same nine years of capital allocation, and the verdict nearly triples depending on which March the calculation stops at.

Consistent, durable earnings

Five questions. A business that answers yes to all five is rare, and most of the ones that do are the ones worth paying a premium for.

• Consistently high profitability?
• Consistently zero or low debt?
• Consistently avoids large capital investment?
• Consistent earnings?
• Consistent growth in earnings?

The business questions

These are the questions we ask before anything else, before the ratios and before the valuation.

• Is it a simple business?
• Has it consistently grown sales and EPS for five to ten years? Consistency matters more than the rate of growth.
• Does the product have strong demand? Is there a long runway for the next ten years? Can it be easily disrupted?
• Is it highly capital intensive? Can it scale without destroying capital at a low or negative ROCE?
• Will it be around in ten years, and better off than it is today?
• How competitive is the industry, and does the business have a moat? Examples include pricing power, high profit margins, superiority over competitors, entry barriers and economies of scale.
• How good is management on capital allocation, on ROE, and on performance against competitors over five to ten years?

Reading annual reports

The auditor's opinion should be read first, and read for qualifications. After that, unusual losses, debt to equity, revenue growth, and receivables compared with revenue, which is the Chapter 3 check done from the source document. Competitors' annual reports should be read as well, as a company's own report says what it wants investors to know and a competitor's report often shows what was left out. On promoter holding, a fall in promoter equity over a single year is fine and usually has an ordinary explanation. A constant downward trend is not.

Growth vs mature balance sheets

The same line item can be healthy on one balance sheet and a warning sign on another. It is important to decide which kind of company is being analysed before judging the lines.

Worked examples

Two numerical examples that explain more of the method than any formula. Both are built as calculators, so the inputs can be changed to see whether the conclusion still holds.

Example 1 — Why growth beats a cheap multiple

Consider two companies, both trading at ₹20 with earnings per share of ₹1.00, i.e. a P/E of 20 each. Company A grows its earnings at 20% a year and Company B at 10%. After ten years, at the same P/E of 20, A is worth ₹123.80 per share and B is worth ₹51.80. The important part is what happens if investors stop believing in A's growth. Even if its P/E de-rates to 15, the stock is still worth ₹92.85, which is nearly double B at its full multiple. This shows that a higher growth rate matters far more to the final return than a cheaper starting multiple.

The table below is that arithmetic. A's growth rate and exit P/E can be changed to find the point at which the conclusion stops holding. B stays fixed at 10% growth and a P/E of 20 as the benchmark.

Example 2 — Operating leverage and profit margins

Consider two companies, both with ₹100 of sales. Company A is the high-margin operator, with ₹88 of costs and ₹12 of pretax profit. Company B runs on ₹98 of costs and ₹2 of profit. Now sales rise by 10% and costs rise by 5% at both companies.

A's profit goes from ₹12 to ₹17.60, an increase of 47%. B's profit goes from ₹2 to ₹7.10, an increase of 255%. The lesson works in both directions. The highest-margin company is the lowest-cost operator and hence has the best chance of surviving a downturn. However, in an upswing it is the lowest-margin company that benefits the most, as a small change in sales is a very large change relative to the little profit it had to begin with. Moving cost growth above sales growth in the calculator shows the same leverage working against the low-margin company.