P/E, P/B and ROE: three ratios that do most of the work

A share price on its own tells you nothing. Four hundred rupees is cheap for one company and absurd for another. A ratio puts the price next to something the business actually produces, so you can compare a bank with a paint maker, or a company with its own past. There are dozens of ratios, and screening apps will happily show you all of them. Three carry most of the weight.

Price to earnings

Divide the share price by the earnings per share, which is the year’s profit split across all the shares. A share at 400 rupees with earnings of 20 rupees per share trades at a P/E of 20. You are paying 20 years of today’s profit for the business, or, turned upside down, the business earns 5 percent a year on the price you paid. That inverted number, the earnings yield, is worth comparing with a fixed deposit rate now and then. It keeps you honest.

The Nifty 50 has spent most of the last two decades somewhere around 20 times trailing earnings, dipping into the low teens in panics and above 30 in euphoria. A single company’s P/E only means something next to its own history and its sector.

A high P/E says one of two things: the market expects profits to grow quickly, or this year’s profit is temporarily depressed. A low P/E also says one of two things: the share is cheap, or the market thinks the profits are about to fall. Working out which is the whole job.

Watch for the ways P/E lies. A one-off gain, such as selling a factory, inflates earnings for one year and makes the ratio look low. Cyclical businesses such as steel, sugar and shipping look cheapest at the top of their cycle, when profits are fat and about to collapse, and most expensive at the bottom, which is usually the time to buy. A company with a loss has no P/E at all, and apps show a dash where the trouble is. And “trailing” P/E uses the last twelve months while “forward” P/E uses somebody’s estimate of the next twelve, so check which one you are looking at before comparing.

Price to book

Divide the share price by the book value per share, where book value is what the accountants say the company’s assets are worth after paying off every liability. A P/B of 3 means you are paying three rupees for every rupee of net assets on the books.

This ratio is most useful where the assets are financial and recorded close to their real value: banks, housing finance companies, insurers. Comparing HDFC Bank with a smaller private bank on P/B is a reasonable conversation. For a software company or a consumer brand, book value leaves out the things that make the business valuable, the engineers and the brand, so P/B tells you little. Nobody buys a toothpaste company for its factories.

Return on equity

Divide the year’s profit by shareholders’ equity, the same book value from above. This is the rate at which the business compounds its owners’ money. A company with a 20 percent ROE that keeps its profits can, in principle, grow its book value 20 percent a year without raising fresh capital. Sustained ROE above 15 percent is a good sign in India; above 25 percent for a decade is rare and usually signals a real competitive advantage.

ROE lies in one specific way. Profit over equity rises when a company borrows heavily, because debt adds assets without adding equity. Split ROE into its three parts, profit margin times asset turnover times leverage (the DuPont breakdown you will see on screener sites), and check where the number comes from. A 20 percent ROE from fat margins is a different animal from a 20 percent ROE from borrowing five rupees for every rupee of equity. Banks run on leverage by design, which is why you judge them on P/B and on the quality of their loans rather than on ROE alone.

The arithmetic that ties them together

P/B equals P/E multiplied by ROE. Write it on the sticky note from the Sensex and Nifty chapter. If a company earns 15 percent on its equity and trades at 20 times earnings, its P/B must be 3. If an app shows you a P/B of 6 for a company with a 15 percent ROE, either the P/E is 40 or something is wrong with the data. The identity is a free error check, and it also explains why high-ROE businesses look “expensive” on P/B: they are supposed to.

A worked example

Two hypothetical companies, both selling at 400 rupees a share.

	Company A	Company B
Earnings per share	20	20
Book value per share	160	80
P/E	20	20
P/B	2.5	5.0
ROE	12.5%	25%
Debt to equity	0.2	2.0

Same price, same profit, same P/E. B looks like the better business because its ROE is double. Then you see the last row: B carries ten times the debt relative to its equity. Its 25 percent is borrowed. In a bad year, A’s profit halves and B’s disappears under interest. Neither company is “the answer”. The point is that any one ratio hid half the picture, and three together with one extra line told it.

How to use them without fooling yourself

Compare within a sector, never across. Look at five to ten years of each ratio rather than today’s value; a company whose P/E has ranged between 15 and 25 and now sits at 14 is a different proposition from one that has never traded below 40. Use the free screeners, Screener.in, Tickertape and the exchange websites, to get the numbers quickly, and then check the ones that matter against the annual report you learned to read in the previous chapter. Data vendors make mistakes, and they are never in your favour at the moment you most need them not to be.

Finally, none of these ratios tells you what a business will earn in five years, which is the thing you are actually buying. They tell you what you are paying for what it earns now. That is the right starting point and the wrong finishing line.

Try this before moving on

Take the three companies from the Basics course and compute P/E, P/B and ROE yourself from the latest annual report: profit, number of shares, equity. Then compare your numbers with what your broking app shows. If they differ, find out why. Usually it is trailing versus forward, or standalone versus consolidated. Occasionally it is a mistake, and you will never trust an app blindly again.