P/E and its traps

P/E = share price ÷ earnings per share

Equivalently, market capitalisation over net profit. One number, quoted everywhere, and more misread than any other in this subject.

What it means

How many rupees you pay for each rupee of annual profit. A P/E of 20 means paying ₹20 for ₹1 a year — and, if nothing changed, taking twenty years to get your money back.

Inverted, it is an earnings yield: a P/E of 20 is 5%. That inversion is useful, because it makes the multiple comparable with the bond yields of the fixed income subject and shows what you are being offered.

Chapter 16 established what justifies a higher multiple: more growth, less risk, better conversion of earnings into cash.

The denominator is the problem

Everything that goes wrong with P/E goes wrong underneath the line.

Which earnings? Trailing twelve months, last reported year, or a forecast. A "P/E of 18" on next year's optimistic estimate and on last year's actuals are different claims. Forward multiples are always lower, because forecasts almost always show growth.

Earnings after what? Chapter 4: net profit includes other income, exceptional items, and the tax position. A company whose profit was lifted by selling land has a flattering denominator and therefore a flattering multiple.

Earnings of what quality? Chapter 14. Two companies reporting ₹100 crore where one collects it in cash and one does not should not trade at the same multiple, and the P/E cannot tell them apart.

The trap that costs the most

A cyclical company at the top of its cycle has a low P/E, and that is exactly when it is most expensive.

Commodity producers, steel, cement, shipping. At the peak, profits are extraordinary and the P/E looks tiny — five, or three. The multiple is low because the market knows the earnings will not persist.

Buying the low multiple means buying peak earnings that are about to fall, and when they do the price falls with them. The P/E was not wrong; it was describing a denominator with a short life.

The reverse also holds: a cyclical at the bottom shows an enormous P/E or none at all, which is frequently the better moment.

The defence: for cyclical businesses, compare price against average earnings over a full cycle, not against this year's. Or use chapter 19's price to book, which is anchored to assets rather than to a volatile profit line.

The value trap

A low P/E is also what a declining business looks like.

A company whose earnings fall 10% a year can trade at eight times and still be expensive, because the multiple is applied to a number that shrinks. The market is not failing to notice it is cheap; it is pricing the decline.

So "low P/E" is the start of an investigation with three possible endings: the market is wrong (rare), the earnings are temporarily high (cyclical), or the business is deteriorating (common). The work is deciding which.

When a high P/E is justified

Also worth stating, because the mirror error is as expensive.

A business earning 30% on capital, growing well, converting earnings to cash and protected by something durable is genuinely worth more per rupee of current earnings than a mediocre one. Chapter 16's arithmetic says so: more growth and lower risk raise the multiple, and chapter 9 says what makes growth sustainable.

The risk is not that high multiples are always wrong. It is that they leave nothing for disappointment. Chapter 2 of the equity subject showed de-rating wiping out years of genuine earnings growth, and a high multiple is where that mechanism has most room to work.

Comparing P/Es properly

Within an industry. Across industries a P/E difference is mostly about growth and capital intensity, not value.

Against the company's own history. A business at 35 times against a ten-year median of 20 is being priced for something different from its past.

Against the market. The relative multiple strips out market-wide re-rating, which chapter 9 of the equity subject noted can make everything look reasonable relative to everything else.

And the question chapter 9 of the IPOs subject insisted on: who was left out of the comparison? A peer set is a choice.

The PEG shorthand, and its limits

PEG = P/E ÷ earnings growth rate

A rough adjustment for growth, with a convention that below 1 is attractive. It is a useful back-of-envelope and it has real weaknesses: it ignores risk entirely, ignores how long the growth lasts, and ignores whether earnings become cash. Treat it as a sorting device, not a conclusion.

The point

P/E is what you pay for a rupee of annual profit, and the denominator can be chosen, inflated by one-offs or about to collapse. A low P/E on a cyclical at its peak is the most expensive thing in the market, and a low P/E on a declining business is the market pricing the decline. A high multiple can be justified and leaves no room for disappointment.