P/E: what it is and why low does not mean cheap

P/E is the market price of a share divided by the earnings attributable to one share; the result is conveniently read as the number of years it would take the company, at unchanged earnings, to "earn back" its own valuation. A low P/E says exactly one thing: the market is paying little for this company's current earnings. There is almost always a reason for that — the earnings are seen as one-off, declining, made at the peak of the cycle or assembled from paper items. That is why the multiple is neither a price tag nor a verdict, but a question to put to the financial statements. The current value for a security is always shown on its card: 3,77, next to the rest of the instrument's data {{instrument:SBER}}.

What exactly is divided by what

The numerator is the price at which the security is trading right now. This is the most honest part of the formula: it updates throughout the trading session and does not depend on anyone's interpretation.

The denominator is earnings per share. And this is where everything interesting begins. Earnings are taken from the financial statements, and the statements come out with a delay and in several versions. The variants of the denominator that give different P/E figures for the same security on the same day:

• earnings for the last completed financial year;
• earnings for the most recently published quarters, added up into a rolling year ("trailing");
• forecast earnings for a future period ("forward") — not a fact, but somebody's opinion;
• earnings stripped of one-off items — and every analyst does the stripping in their own way.

Hence the practical rule: the P/E of two companies can be compared only when you know that the denominator was calculated in the same way. It is the same trap as with bonds, where yield is measured in several different ways and the discrepancy between brokers is explained by methodology, not by error.

Why the denominator is the weak spot

Earnings are a calculated figure, not a fact of cash coming in. They can be legitimately increased by a revaluation of assets, the sale of a stake in a subsidiary, a foreign exchange difference, or the release of a previously created provision. All of this lands in the very line you are dividing the price by — and disappears in the next period.

The reverse is also true. A one-off write-off, an impairment provision or a large non-cash expense can crush the earnings of a company whose business has not changed at all. The P/E will soar instantly, and the security will look "expensive" right up until the next publication.

This should be checked not in the press release but in the cash flow statement: operating cash flow shows whether the money came in, whereas earnings show how it was recognised. A gap between earnings and cash flow that persists for more than one period is a reason to treat a low P/E as a warning. Primary sources on issuers are collected in the financial statements section.

The cycle: why everything looks cheap at the peak

At commodity, steel and chemical companies, earnings move in waves along with the price of the product. At the top of the cycle the denominator is at its maximum and the P/E at its minimum — and the security looks cheapest at precisely the moment when earnings are about to head down. A period later the price will have fallen, earnings will have fallen further, and the multiple will turn out to be higher than it was in the "expensive" market.

So for cyclical stories — say, {{instrument:LKOH}} or {{instrument:GAZP}} — the value of the multiple at a given moment is less informative than its position relative to the company's own history and the phase of the cycle. A static comparison with the "market average" here gives the opposite of the right answer.

When the multiple simply does not work

• Banks and insurers. Earnings depend on provisions and on the interest rate; valuation is usually done through capital rather than through P/E in its pure form. Here it is more useful to watch how the regulator's signal is worded than the multiple itself.
• Companies with heavy debt. P/E ignores the capital structure: it looks only at the shareholder's portion. A company with a heavy debt load and a company with no debt are worth different amounts at the same P/E. For such comparisons EV/EBITDA is used.
• Fast-growing and loss-making companies. The denominator either does not exist or does not reflect what the security is being bought for.
• Securities with low liquidity. The numerator is then set by infrequent trades rather than by the market; this is addressed directly in the piece on how liquidity is noticed only when it runs out.

How to check a low P/E, step by step

Step 1. Find out which period the earnings are taken from and whether a one-off item has found its way into them.

Step 2. Compare earnings with operating cash flow for the same periods.

Step 3. Compare the multiple not with the market as a whole, but with the company's own history and with its sector peers.

Step 4. Look at the debt: if it is large, the cheapness on P/E may be an optical illusion.

Step 5. Check the payouts — a low P/E with sustainable dividends and a low P/E with dividends cancelled mean different things. The nearest record dates: {{dividend_calendar|limit=5}}, the full list is in the dividends section, and the securities themselves are in the stocks section.

There is one more layer that is entirely absent from the multiple: the company's earnings are subject to its own tax, and the investor's income to another, and the rules for determining the tax base are set out separately: 13%. P/E is calculated before this layer, which means that the "payback" it shows is a figure for the company, not for your portfolio.

Finally, an honest caveat about the limits of the method. P/E is a snapshot of the relationship between price and recognised earnings. It contains no information about the quality of revenue, about the capital expenditure that will be needed to sustain those earnings, or about who will receive the earnings of future periods. No multiple answers the question "is this cheap?"; it only points to where to look for a mismatch between the valuation and the business. Definitions of related terms are in the glossary.