The Peak-Earnings Trap: Cyclical Stocks at Low P/E

There is an umbrella shop near my old office that does roaring business for exactly four months a year. Walk in during July, mid-monsoon, and the owner is run off his feet — stock flying out, prices firm, money pouring in. If he showed you his books that week and offered to sell you the shop, the numbers would look spectacular. Record sales. Fat margins. A "cheap" price against this year's profit.

Now imagine buying it on that basis — in September, just as the rain stops.

That, in one image, is the peak-earnings trap: paying up for a business at the very top of its cycle, fooled by profits that were never going to last. And the cruellest part is the signal that's supposed to protect you — a low price-to-earnings ratio — is the thing that lures you in.

Why "cheap" lies at the top of a cycle

First, the term, because everything turns on it. The price-to-earnings (P/E) ratio is the company's market value divided by its yearly profit — roughly, how many years of current earnings you're paying for. For most businesses, a low P/E is a green light: you're paying little for each rupee of profit.

For a cyclical — a business whose profits swing wildly with a commodity price or the economy, like steel, sugar, cement, or metals — that rule flips upside down. At the top of the cycle, profits are abnormally high. Divide a normal-looking price by an abnormally high profit and you get a small number. The P/E looks cheap precisely because the "E" is about to fall off a cliff.

Peter Lynch, who ran the legendary Magellan fund, said it without softening: "Buying a cyclical after several years of record earnings and when the P/E ratio has hit a low point is a proven method for losing half your money in a short period of time." (Peter Lynch, One Up on Wall Street). The shop is never cheaper-looking than in the last week of monsoon.

What the steel cycle actually did to its books

You don't have to imagine this. India's largest steelmaker lived it in public.

In FY22, the post-pandemic boom sent steel prices to the roof. Tata Steel reported a consolidated net profit of ₹41,749 crore on sales of about ₹243,959 crore — its highest profit ever, up from ₹8,190 crore the year before (Screener.in, Tata Steel consolidated). Five-fold, in twelve months. That was peak monsoon.

Then the rain stopped. Watch what happened to the margins quarter by quarter — the operating profit margin, meaning operating profit as a percentage of sales, is the cleanest read on a cyclical's health (Inve data, 2026):

These four quarters are picked to trace the peak-to-trough path, not to flatter the story — the in-between quarters (Q4 FY23 onward) wobbled along at a single-digit-to-low-teens margin rather than recovering cleanly. In about fifteen months the operating margin collapsed from 24% to single digits, and a quarter that once threw off ₹7,714 crore of profit turned into a ₹6,511 crore loss — that swing all the way to a reported loss was driven partly by exceptional write-downs, but the underlying squeeze is the operating margin falling by two-thirds. For the full year FY24, Tata Steel reported a consolidated loss after tax of ₹4,910 crore (Tata Steel FY2024 results). The business hadn't broken. The cycle had simply turned, exactly as steel cycles always do.

Here is the trap in numbers. An investor who bought near the FY22 top, anchoring to ₹41,749 crore of profit, was buying at what looked like a single-digit P/E — a screaming bargain by ordinary rules. Two years later the company they "bought cheap" was losing money. The umbrella shop's July books told them nothing about September.

The number to use instead: earnings through the cycle

So how does an owner avoid this? Not by predicting the next steel price — nobody reliably can. By refusing to value a cyclical on a single year's profit at all.

The fix is normalized, or through-the-cycle, earnings: roughly, the average profit the business earns across a full boom-and-bust, not the figure at either extreme. Don't ask "what did it earn at the peak?" Ask "what does it earn in an average year?" For the umbrella shop, that's not the July till — it's the twelve-month total divided by twelve.

Steel looks roughly mid-cycle now — though that is a judgement call, not a fact; as said above, nobody can reliably time the cycle, and where exactly we sit is an estimate that could be wrong. With that caveat, the contrast is instructive. In FY26 Tata Steel earned a consolidated profit after tax of ₹10,886 crore (Tata Steel FY2026 results) — a quarter of the FY22 peak, but more likely repeatable money than a sugar high. Against the current market value of roughly ₹2.7 lakh crore, that's a P/E near 25 and an EV/EBITDA — enterprise value over operating earnings, a debt-aware cousin of the P/E — of about 10.6 (Inve data, 2026). Notice the inversion Lynch warned about: the stock looks expensive on normalized mid-cycle earnings and looked cheap on peak earnings. The expensive-looking version is the more honest one to value against. (For the full mechanics of doing this, see how to value a cyclical stock.)

A caution before you flip the rule the other way: a high trailing P/E on a cyclical is not a buy signal either. It can mean the company is near a trough with depressed earnings — or that the business is genuinely broken, with earnings collapsing for good. A high P/E is just as ambiguous as a low one on a cyclical; both send you back to normalized earnings and the survival checks below, not to a snap decision.

A practical owner's habit: when you see a tempting low P/E on a steel, sugar or metal stock, pull up five to seven years of operating margin. If this year's margin towers over the others, you're not looking at a cheap stock. You're looking at a peak.

What a five-year owner should actually check

If you're going to own a cyclical, own it like a shopkeeper, not a tourist who wandered in during the rains. Three checks do most of the work.

Where are margins versus their own history? Today's operating margin against the last five to seven years tells you, at a glance, whether you're near the floor or the ceiling. Buying near the floor tends to be uncomfortable and, for a business that survives the trough, usually rewarding; buying near the ceiling feels safe and usually isn't. That survival clause matters: not every cyclical comes back. An over-leveraged one can go bankrupt in the down years and never recover — the floor is only a floor if the company is still standing when the cycle turns. Tata Steel recovered; plenty of weaker commodity players have gone to zero from a similar-looking "cheap near the bottom" setup. So pair this margin check with the debt check below before you treat any trough as a bargain. This is the cyclical's cousin of the margin of safety — your protection is buying below mid-cycle worth, not below peak-year worth.

Did the company use the boom to get stronger? The best cyclical operators bank the upcycle cash to cut debt, so they survive the down years that always come. Tata Steel, on its results calls, has guided to keeping net debt-to-EBITDA below 3x through the cycle — a commitment about surviving the trough, not chasing the peak, and one its FY26 results suggest it has held to (Inve data, 2026). Whether a management says one thing on the boom-time call and quietly drops it when the cycle turns is exactly the kind of follow-through you can read on Inve's concall summaries instead of digging through years of transcripts yourself.

Are you inside your circle of competence? Lynch again: "It's perilous to invest in a cyclical without having a working knowledge of the industry." (Peter Lynch, Beating the Street). If you can't say where steel demand and prices sit in their cycle, you're not investing — you're guessing the weather. (More on knowing your edge: circle of competence.)

None of this is a view on Tata Steel as a buy or a sell. It's the lens. A great cyclical bought at the top can lose you money for years; an ordinary one bought near the bottom can do the opposite. The business is only half the question — the price relative to where you are in the cycle is the other half, and it's the half the peak-earnings trap is built to make you forget.

Frequently asked questions