How to Value a Cyclical Company (and Why P/E Betrays You)

A cyclical company is dangerous to value on a simple P/E because the ratio moves in the opposite direction to the business. At the top of the cycle earnings are swollen and the P/E looks temptingly low, while at the bottom earnings are crushed and the P/E looks alarmingly high, which is why professionals lean on normalised earnings, price-to-book and cross-cycle EV/EBITDA instead.

If you take one idea from this piece, take that inversion. For steel, cement, base metals, autos and other economically sensitive businesses, a low P/E is often a warning rather than a bargain, and a high P/E is often noise rather than a red flag. The trailing earnings number in the denominator is the problem, not the price.

Why a single year of earnings lies

Most valuation shortcuts assume that this year’s profit is a fair guide to next year’s. For a stable consumer-staples business, that assumption is roughly fine. For a cyclical, it is close to worthless.

Commodity and heavy-industry profits are geared to two things that swing hard: volumes and margins. When demand runs hot, factories run near full capacity, prices firm up, and operating leverage does the rest, so profits can multiply. When demand cools, the same fixed-cost base works against the company, and profits can fall much faster than revenue. A business whose sales move by a modest amount can see profits move by a large multiple of that.

So the earnings line, the thing sitting under the P in P/E, is the single most unreliable input you could choose for these companies.

The trap, shown plainly

Here is the pattern that catches people out. Imagine a metals company earning far above its long-run average at a cyclical peak, and far below it at a trough. The share price, set by a forward-looking market, usually does not fall as much as peak earnings, nor rise as much as trough losses.

The numbers below are illustrative, chosen to show the mechanics rather than to describe any real company.

Cycle stage	Earnings vs mid-cycle	Share price	Reported P/E	What it feels like	What is often true
Peak	Well above average	Elevated	Low	”Cheap, look at that P/E”	Earnings near a top
Mid-cycle	Around average	Fair	Average	”Nothing special”	The most honest reading
Trough	Well below average	Depressed	High or not meaningful	”Expensive, avoid”	Earnings near a bottom

Read the two right-hand columns together. The emotion the ratio triggers is the reverse of where the business actually sits. That is the whole trap in one line.

A low P/E on a cyclical at the top of its cycle is not a discount. It is the market telling you the E in the denominator is unsustainable.

None of this is a comment on whether any stock is worth owning. It is a comment on why one number, used alone, points the wrong way.

Tool one: normalised or mid-cycle earnings

The cleanest fix is to stop using a single year. Instead, estimate what the company earns through an average year across a full cycle. In practice, analysts do this a few ways:

• Average the history. Take earnings or margins across several years that span at least one full up-and-down cycle, then work with that average rather than the latest print.
• Apply a mid-cycle margin. Estimate a normal operating margin for the business, apply it to a reasonable through-cycle revenue level, and derive a normalised profit from there.
• Sanity-check against capacity. Ask what the company earns at a sensible utilisation level, not at a boom-time sprint or a slump-time crawl.

Once you have a normalised earnings figure, a “normalised P/E” becomes far more informative than the reported one. It answers a better question: not “what is it earning right now,” but “what does it earn in a typical year, and what am I paying for that.”

The honest caveat: normalising is judgement, not arithmetic. Cycles differ in length and depth, and a structural shift, say a lasting change in demand or a permanent cost advantage, can make the past a poor guide. Treat a normalised number as a considered estimate, not a precise fact.

Tool two: price-to-book, the asset anchor

When earnings are unreliable, the balance sheet becomes useful. Price-to-book compares the market value of equity to its accounting net worth, and for asset-heavy cyclicals it tends to swing far less violently than earnings.

The logic is intuitive. A steel plant or a cement kiln does not vanish when the cycle turns; the assets are still there. Book value gives you an anchor that does not collapse the way a single year of profit can. Many investors watch how a cyclical trades relative to its own history of price-to-book, treating stretches well above or well below that range as information about sentiment rather than as a signal to act.

Two health warnings. First, book value can be flattered or distorted by old asset values, write-downs and accounting choices, so it is an anchor, not gospel. Second, price-to-book is most meaningful for genuinely asset-heavy businesses; for an asset-light cyclical it tells you much less.

Tool three: EV/EBITDA across the whole cycle

Enterprise value to EBITDA is popular for cyclicals because it steps around two things that muddy the P/E: differences in debt loads and differences in depreciation and tax. Capital-intensive cyclicals often carry meaningful debt, and EV captures that where market cap alone does not.

The discipline that matters is the phrase “across the whole cycle.” A single year’s EV/EBITDA falls into exactly the same trap as the P/E: it looks low on peak EBITDA and high on trough EBITDA. The professional habit is to look at the multiple on normalised EBITDA, or to study the range it has traded through over many years, rather than fixating on today’s snapshot.

Read the cycle, not just the multiple

Every tool above still needs one input no ratio can give you: a view on where demand and margins sit right now. That is the real work.

Useful questions, kept general:

• Demand. Is end-demand near a high or a low? For an auto maker, that might be about the broad replacement and upgrade cycle; for cement, about the construction and housing backdrop; for a base metal, about the global picture, since prices are set worldwide.
• Capacity and supply. Is the industry adding capacity into strength, which often sows the seeds of the next downturn, or has under-investment tightened supply?
• Margins versus history. Are current operating margins unusually fat or unusually thin compared with the company’s own long-run record?
• Balance sheet. Can the company comfortably carry its debt through a lean stretch? Cyclicals with heavy leverage face a very different downturn from those without.

Answer those, and the multiples start to mean something. Skip them, and even a normalised P/E is just a tidier version of the same guess.

Tools like Altys exist to make this kind of cross-cycle, multi-year context easier to assemble, but the method matters far more than any single screen.

The practical takeaway

For a cyclical, never let a single year of earnings decide the story. Treat the reported P/E as a mood ring, not a valuation, and remember that it glows “cheap” at the top and “expensive” at the bottom.

Do three things instead. Normalise the earnings across a full cycle before you compute any multiple. Cross-check with price-to-book for asset-heavy names and with EV/EBITDA over many years, never a single one. And form a plain-language view on where demand, capacity and margins actually sit today. The number is easy; the cycle is the hard part, and it is the part that counts.

Altys Labs is not a SEBI-registered Research Analyst or Investment Adviser. This article is educational and general in nature. It is not investment advice, and it makes no recommendation about any security. Sectors and companies are named only as neutral illustrations, and any figures are rounded or clearly illustrative. Do your own research and consult a registered adviser before investing.