Normalised Earnings: Valuing Cyclical Companies

Normalised earnings estimate what a cyclical company earns in an average year across a full boom-and-bust cycle, rather than at a single peak or trough. You average margins or returns over five to ten years and apply them to current sales, so the valuation reflects mid-cycle reality instead of a misleading snapshot. Get this wrong and a cyclical stock will look cheap exactly when it is most dangerous.

Where this fits: normalising is a prerequisite for the “reasonable price” test in our Five Criteria framework, and it sits alongside what stock valuation is as a core skill for anyone valuing an economically sensitive business.

Why Cyclical Companies Break Ordinary Valuation

Most valuation shortcuts assume that this year’s earnings are a fair guide to next year’s. For a software company with recurring revenue or a consumer staple that sells the same soap in every economy, that assumption is roughly fine.

For a cyclical company it is dangerous.

A cyclical business earns feast-or-famine profits that swing with the economy, commodity prices or capital spending. Think miners, steelmakers, carmakers, chemical producers, homebuilders, airlines, shipping lines and semiconductor foundries. Their sales and margins can triple in a boom and vanish in a slump.

The trap is mechanical. At the top of the cycle, earnings are high, so the price-to-earnings ratio looks low. The stock appears cheap just as profits are about to fall. At the bottom, earnings collapse or turn negative, the P/E looks enormous or meaningless, and the stock appears expensive just as profits are about to recover.

This is the opposite of how a multiple is supposed to behave, and it is why Peter Lynch warned investors to treat cyclicals as a category of their own rather than lumping them in with steady growers.

What Normalised Earnings Actually Mean

Normalised earnings are an estimate of what a business earns in a typical, mid-cycle year. You strip out the distortion of where we happen to sit in the cycle and ask: across a full boom and bust, what does this company earn on average?

There are two layers to normalisation.

• One-off adjustments. Remove items that are genuinely non-recurring, such as a legal settlement, a restructuring charge, an asset write-down or a gain on a disposal. This applies to every company, cyclical or not, and overlaps with quality of earnings.
• Cycle adjustments. Smooth the natural boom-and-bust swing by averaging profitability over a full cycle. This is the part unique to cyclicals, and it is where most of the value of the exercise lies.

The output is a single number, “normalised” or “mid-cycle” earnings, that you then feed into a multiple or a discounted cash flow. It is an estimate, not a fact, so treat it with the humility any estimate deserves.

Three Ways to Normalise

There is no single correct method. The right one depends on what data you trust and how the business makes money. Here are the three most practical approaches.

1. Average the margin, apply to current sales

This is the cleanest method for most cyclicals.

1. Pull operating margin for each of the last seven to ten years.
2. Calculate the simple average. Make sure your window includes at least one downturn.
3. Apply that average margin to the current year’s revenue.
4. Tax the result to get a normalised net profit.

The logic is that revenue tends to grow over time with the business, but margin is the part that swings with the cycle. Applying a through-the-cycle margin to today’s larger revenue base gives a fairer picture than simply averaging old profit figures that were earned when the company was smaller.

2. Average return on capital, apply to the asset base

For capital-intensive businesses, normalising the return on invested capital can be more robust than margin.

Take the average return on invested capital across the cycle and apply it to the current invested capital base. This works well when a company has been building capacity, because it links earning power to the assets actually in the ground or on the factory floor.

3. Average the earnings themselves (with care)

The crudest version is to average reported earnings per share over a full cycle, as Benjamin Graham did with his ten-year method. The weakness is that it ignores growth in the business over the window, so it can understate the earning power of a company that has genuinely expanded. Use it as a sanity check rather than your primary number.

A Worked Example (Hypothetical)

Imagine a company, Cyclica plc, a hypothetical maker of industrial components whose profits swing with factory construction. Its reported numbers over a cycle look like this.

Year	Revenue	Operating margin	Operating profit
1 (slump)	£1,000m	2%	£20m
2	£1,100m	6%	£66m
3	£1,250m	11%	£138m
4 (peak)	£1,400m	16%	£224m
5	£1,300m	12%	£156m
6	£1,150m	7%	£81m
7 (current)	£1,300m	14%	£182m

In the current year, Cyclica earns £182m. If the market values it at £1,460m, the trailing P/E on operating profit is about 8x. That looks cheap.

Now normalise. The average operating margin across the seven years is about 9.7%. Apply that to current revenue of £1,300m and you get normalised operating profit of roughly £126m. On the same £1,460m valuation, the multiple on mid-cycle earnings is closer to 11.6x, not 8x.

The stock is not nearly as cheap as the headline suggested, because year 7 is a near-peak year. This is exactly the distortion normalisation is designed to expose, and it is a textbook case of the kind of valuation mistake beginners make.

Normalised P/E and the Shiller Approach

The best-known application of this idea is the cyclically adjusted price-to-earnings ratio, often called CAPE or the Shiller P/E, which averages inflation-adjusted earnings over ten years. Robert Shiller popularised it for valuing the whole US stock market, and the principle carries straight over to individual cyclicals.

The mechanics are simple. Where an ordinary P/E ratio divides price by the latest year’s earnings, a normalised P/E divides price by average earnings across a cycle. A high normalised P/E says you are paying a rich price for average earning power; a low one says you are not.

The table below shows how differently the two multiples can read for the same stock.

Measure	Peak year	Trough year
Reported earnings	High	Low or negative
Trailing P/E	Looks low (cheap)	Looks high (dear)
Normalised earnings	Steady	Steady
Normalised P/E	Reveals the real price	Reveals the real price

The whole point is that the normalised figure stays steady while the reported figure whipsaws. That stability is what lets you compare the valuation today with its own history.

Feeding Normalised Earnings Into a DCF

A discounted cash flow inherits whatever starting point you give it. Begin a cyclical DCF from peak free cash flow and every forecast year is built on sand.

For cyclicals, start from mid-cycle free cash flow instead. Average free cash flow across the cycle, or apply a normalised margin to current sales and work down to free cash flow from there. Our full method is in what discounted cash flow is, and you can run the numbers in the DCF intrinsic value calculator.

The same caution applies to the FCF yield leg of the price test. A 9% FCF yield on peak cash flow might be a 4% yield on normalised cash flow, which changes the conclusion entirely.

The Balance Sheet Comes First

Normalising tells you what a cyclical is worth. It does not tell you whether it will survive the next trough to realise that value. For cyclicals, survival is the first question, not the second.

A company with heavy debt can be wiped out in a downturn before the cycle turns, however attractive its mid-cycle earnings look. Before you value a cyclical, check that it can endure a bad year or two:

• Net debt to mid-cycle EBITDA comfortably within the Five Criteria threshold of 2x, not 2x on peak EBITDA.
• Interest coverage that holds up when profits halve.
• A sensible maturity profile on long-term debt, so nothing large comes due in the teeth of a slump.

Read the balance sheet with the trough in mind. The best cyclical operators run conservative balance sheets precisely so they can buy assets cheaply when weaker rivals are forced to sell. That discipline is often the real moat in a commodity business.

Frequently Asked Questions

What is the difference between normalised and reported earnings? Reported earnings are what a company actually booked in a given period. Normalised earnings adjust that figure to remove one-off items and, for cyclicals, to reflect a mid-cycle average rather than a peak or trough year. The goal is a number that represents sustainable, repeatable earning power.

How many years should I average to normalise earnings? Use at least one full cycle. For most industrials and commodities that means five to ten years, ideally spanning one downturn and one upturn. Averaging only good years defeats the purpose, so make sure your window includes a recession or a price trough.

Is a low P/E on a cyclical always a trap? Not always, but it is a warning. A single-digit P/E on a miner or carmaker at the top of the cycle usually reflects peak earnings that will fall, not a bargain. Recalculate the multiple on normalised earnings before concluding anything.

Can I use normalised earnings in a DCF? Yes, and for cyclicals you should. Start a discounted cash flow from mid-cycle free cash flow rather than the latest year, or your forecast will inherit the distortion. See our full DCF walkthrough for the method.

Your Next Step

Pick one cyclical company you follow and recalculate its P/E on a seven-to-ten-year average margin. If the normalised multiple looks very different from the trailing one, you have just avoided a common trap.

To go deeper, read when is a stock cheap? a practical framework and our guide to the largest risk to your wallet: overpaying for stocks. When you are ready to value one, use the DCF intrinsic value calculator with a mid-cycle starting point.

---

About the author: Cameron Hayes is a senior investor relations and finance professional who has worked in IR at Nasdaq Copenhagen-listed companies including Nilfisk and Maersk Drilling. He holds an MSc from Copenhagen Business School and a BCom from the University of Ottawa, and founded Gingernomics to teach individual investors the Five Criteria framework. More about Gingernomics.

The content on Gingernomics is for educational and informational purposes only and does not constitute financial advice. Always do your own research and consult a licensed financial advisor before making any investment decisions. Past performance is not indicative of future results.