Valuing a Company on Earnings It Is Not Making Right Now
Cyclical companies earn a lot at the peak and little at the trough, so their current earnings mislead. Normalized earnings estimate what they make in an average year, which is what a valuation should use.
The Trap in Current Earnings
Valuing a company often starts with its earnings, and for a stable business that works. For a cyclical company, whose profits swing dramatically with an economic or industry cycle, current earnings can be deeply misleading, because they reflect where the company sits in its cycle rather than what it earns on average.
Normalized earnings are an estimate of what a company earns in a typical, mid cycle year, stripping out the distortion of being at a peak or a trough. Using them, rather than current earnings, is essential for valuing any business whose profits are cyclical.
A cyclical company looks cheapest exactly when its earnings are about to fall and most expensive when they are about to rise. Current earnings point you the wrong way at both extremes.
Why Cyclical Earnings Deceive
The problem is the interaction of the cycle with the price to earnings ratio. At the peak of a cycle, a cyclical company earns unusually high profits, so its price to earnings ratio looks low, appearing cheap. But those peak earnings are unsustainable, and as the cycle turns down, earnings collapse and the stock falls.
| Cycle position | Current earnings | P/E appears | Reality |
|---|---|---|---|
| Peak | High | Low, looks cheap | Expensive, earnings will fall |
| Trough | Low | High, looks expensive | Cheap, earnings will recover |
At the trough the reverse happens: earnings are depressed, the ratio looks high and expensive, but the company is actually cheap because earnings will recover. The naive use of current earnings makes a cyclical company look cheap at the top and expensive at the bottom, which is exactly backwards.
How Normalization Works
Normalizing earnings means estimating what the company would earn in an average year across the cycle, not at any particular point. Several approaches exist.
One averages the company earnings over a full cycle, smoothing the peaks and troughs into a representative figure. Another applies a normal, mid cycle profit margin to current revenue, since margins swing more than sales over a cycle. A third estimates mid cycle volumes and prices for a commodity producer and works out the earnings those would generate.
The goal in each case is the same: a figure representing sustainable, through the cycle earning power, which is what a long term valuation should rest on rather than a single year that happens to be a peak or a trough.
The Cyclically Adjusted Ratio
The same logic applies to whole markets. A well known measure, the cyclically adjusted price to earnings ratio, values a market against its average inflation adjusted earnings over many years rather than the current year, precisely to avoid being misled by where the cycle happens to be.
This averaging is meant to reveal whether a market is expensive or cheap relative to its sustainable earnings, filtering out the distortion of a boom or a bust year. It rests on the same principle as normalizing a single company earnings: judge value against average earning power, not against a possibly unrepresentative single year.
The Judgement Involved
Normalization is not mechanical, and that is its weakness. Deciding what constitutes a normal year requires judgement about the cycle, and reasonable people disagree. Is the current period a normal environment or an unusually strong or weak one? What margin is sustainable? These questions have no certain answers.
The danger is that normalization can be manipulated to produce a desired valuation, by choosing a favourable definition of normal. An optimist normalizes to a high mid cycle figure and finds the company cheap; a pessimist does the opposite. The discipline is to base the normal figure on a genuine long run history rather than on a convenient assumption, and to be honest about the uncertainty.
When Not to Normalize
Normalization assumes the cycle will repeat and the company will return to its historical average earnings. This can be wrong if the business has structurally changed, if a cyclical industry is in permanent decline, or if past averages no longer apply.
Applying a historical mid cycle margin to a company whose industry has been permanently impaired overstates its earning power and its value. Normalization must be checked against whether the past is still a reasonable guide to the future, since averaging historical earnings for a business that will never see those conditions again produces a false picture of value.
The Bottom Line
Cyclical companies earn far more at the peak than the trough, so their current earnings mislead, making them look cheap at the top and expensive at the bottom. Normalized earnings estimate sustainable mid cycle earning power, by averaging over a cycle or applying normal margins, which is what a valuation should use. The approach requires honest judgement about what normal means, is vulnerable to manipulation toward a desired answer, and fails if the business has structurally changed so that its past is no longer a guide to its future.