Normalised earnings: valuing a business through the cycle
Use an average across a full cycle, not the latest year. The latest year is the one number guaranteed to mislead you.
Value a company whose earnings swing wildly and predictably.
How Normalised earnings: valuing a business through the cycle works, in one picture
The same argument as the text, as a chain. Each step is what makes the next one possible.
The market moves first, and recovers first
Shares fall before the data does and start climbing while the news is still uniformly awful. Waiting for the news to improve means buying after the recovery has happened.
- 1
The P/E inverts on a cyclical, and this catches everyone
At the top of the cycle, earnings are at a peak, so the P/E looks tiny and the stock screens as cheap, moments before profits collapse. At the bottom, earnings are near zero and the P/E is enormous or meaningless, which is often when the shares are genuinely worth owning.
For a cyclical, a low P/E is a warning and a high one can be an opportunity. It is the most counter-intuitive rule in equity analysis.
- 2
So average the earnings across a whole cycle
Take profit across seven to ten years, spanning at least one peak and one trough, and use the average as your estimate of mid-cycle earning power. Then apply a sober multiple to that, not to this year's figure.
- 3
Adjust for the size the company is now
A company that has doubled its capacity since the last trough will earn more at the next peak. Averaging raw historical profit understates it. Normalise using margins on today's revenue base rather than yesterday's profits.
- 4
Then check the balance sheet, because that is what kills them
Cyclicals do not die of low prices. They die of debt taken on at the top of the cycle that has to be serviced at the bottom. Interest coverage at trough earnings, not current earnings, is the number that matters.
You can state a company's mid-cycle earnings, and its interest coverage at the trough.
Read next
Use an average across a full cycle, not the latest year. The latest year is the one number guaranteed to mislead you.
Every one shows its exact method, and the circumstances in which it is wrong. Free, and no account to look.
