Comparing against the sector: why absolute multiple levels are useless

The claim that "a P/E below such-and-such means cheap" does not work, because the normal level differs by industry by nature.

Why the levels differ

Fast-growing industries trade at a higher multiple of current profit: the market pays for the future.

Capital-intensive, slow-growing industries trade lower.

Banks are valued on capital rather than revenue, because their business is structured differently — Sectors of the Russian market: what it is made of.

Median rather than average

An industry average is easily distorted by one company with an extreme value — near-zero profit producing an enormous multiple, for instance. The median is more robust.

Its own history

A company's metric relative to its own average over several years is often more informative than a cross-industry comparison: it answers whether the market's attitude to this business has changed.

What the comparison does not give

The reason the gap appeared. Cheapness relative to the industry is the start of analysis, not the conclusion: The value approach: buying below what it is worth.

In practice

A screener lets you filter by multiple within an industry and compare against the median — The screener: searching for securities rather than confirming your view.