How do you tell a cheap company from an expensive one?

Your intrinsic-value course valued a company from the inside — its own cash flows. This unit adds the other lens: relative valuation, judging a company by comparing its price multiples to those of its peers. It's faster than a full valuation and it's how the market actually talks about "cheap" and "expensive" day to day.

Multiples are price-per-unit-of-something

A multiple puts price over a fundamental so two companies of different sizes become comparable. The workhorses:

• P/E — price relative to yearly earnings. A P/E of 20 means you pay 20 for each 1 of annual profit. The most quoted multiple in the world.
• P/S — price relative to sales. Useful when a company has little or no profit yet (young growth firms).
• P/B — price relative to book value (net assets). Central for asset-heavy businesses like banks.
• EV/EBITDA — enterprise value relative to a rough cash-earnings measure. A favourite because it looks through differences in debt and tax.

Each divides a valuation — the equity price for the first three, enterprise value for EV/EBITDA — by "something the business produces or owns." Higher multiple = you're paying more per unit of that something.

Cheap and expensive are relative words

A multiple alone is meaningless — the entire method is comparison. A P/E of 25 is expensive for a bank and cheap for a fast-growing software firm. So you always compare a company's multiple to a reference:

• its peers in the same industry,
• the sector average,
• and its own history.

Trading below your peer group might mean cheap. Trading above might mean expensive. "Might," because a gap always demands the next lesson's question: cheap for a reason, or genuinely mispriced?

A worked example

Five regional banks, ranked by P/E: 9, 10, 11, 12, and one outlier at 6. On a purely relative read, the bank at 6 looks cheap versus a peer group clustered around 10-11. That's a finding, not a verdict — it's the flag that says "look here and ask why." Maybe the market is right (this bank has bad loans); maybe it's overlooked. Relative valuation's job is to surface the gap efficiently; explaining the gap is the analyst's job.

In the data

A stock screener gives you a peer group's line-up in one table. Here are US-listed diversified banks above $10 billion, largest first, with each one's price and trailing earnings per share:

What a screener line-up usually does not carry is the multiples themselves, so a P/E per peer is price divided by earnings per share, worked out by you. And every row is the latest day, so the ranking you build is today's rather than a stated date's.

Try it now

1. In the table above, count how many of the figures you would need for a valuation comparison are actually there. Then count how many distinct banks there are, not rows.
2. Build the missing column. A P/E per row is the adjusted close ÷ earnings per share: six divisions. Check one against the published figure for that bank:

For a bank, record the price-to-book alongside the P/E: it is the multiple that industry is actually read on. 3. Rank the group. Who is the most expensive, who is the cheapest, relative to this group? 4. Circle the biggest outlier in either direction. You have just done the fast version of what analysts do all day: find the odd one out, then go ask why it is odd. Before you file the ranking, write today's date on it, and the section above says why that is not a formality.