You have read the company story. The products are excellent, customers keep coming back, and profits are growing. Then an analyst describes the shares as expensive, which feels rather like being told that an excellent restaurant is a disappointing place to have dinner. The missing detail is the bill: a wonderful meal can still leave you wondering whether the potatoes really needed their own finance agreement.

So, let's jump in.

A share is a small ownership stake in a business. When someone calls it expensive, they usually mean its price looks high relative to the financial benefits they expect that ownership to deliver. They are making a judgement about the price as well as the company.

One common measure is the price-to-earnings ratio, or P/E. It compares the share price with annual earnings per share: the profit attributable to each share.

Share price ÷ annual earnings per share = P/E ratio

A £20 share with annual earnings per share of £1 has a P/E of 20. Investors are paying £20 for each £1 of annual earnings, but that £1 is not necessarily paid out to them. Earnings and dividends are different things; a P/E of 20 is not a promise that you will get your money back in 20 years.

Meet Rather Good Biscuits plc, an entirely fictional company. It sells biscuits through supermarkets, earns a profit after its costs, and attracts repeat customers through generous chocolate coverage. For this example, assume its share count stays unchanged and its earnings figures are comparable.

Initially, its shares cost £20 and its earnings over the preceding 12 months are £1 per share. A year later, annual earnings per share have risen by 10% to £1.10. The business has delivered more profit, and the chocolate department has every reason to feel pleased.

However, investors have become less confident about future growth. Suppose the price they will pay now represents 15 times annual earnings, down from 20 times. Here is the result:

Profits per share rose 10%. The share price fell 17.5%.

Better biscuits. A smaller multiple.

EARNINGS PER SHARE

+10%

£1.00 → £1.10

The business earns more.

SHARE PRICE

−17.5%

£20.00 → £16.50

Investors pay a lower multiple.

P/E: 20 → 15
£1.10 earnings × 15 = £16.50 share price

Fictional example. Price change excludes dividends, dealing costs and tax.

These are invented figures illustrating a possible outcome, not a forecast. The price change excludes dividends, dealing costs and tax. The calculation is £1.10 × 15 = £16.50, a £3.50 fall from £20.

This is multiple compression: the “multiple” is the P/E, and investors are now paying a smaller multiple of earnings. The phrase sounds like a problem requiring an engineer. Here, it means the valuation has become less generous, outweighing the improvement in profits.

The important question is what the starting price already assumed. In our fictional example, imagine investors originally expected earnings per share to grow by 20%. Delivering 10% growth would then be both an improvement on last year and a disappointment against expectations.

That does not mean missing a forecast mechanically produces a particular price fall. Our move from a P/E of 20 to 15 is an assumption, not a formula for predicting investor behaviour. It shows how an improving business and a falling share price can coexist.

Nor does the example prove the original £20 price was wrong. If the company subsequently grew much faster than expected, that price might look reasonable with hindsight. One disappointing year does not settle the long-term argument.

Investors may pay more for each pound of current earnings when they expect stronger growth or have greater confidence in the business's prospects. Rather Good Biscuits might win more shelf space, retain loyal customers and turn higher sales into higher profits. A premium can have a sensible explanation.

But the size of that premium still matters. Rising cocoa costs or cheaper competitors might undermine the story, and customer loyalty has limits when the supermarket's own biscuits are sitting six inches away. Equally, a low P/E might signal trouble ahead rather than an overlooked bargain.

The comparison might be with similar businesses, the company's own history, or an analyst's estimate of its worth. None is a universal price list. Businesses differ, and an old valuation is not automatically the right one today.

Check the earnings period, too. Trailing P/E uses the previous 12 months' earnings; forward P/E uses forecast earnings for a specified future period. Those forecasts can change, so two commentators may be using different numbers rather than disagreeing about the same calculation.

P/E also has blind spots. A one-off gain can make earnings look unusually healthy, while a loss-making business generally has no useful P/E. Debt and the durability of profits need attention beyond the headline ratio.

When the next analyst calls a share expensive, ask what earnings they expect, over what period, and why those earnings deserve their chosen valuation. Then ask what could make that argument wrong. The biscuits can be excellent; the bill still deserves a look.

  • Business quality and share valuation answer different questions. Admiring a company does not establish what its shares are worth.
  • Expectations matter. Rising earnings can coincide with a falling share price if investors pay a lower multiple.
  • P/E starts the investigation. Check the earnings period, the comparison and the assumptions behind the number.