When earnings fail you, what do P/B and P/S measure instead?
P/E stops working the moment earnings turn negative or lumpy. Two other price multiples pick up where it leaves off by dividing price against something steadier than profit: assets, or sales.
Price-to-book (P/B)
Book value is what the balance sheet says the company's net assets are worth — total assets minus total liabilities, the accountant's view of owner's equity. P/B compares market value to that figure.
- A P/B of 1 means the market values the company at exactly its accounting net worth.
- Below 1 means the market values it at less than its stated net assets — an observation that can mean anything from deep pessimism to assets the market doubts are worth their book figure.
P/B earns its keep for asset-heavy businesses — banks, insurers, real-estate and industrials — where the balance sheet genuinely reflects the value of the business. It is far less useful for an asset-light software firm whose real value (brand, code, people) barely appears on the balance sheet at all, which is why such companies routinely trade at a P/B of 10 or more without it meaning much.
Price-to-sales (P/S)
P/S divides market value by annual revenue. Its superpower is that revenue is almost always positive, so P/S keeps working when P/E can't — for loss-making growth companies, early-stage firms, or a cyclical business in a down year.
The trap is right there in the definition: sales are not profits. A company can grow revenue impressively while losing money on every sale. A low P/S looks cheap, but it says nothing about whether those sales ever convert to profit. P/S is only comparable within an industry, because a supermarket (huge sales, thin margins) and a software firm (smaller sales, fat margins) live in completely different P/S worlds.
Choosing between them
Think of it as matching the tool to the business. Steady profits → P/E. Asset-driven balance sheet → P/B. No profits yet, but real revenue → P/S. No single multiple is "best"; each measures price against a different fundamental, and each has a blind spot.
In the data
A bank first, because price-to-book is the multiple that was built for it:
Read the two labels closely. "MRQ" means book value as of the most recent reported quarter; "TTM" means the last twelve months of sales. The two multiples are therefore measured over different windows, one a snapshot and one a year, which is worth stating whenever they are quoted together. And a "book value" printed on a company profile is usually per share, not the equity total on the balance sheet.
Try it now
- Note the bank's price-to-book. For a business that essentially is a pile of financial assets and liabilities, book value is the natural anchor, and this is the figure the market actually quotes.
- Now the same multiple for an asset-light business:
Its price-to-book is a different order of magnitude, and that is not a verdict on either company. Brand, software and people barely appear on a balance sheet, so book value stopped describing the business. 3. Read the suffixes on both tables and say what window each of the two multiples is measured over. They are not the same window, and a sentence quoting them side by side without saying so has quietly compared a snapshot against a year. 4. Now a company growing its revenue with no profit yet:
Its trailing P/E reads 0 and its headline P/E is blank, while price-to-sales still reports. Read the quarterly revenue growth to see what the market is paying sales multiples for. Say which multiple fits which of the three businesses on this page and why — matching the tool to the company is the practitioner's habit.