When prices rise quickly, someone always asks whether it is a bubble. Often the honest answer is: not necessarily. Prices can rise strongly for sound reasons. A share market recovering from a crash, a housing market catching up after years of weak growth, or a commodity responding to a genuine shortage can all show rapid gains without being in a bubble.
The more useful question is not how fast prices are rising, but how high they are compared with the things that ultimately support them. For shares, that is the earnings and cash flow of the businesses behind them. For housing, it is the rents homes could earn and the incomes of the people who buy them. These comparisons are called valuation measures, and they provide the best single clue to whether a market has moved into bubble territory.
This article explains the main valuation measures, how to read them against history and against reason, why they can stay stretched for a long time, and how the same thinking applies to business decisions. It is part of GoCore’s series on bubbles; the full set of warning signs is covered in Warning signs of a bubble.
Why price alone is misleading
Consider two markets that have each risen 50% in three years.
The first had fallen 40% in a recession before the rise began. After the rise, prices are only slightly above where they started before the recession, and earnings have recovered strongly. Valuations are near their long-run average.
The second had risen steadily for a decade before the latest surge. Earnings have grown modestly. After the rise, prices are far higher relative to earnings than they have ever been.
The headline gain is identical. The risk is completely different. Only a valuation measure reveals the difference.
Valuation measures for shares
The price-earnings ratio
The most widely used measure for shares is the price-earnings ratio, or P/E: the share price divided by earnings per share, or for a whole market, total market value divided by total earnings. A P/E of 15 means investors are paying $15 for every $1 of annual earnings.
Over long periods, the P/E of broad share markets has tended to move around a long-run average. For the main US share index, that average has historically been roughly 14 to 16 times earnings, depending on the period and method. During the late-1990s technology bubble, the ratio on operating earnings (which excludes one-off items) rose above 30, roughly double the long-run norm.
A high P/E is not proof of a bubble. It may reflect low interest rates, strong expected growth or temporarily depressed earnings. But a P/E far above its long-run average, sustained for some time, is one of the clearest signs that prices depend on very optimistic expectations.
Which earnings?
The P/E depends heavily on which earnings are used:
- Reported earnings follow accounting standards and include one-off items such as write-downs.
- Operating earnings exclude one-off items, giving a cleaner view of the underlying business, but can be flattered by generous exclusions.
- Forecast earnings are analysts’ estimates for the coming year. They tend to be optimistic in booms.
Comparisons should always use the same type of earnings consistently over time.
The cyclically adjusted P/E
Earnings swing with the economic cycle. At the top of a boom, earnings are high, which makes the P/E look lower than it really is. In a recession, earnings collapse, which makes the P/E look misleadingly high.
To smooth this out, the economist Robert Shiller popularised the cyclically adjusted price-earnings ratio, often called CAPE: price divided by the average of the previous ten years of earnings, adjusted for inflation. Because it averages across a full cycle, it gives a steadier view of valuation. Shiller’s book Irrational Exuberance, published in 2000 near the peak of the technology boom, used this measure to argue that US shares were unusually expensive.
Other measures
Analysts also use:
- dividend yield: annual dividends divided by price; low yields suggest high prices
- price to book value: market value compared with the equity shown on balance sheets
- market value to the size of the economy: total share market value compared with national income
Each has weaknesses, and none should be relied on alone. When several point the same way, the evidence is stronger.
When there are no earnings
New technology companies often have no earnings at all, which makes the P/E useless. Investors then turn to other measures, such as price compared with sales, or with numbers of users or subscribers.
These measures can be reasonable for young businesses that are deliberately investing for growth. But they also give speculation more room, because they say nothing about whether the business will ever earn a profit. During the late-1990s boom, some companies were valued on website visitors, sometimes called “eyeballs”, with little attention to how those visitors would produce revenue. When a market shifts towards valuing companies on measures increasingly distant from profit, it is worth asking why.
Valuation measures for housing
Price-to-income ratio
For housing, a common measure is the price-to-income ratio: the typical home price divided by typical household income. It indicates how many years of income a home costs. When the ratio rises far above its historical level, buyers must borrow more relative to their incomes, which makes the market more dependent on low interest rates and easy credit.
Rental yield
The rental yield is the annual rent a property could earn divided by its price. It works like an earnings yield for housing. When prices rise much faster than rents, yields fall. A yield well below the interest rate on a mortgage means that owning a property as an investment depends heavily on future price gains rather than on the income it produces.
Price-to-rent ratio
The inverse of the rental yield, the price-to-rent ratio, compares the cost of buying with the cost of renting a similar home. When buying becomes far more expensive than renting, and remains so, it suggests that buyers expect continued price growth to make up the difference.
Housing is local
Housing markets are local, and national averages hide large differences. A national price-to-income ratio may conceal one city that is extremely stretched and another that is modest. Valuation analysis for housing works best for specific cities or regions.
Two benchmarks: history and reason
Valuations can be judged against two different benchmarks.
Compared with history
The first compares current valuations with their long-run average. This rests on the idea of mean reversion: over long periods, valuation measures have tended to return towards their averages, either because prices fall or because earnings and incomes catch up.
Historical comparisons are powerful because they draw on real experience across many conditions. Their weakness is that the world changes. If interest rates, tax rules or the structure of the economy have changed permanently, the old average may no longer be the right anchor.
Compared with reasonable levels
The second benchmark asks what valuation is reasonable given current conditions. If long-term interest rates are much lower than in the past, a higher P/E or a lower rental yield may be justified, because the alternative investments also yield less.
This is a legitimate argument, and it is often correct up to a point. But it can also be stretched. During bubbles, the reasoning shifts from “valuations can be somewhat higher” to “valuations no longer matter at all”. The test is whether the case for higher valuations rests on specific, measurable changes, and whether those changes can justify the full extent of the rise.
Starting from undervalued levels
A strong rise in prices from undervalued levels is not a bubble. After the share market crashes of 1974, 1987, 2002 and 2008–09, prices rose rapidly in the following years. Much of that rise was a recovery from levels that had been unusually low relative to earnings.
This is why the extent of overvaluation, rather than the speed of the rise, gives the best clue to the probability of a bubble. A market that has doubled from deeply undervalued levels may be fairly priced. A market that has risen 20% from already very high valuations may be in more danger.
Why valuations can stay stretched
One of the most frustrating features of valuation measures is that they say little about timing. Markets can stay overvalued for years.
There are several reasons:
- Momentum. Rising prices attract buyers, and buyers push prices higher.
- Easy credit. As long as lenders keep lending, buyers can keep paying high prices.
- Normalisation. After prices have been high for a while, people begin to see those levels as normal. The judgement that a price is “permanent” tends to be based on how long it has held, rather than whether it makes sense.
- Incentives. Professional investors who avoid an overvalued market may underperform for long enough to lose clients or jobs.
Valuation tells you about risk and long-term returns. It does not tell you when prices will turn.
Valuation and long-term returns
What valuation does predict, with reasonable reliability, is long-term returns. Historically, buying shares when valuations were very high has tended to produce poor returns over the following decade, while buying when they were low has tended to produce good ones. This relationship is weak over one or two years, because prices can move a long way in either direction, but it has been much more consistent over ten.
Applying valuation thinking to business decisions
The same discipline is valuable well outside financial markets.
Buying a business. The price of a small business is often expressed as a multiple of its annual profit. During periods of enthusiasm for a particular sector, those multiples can rise well above their usual levels. Comparing the asking multiple with what similar businesses have sold for over many years, and with what the business’s earnings can realistically support, helps avoid paying a bubble price.
Buying property for the business. A rental yield calculation shows whether buying premises makes more sense than leasing. If yields are very low, leasing may be the better choice, preserving capital for the business itself.
Equipment and inputs. Prices for second-hand equipment, materials and specialist labour can spike during booms in particular industries. Comparing current prices with longer-term levels can reveal when it is better to wait.
Valuing your own business. Owners considering selling, or raising capital, can be tempted by boom-time valuations. Understanding that these may not last helps in making decisions about timing and structure.
A worked illustration
This is an illustration with round numbers, not data about any real market.
An investor compares two share markets.
| Market A | Market B | |
|---|---|---|
| Rise over the past three years | 50% | 25% |
| Current P/E | 15 | 32 |
| Long-run average P/E | 15 | 15 |
| Earnings growth over three years | 45% | 10% |
Market A’s rise has been driven almost entirely by earnings growth. Its valuation is at its long-run average. Market B has risen less, but its valuation has more than doubled relative to its history, with modest earnings growth.
Despite the smaller rise, Market B carries far more valuation risk. If its P/E returned to its long-run average while earnings stayed the same, prices would fall by more than half.
Common mistakes
Judging by the speed of the rise. Valuation matters more than momentum.
Using a single measure. Several measures pointing the same way give stronger evidence.
Mixing types of earnings. Compare like with like.
Assuming high valuations must fall soon. They can stay high for years.
Accepting every argument that valuations no longer matter. Specific, measurable changes can justify somewhat higher valuations; they rarely justify any valuation at all.
Questions to ask
- How do current valuations compare with their long-run averages?
- Are higher valuations justified by specific changes, such as lower interest rates, and to what extent?
- Has the rise come from growing earnings or incomes, or from rising multiples?
- Did the rise start from undervalued, fair or already high levels?
- For your own business: are you paying, or relying on, a valuation multiple that is unusually high by historical standards?
Bringing it together
Rapidly rising prices raise the question of a bubble, but they do not answer it. Valuation measures, such as price-earnings ratios for shares and price-to-income ratios and rental yields for housing, compare prices with what ultimately supports them. Judged against history and against reasonable levels, they provide the best clue to how stretched a market has become.
Valuation does not reveal timing. Markets can stay expensive for a long time. But it reliably indicates risk and long-term returns, and the same discipline helps business owners avoid paying bubble prices for businesses, property and equipment.
Sources: an introductory chapter on asset bubbles written in the mid-2000s, Robert Shiller’s Irrational Exuberance, and widely documented market history. Figures in the worked illustration are not data. This article is general information, not financial or investment advice.
