Economists and traders often split economic data into two broad types. Hard data measures what has actually happened, such as how much was produced, sold or earned. Soft data captures what people and businesses say they feel or expect, usually through surveys. Knowing the difference helps you judge how much weight to place on each release, and why markets sometimes react strongly to one and shrug off the other.
What counts as hard data
Hard data records real activity. Examples include gross domestic product, industrial production, retail sales, payrolls, trade figures and inflation. These numbers come from official statistics agencies or detailed records of transactions and output.
The strength of hard data is that it reflects reality rather than opinion. The weakness is timing. It is often published weeks or months after the period it covers, and it can be revised significantly later. GDP is a good example: early estimates can change as more information arrives. Our guide to China GDP shows how traders handle one of the most closely watched hard releases.
What counts as survey data
The survey side includes business and consumer surveys. Purchasing managers’ indices, known as PMIs, ask companies whether orders, output and employment are rising or falling. Consumer confidence surveys ask households how they feel about their finances and the economy. Sentiment surveys from groups such as the University of Michigan and the Conference Board are widely followed in the US.
These surveys are quick. Many are released within days of the month ending, and some, such as flash PMIs, arrive before the month is even over. That speed makes them valuable as early signals.
Why the two can disagree
Survey and activity data do not always line up. People may say they feel worse about the economy while continuing to spend. Businesses may report gloom in surveys while their actual output holds up. Political events, news headlines and inflation worries can sway how people answer surveys without immediately changing their behaviour.
There are also periods when surveys turn first and hard data follows later. In those cases, surveys act as a useful early warning. The challenge for traders is that it is often unclear, in real time, which situation they are in.
How markets react
Markets trade on surprise relative to expectations, so both types can move prices. Survey data can cause sharp moves because it arrives early and shapes the story before hard data confirms or contradicts it. Hard data often carries more weight with central banks, which want evidence that conditions have truly changed before adjusting policy.
When the two diverge for long periods, attention turns to which side is right. If survey data stays weak while activity holds up, investors may eventually decide the surveys were overly pessimistic. If hard data starts to roll over, the surveys may be vindicated.
Applying the idea to company results
The same principle applies to company news. Management commentary about the outlook is a kind of survey signal, while reported sales and profits are hard data. When you read an earnings report, compare what executives say with what the numbers actually show.
Using both sensibly
Look at trends rather than single releases. One weak survey may be noise, but several months of weakness carry more weight. Compare survey readings with related hard figures to see whether they are moving in the same direction. Be aware of revisions to hard data, which can change the story after the fact.
Most importantly, avoid treating either type as the whole picture. Surveys offer speed; hard figures offer substance. Combining them gives a more balanced view of the economy.
Bringing it together
Hard data measures real economic activity but arrives later and can be revised. Survey data captures sentiment and expectations quickly but does not always match behaviour. Traders use survey readings for early signals and hard figures for confirmation, watching closely when the two disagree.
If you want to understand how you react to economic news and risk, our free trader assessment is a helpful starting point.
