What is a stock market correction?
A stock market correction is a commonly used name for a fall of at least 10% from a recent high in a broad market index. The phrase describes what has happened to prices. It does not tell you what happens next, and it is not a signal that a recovery is due or that a further fall is inevitable.
Corrections are part of the uneven path of markets. Prices reflect changing expectations about companies, interest rates, inflation, economic growth and risk. Those expectations can shift quickly. A broad index may fall even when many of its companies are doing well, because a few large companies have heavy weight. An individual share can fall sharply without the whole market meeting a correction definition.
The 10% convention is not a universal legal or scientific boundary. Index providers and market commentators use different reference points: an intraday high or closing high, a price index or a total-return index, and different broad markets. A fall of 9.9% and one of 10.1% are not fundamentally different events. The label is a shorthand that helps people discuss the scale of a decline.
How often do corrections happen?
There is no calendar. Corrections do not arrive on a dependable schedule, and a historical average cannot say when the next one will begin. To answer frequency carefully, a source would need to state the index, observation dates, price series, peak and trough rule, and whether overlapping declines count separately. Without those choices, a single number can mislead. We leave the frequency as a description rather than quote a number without a clearly matched public series.
A correction becomes a bear market under a common convention when a broad market index falls 20% or more from a recent peak. This 20% marker is also a convention, not a law of nature. A decline can pass through the 10% correction threshold on its way to 20%, so the labels describe different depths along the same path rather than mutually exclusive events.
The distinction can be useful when reading headlines, but it should not become a forecast. A correction can stop before a 20% decline, or continue beyond it. Neither label tells you whether prices will recover soon, how long the decline will last, or whether a particular investment will fall by the same amount. The broader index and an individual holding can take very different paths.
A worked example
Imagine a broad index starts at 100 points and later reaches 120. It then falls to 108. The decline from the peak is 12 points divided by 120, or 10%, so it meets the common correction threshold. If it later falls to 96, that is a 20% decline from 120 and meets the common bear-market convention. The round numbers are invented for illustration, not a description of a real index.
These percentages use the peak as the reference, not the original starting level. After a 20% fall from 120 to 96, a rise of 20% would take the index only to 115.2, still below its old peak. Percentage losses and gains compound on different bases. This is one reason a drawdown chart can add context that a headline label misses.
How to read the label
Check which market and dates the statement refers to. Is it the S&P 500, a UK index, a sector, or one company? Is the fall measured from a closing high or an intraday high? Does the source include dividends? Those details matter when comparing two reports. A price-only index and a total-return series can differ because one includes reinvested distributions.
Then look beyond the percentage. The time taken to fall, the time spent below the old high, and the path of recovery are different pieces of information. Maximum drawdown records the largest peak-to-trough decline in a chosen period, but does not by itself say how long recovery took. The companion note on maximum drawdown explains that measure and its limits.
A correction headline also does not describe the breadth of market conditions. Some industries may be rising while others fall. Investors sometimes focus on the index level because it is easy to summarise, but an index is a weighted basket, not every company or household portfolio. A diversified portfolio can behave differently depending on its holdings and weights.
Can a backtest tell you how you would feel during one?
A historical simulation can show the declines in its selected data and rules. It cannot recreate uncertainty, changing personal circumstances, taxes or every real-world cost. StrategyLens figures come from simulated backtests on past data. The runs charge no trading costs, and the buy-and-hold comparison is charged the same, under ADR 0033. Real trading costs money, so real results can be lower, especially for strategies that trade often.
The same caution applies to simple rules that leave the market after a threshold. Such a rule must specify whether it uses closing prices, when the signal is known, and when the next trade can occur. Small implementation choices can change a test. The note on backtesting explains why rules and timing need to be written down before interpreting results.
A correction is therefore best treated as a compact description of a decline already measured. It can help organise a conversation, but it cannot make uncertainty disappear. Avoid reading the label as a promise of a rebound, a reason to panic, or a complete account of risk. Read the underlying market, the period and the measure.
This is research, not financial advice. Past results, real or simulated, do not tell you what will happen next.
The 10% and 20% thresholds in this note are common market conventions, not official thresholds set by a regulator. FINRA’s investor education material describes a market correction as a drop of 10% or more from a recent peak and a bear market as a decline of 20% or more. The definitions are useful shorthand, but the underlying series and measurement dates still need to be checked.
What is a stock market correction?
A commonly used definition is a fall of at least 10% from a recent peak in a broad market index. It is a description of a fall already observed, not a forecast.
Do corrections come on a schedule?
No. There is no fixed schedule. Frequency depends on the index, dates and definition used; avoid treating a historical average as a timetable.
Is a correction the same as a bear market?
No. A correction is commonly described as a fall of 10%; a bear market is commonly associated with a fall of 20% or more, though definitions vary.