What Is a Market Correction? A Simple Guide (2026)

A market correction is a drop of 10% to 20% in the price of an asset, index, or the overall stock market from its most recent peak. It is a routine pause in a rising market rather than a collapse, and the word most investors use for it says it plainly: prices are being corrected back toward their longer-term trend.

That definition gets you the headline answer fast. The parts people usually miss come next: how the 10% line gets measured, why a correction can quietly turn into something worse, and what a correction looks like when the asset in question is gold or copper rather than the S&P 500. Statistics in this guide are current as of October 2026.

None of this is individual financial advice. Every investor’s situation is different, and past market moves say nothing reliable about future returns.

What Is a Market Correction?

A market correction is a decline of roughly 10% or more from a recent high, usually measured on closing prices rather than intraday swings. It happens most often after a market has climbed quickly, when prices have moved further and faster than the underlying businesses justify.

The distinction that matters is between a correction and an ordinary bad day. Individual stocks fall 3% to 5% in a session all the time, and indexes do it too. That noise is not a correction. The 10% line exists because anything smaller is statistically ordinary and says very little about the market’s direction.

Charles Schwab describes the term as a fairly neutral one, and the etymology is worth knowing. A correction is what happens when a price that drifted too far above its long-term average gets pulled back toward it. On a chart it usually looks like a market that went up in a straight line, stalled, slid for a few weeks or months, and then either resumed or did not.

One more piece of context, and it is the one that trips up most beginners: a correction is identified after the fact. Nobody rings a bell when it starts. It only becomes a correction once the decline has already happened, and it is only officially finished once the index prints a new high.

How Is a Correction Different From a Bear Market?

A correction is a decline of about 10% to 20% from a recent peak. A bear market is a decline of 20% or more. That is the whole difference, and it is a convention rather than a law.

Two things muddy it in practice. First, a move can travel well past the 10% mark and still get described as a correction rather than a bear market, because the label depends on where the index eventually bottoms. A decline that stretches to 18% and recovers was a correction the whole time, even though it looked terrifying in the middle.

Second, the thresholds shift depending on what you are looking at. A 10% drop in gold is a serious event. The same percentage move in a single speculative stock may be an ordinary Tuesday. The Dow, the S&P 500 and the Nasdaq also behave differently from one another, so “the market is in a correction” always means a correction in some particular index.

Both labels describe the past, not the future. A correction is not a promise that a bear market is coming, and a bear market is not the end of the world, though it is genuinely uncomfortable for anyone drawing income from a portfolio.

How Much Can a Market Correction Fall?

A correction can land anywhere from just over 10% to around 20%. Deeper than 20% and the label changes, and deeper than 30% is usually described as a crash. The rough categories below are the ones you will hear in headlines.

Label	Approximate decline from recent peak	What it usually looks like	Typical shape of the recovery
Pullback	Roughly 3% to 9%	A few sharp sessions that give most of the move back within days or weeks	Quick, often complete within weeks
Correction	Roughly 10% to 20%	Weeks to several months of choppy, lower prices with frequent sharp rallies	Usually months, frequently within a year
Bear market	20% or more	A sustained grind lower, interrupted by rallies that keep failing at old resistance	Long: an average bear market has run about 14 months since 1966
Crash	30% or more, often in days	Gaps, limit moves, panic volume, forced selling across every asset class at once	Fastest when it comes from a credit or leverage problem

The 2020 selloff is the useful example of why a crash and a bear market are different labels for different things. The S&P 500 went from its February 19, 2020 peak to its March 23, 2020 low in 33 days, a fall of about 34%. Nobody would call that a slow grind. It was violent, brief, and unusually fast by historical standards.

Watch how headlines use these words loosely. A single down day of 4% is called a selloff. A 10% two-week slide is a correction. The precise label matters less than the number behind it, so whenever you read a scary headline, look for the percentage and the index first.

What Causes a Market Correction?

There is no single trigger, but corrections tend to fall into a handful of recognizable patterns. Knowing which one you are looking at tells you a lot about what to watch next.

Rate-driven corrections

When central banks raise interest rates, the value of future earnings drops because future cash flows are worth less today. Bonds sell off, and equities repriced against those higher yields follow. The inflation-fighting tightening cycle of 2022 is the clearest recent example, and it drove a full bear market rather than a tidy correction.

Earnings-driven corrections

When companies report results that fall short of what prices already assumed, the gap closes violently. A single quarter can erase several months of gains if guidance is cut as well as earnings missed. This is why earnings seasons produce a surprising share of corrections.

Valuation-driven corrections

This is simply prices outrunning the businesses. When a market has climbed for a long stretch on optimism, the price-to-earnings ratios behind it get stretched, and any disappointment has an outsized effect. The AI-driven valuation debate that ran through 2026 was discussed by central bank economists as a plausible trigger for exactly this kind of correction, which tells you how much weight the argument carries on both sides.

Event-driven corrections

Geopolitical shocks, surprise policy announcements, and sudden trade or tariff headlines can knock a market off its trend without any change in the underlying economy. The April 2025 tariff announcement, which erased large gains in a matter of days, was the clearest recent instance. These pullbacks usually resolve fast because there is nothing structural behind them.

Liquidity-driven corrections

When heavily borrowed positions unwind, sellers become forced sellers. The August 2024 move, where a Bank of Japan rate increase helped trigger a carry-trade unwind and coincided with the Nikkei’s largest one-day drop in roughly 40 years, is the textbook case. Nothing about the companies changed; the flow of money did.

Profit-taking sits across all five of these. Investors who bought early in a strong run bank some gains, and that selling pressure alone can tip an index into correction territory without any news at all.

Is a correction the same as a recession?

No, and the two are frequently confused. A correction is a price event that usually lasts weeks. A recession is an economic contraction, formally defined as two consecutive quarters of declining gross domestic product in the US, and it usually arrives with rising unemployment.

They overlap often enough that neither reliably predicts the other. Corrections can happen in a healthy economy, and a recession can begin without any correction at all, because the market has already priced the slowdown. What you can say is that a correction driven by falling earnings or rising defaults deserves more attention than one driven by a single headline.

Which Markets Are Most Vulnerable?

The same 10% rule applies everywhere, but the percentage that counts as ordinary depends entirely on the asset. A 10% drop in a currency is routine. The same drop in a leveraged commodity position can be a disaster.

Equities

Broad stock indexes are the reference point for the whole conversation. Within them, richly valued growth shares and anything priced on distant expectations tend to fall hardest, because their valuations depend most heavily on rates remaining low and earnings arriving on schedule.

Bonds

Bond corrections work in reverse to stocks. Rising inflation expectations push yields up and prices down, and long-duration bonds move more than short ones. A stock correction and a bond correction can happen at the same time for opposite reasons, which is exactly why a balanced portfolio behaves better than either part alone.

Currencies

Currencies correct constantly and quietly, mostly because central banks adjust policy. They rarely get correction coverage because the moves are small and nobody’s retirement depends on a single currency pair.

Precious metals

Gold and silver correct against real interest rates and against dollar strength, so a metals correction often arrives when equities are fine, or when a flight to safety has already pushed the metal up too far. The August 2026 Reddit discussion about whether an AI-driven selloff would send investors toward precious metals is a good illustration of how retail investors think about that relationship: metals get treated as protection, then sold when they fall for reasons that have nothing to do with the equity market.

Industrial commodities and mining shares

These are leveraged on the industrial cycle, so they correct harder and earlier than the broader market when growth expectations shift. Mining equities amplify the underlying metal move in both directions, which makes them the most volatile large sector in most portfolios and the first place a correction shows up. They also tend to recover fastest once industrial demand expectations stabilize, because the operating leverage works both ways.

Energy assets follow their own driver almost entirely: supply decisions and inventory data rather than equity sentiment. Crypto trades on the widest band of all, where 10% to 20% daily moves are common enough that corrections there look unremarkable next to stock-market equivalents.

What Should Investors Do During a Correction?

Almost all sensible advice during a correction is boring, and that is the point. The steps below are general process, not a recommendation to buy or sell anything.

Check what you actually need in the near term. Money you must spend within the next couple of years does not belong in assets that can fall 20%. Separating near-term spending from long-term capital is the single most useful thing to do when markets turn.

Look at your diversification rather than your individual holdings. Everyone stares at the position that dropped hardest. The question that matters is whether the whole portfolio got hit or whether one concentrated bet did, because those call for very different responses.

Rebalance if you have drifted. A correction pushes equities down and bonds or cash up, which can quietly move you away from your target allocation. Selling some of what held up to bring the mix back is a rules-based action, and rules-based actions survive bad weeks far better than judgment calls do.

Check your time horizon and your age. Someone with 25 years of savings in front of them can absorb a 20% drawdown and wait. Someone drawing retirement income cannot, and the sequence of returns matters more to them than the average annual return does. This is the group most likely to need to cut risk, and doing it in a correction is uncomfortable but rational.

Separate price moves from broken businesses. A stock falling while its revenue, margins and balance sheet hold up is a different situation from a stock falling because the business is deteriorating. A correction that takes quality companies down 20% is an opportunity; a 20% decline in something with real problems is not, and the arithmetic looks identical on the chart.

Do not try to catch the exact low. Nobody does it consistently. If you want to deploy cash in stages, use fixed amounts on fixed dates rather than trying to read the bottom, and accept that the first tranche may land before the low.

Investor behaviour during corrections is worth taking seriously, because it is consistently bad. Long-running forum threads on the topic keep circling the same double-bind: people spend years calling for a correction, then treat the arrival of one as an emergency. The documented pattern of selling during declines and buying during rallies is exactly backwards, and it is the reason so many investors hold a losing position they could have held a winner instead.

How Can You Tell Whether a Correction Is Over?

No single chart confirms a bottom, and any honest answer to this question starts there. What follows are the things investors actually watch, not signals that reliably time anything.

New highs are the official ending. A correction is considered over once the index closes above its previous peak. Until that happens, the decline can still extend, and plenty of “corrections” turn out to be the first leg of something larger.

Volatility eases and stays eased. Sharp, erratic daily moves tend to fade as the dust settles. Spikes that keep coming back suggest the process is unfinished.

Market breadth improves. When more stocks participate in the advance rather than a handful of heavyweights carrying the index, the recovery is broader and usually more durable.

Economic data stops deteriorating. The soft-landing outcome, in which growth slows without tipping into recession, is the version that ends corrections fastest.

Bad news stops causing worse reactions. Once disappointing figures produce smaller declines, the market is absorbing them rather than panicking. That change in reaction is often more informative than the news itself.

Price action stabilizes on longer windows. A market that stops making new lows on a multi-month view has usually found a floor, even if it does not look obvious day to day.

Frequently Asked Questions

Conclusion: What to Do First During a Market Correction

A market correction is a decline of roughly 10% to 20% from a recent peak, and it is the ordinary cycle rather than an exception. Since 1950, the S&P 500 has fallen more than 10% within the year in about half of all years on record, so a dip of this size is closer to normal than most people assume.

Three things to check first, in order: what you need from the portfolio in the next couple of years, whether you are diversified, and whether any individual holding is falling because its business is deteriorating rather than because the market is. If all three check out, the decision usually takes care of itself.

Everything above is general information about how markets behave, not individual financial advice. Past performance is no guarantee of future results, and no one can tell you in advance whether the next correction ends at 12% or becomes something much larger.

Source: https://www.pgm-blog.com/what-is-a-market-correction/