What is a Market Correction? Definition, Formula, and Example
A market correction is a decline of 10% to 19.9% in a stock index or asset from its most recent peak, distinct from a bear market which begins at a 20% drawdown.
What is a Market Correction?
A market correction is a decline of at least 10% but less than 20% in a broad index — most commonly the S&P 500 — measured from its most recent closing high. The 10% threshold is a convention, not a law of physics, but it is the standard used by every major data provider and financial news desk, which makes it a self-reinforcing event: once an index crosses −10%, headlines declare a correction, algorithms tag it, and positioning shifts. A decline of 20% or more graduates to a bear market; anything under 10% is a pullback. The term applies to individual stocks, sectors, and asset classes as well, and single-stock corrections of 10% in a day are routine around earnings.
How a Correction is Identified
The measurement is simple arithmetic against the trailing peak:
Correction Depth = (Current Close − Peak Close) ÷ Peak Close
Rules that govern the classification:
- Measured on closing prices, not intraday lows (though intraday breaches are widely reported).
- The peak resets only after the index recovers to a new all-time or cycle high.
- −10.0% to −19.9% = correction; ≤ −20% = bear market; 0% to −9.9% = pullback.
Historically, the S&P 500 has averaged roughly one 10% correction every 1.5 to 2 years. The average correction since 1950 runs about −14% and takes roughly four months from peak to trough and another four months to recover — though the distribution is wide.
Worked Example
The February–March 2018 correction is a clean case. The S&P 500 closed at a peak of 2,872.87 on January 26, 2018. Nine trading days later, on February 8, it closed at 2,581.00 — a decline of 10.2%, officially a correction. The trigger was a spike in wage inflation data and the implosion of short-volatility products ("Volmageddon," February 5, when the VIX jumped 116% in one session). The index bottomed, retested lows in early April, and recovered to new highs by August 2018 — peak to recovery in about seven months. Individual names corrected harder: FB (now Meta) fell over 20% from its July 2018 peak on its own idiosyncratic correction later that year.
When Traders Use the Concept
- Regime classification: Systematic strategies de-risk or shift factor exposure when a correction is declared, because volatility and correlations rise through the −10% threshold.
- Buying discipline: Long-term investors treat corrections as scheduled rebalancing or accumulation points — historically, buying the S&P 500 at −10% has produced positive 12-month forward returns the large majority of the time.
- Volatility trading: Correction onset reliably lifts the VIX above 25 and flips futures into backwardation, creating defined setups in volatility products.
- Risk budgeting: Portfolio managers size drawdown tolerance around the statistical reality that a 10%+ hit arrives roughly every other year.
Limitations and Common Misconceptions
The 10% line is arbitrary — a −9.8% decline and a −10.2% decline are economically identical, yet only one triggers the label and its associated flow effects. Corrections also carry no predictive content: most never become bear markets (roughly three-quarters of post-WWII corrections stopped before −20%), so selling at −10% to "avoid the crash" is a losing trade on average. The label is backward-looking by construction; you only know the peak after the decline has happened. And for individual stocks, 10% moves are noise, not signal — applying index-level correction logic to a high-beta name produces meaningless classifications.