The S&P 500 (^GSPC +0.28%) closed at a record 7,818.93 on Oct. 6. Then tech stocks got hit.
The Nasdaq Composite (^IXIC -0.02%) dropped about 4.5% last week, its worst week since June, as chipmakers slid on worries about how long the spending frenzy around artificial intelligence (AI) can last.
The S&P 500 lost about 2% in the same five trading days.
As I write, the S&P 500 sits only about 1% under the record. It’d take a close around 7,037 to enter a correction, the common term for a fall of 10% or more.
But a week like that can make an index fund holder wonder what a deeper drop would be like.
Since 1950, the index has seen 24 declines of 10% or more from a record close. And all of them eventually ended with a new record close. The median wait from the bottom to that new high was around five months.
But the waits varied a lot. The quickest took around a month, and the slowest took almost six years.
Image source: Getty Images.
Corrections didn’t last long
I measured every decline using daily closing prices, from a record close to the lowest close after it, then to the next record close. (The index didn’t beat its 1929 high until September 1954, so the first of the 24 declines was in 1955.)
Thirteen of the declines stayed below 20%. For these, the median wait from the bottom to a new record was around three months. And none of the 13 took more than about six and a half months.
The latest one got close to the 20% line. After a record close of 6,144.15 on Feb. 19, 2025, the S&P 500 dropped as tariff worries mounted. It finished at 4,982.77 on April 8 — down almost 19% from that February record.
Less than three months later, on June 27, it closed at a new record of 6,173.07. Of course, no one knew on April 8 that the bottom was in. The index had just fallen more than 12% in four trading days.
Bear markets took much longer
The other 11 declines topped 20%, the usual line for a bear market. From the bottom, the median wait for a new record was around 15 months.
But the range was huge. After the August 1982 low, the index needed under three months to hit a new record.
The 2020 crash sent the S&P 500 down around 34% in just over a month. Still, it closed at a record on Aug. 18, 2020, under five months after bottoming.
At the other end, the drops that ended in 1974, 2002, and 2009 each took about four to six years to come back. After the October 1974 low, the index didn’t close at a record again until July 1980.
The numbers start the clock at the bottom, which nobody can spot in real time. Measured from the old record instead, the median wait for a new one was about 14 months. That ran roughly seven months when the drop stayed a correction, and around two years when it became a bear market.
What should a holder plan for?
Declines like those have come along about once every three years on average since 1950, but never on a schedule. Even after last week, the index is up over 12% in 2026 as of this writing.
For a holder, I think the slow cases arguably matter more than the median. Seven of the 11 bear markets took more than a year from the low to a new record, and three took four years or longer.
Today’s Change
(0.28%) +21.84
Index Level
7,712.57
Key Data Points
Day’s Range
7,668.03 – 7,748.42
52wk Range
6,316.91 – 7,844.52
Notably, these numbers are on price alone. Someone who reinvested dividends would usually have broken even a bit sooner.
So I’d only hold money in an S&P 500 index fund if it can ride out a few rough years. Money needed earlier than that might be stuck waiting at the wrong time.
Last week’s slide isn’t near any of these 24 declines yet, and it may never become one. But if it does, a new record has followed every past drop that big since 1950. Over half of those comebacks took six months at most from the bottom, while one took almost six years.
