Recession and the Stock Market: When Do Stocks Bottom?

Why timing matters more than the label

"Are we in a recession?" is one of the most searched market questions, and one of the least useful for trading. In the U.S., recessions are dated by the Business Cycle Dating Committee of the National Bureau of Economic Research (NBER), usually many months after they start. By the time the label shows up, the market has usually already moved.

What you need as a trader isn't the label, it's the sequence: when did stocks peak, when did they bottom, and how did those dates line up with the economy? We measured it on the S&P 500 since 1999.

Every drop of 10% or more since 1999

Measured on SPY's weekly closes, dividends included, from a record high to the lowest point before the next record:

• March 2000−45.7%
• October 2007−54.6%
• July 2015−11.2%
• September 2018−17.1%
• February 2020−31.8%
• December 2021−23.9%
• February 2025−16.9%

Exact values from table 2 in this article. Bar lengths share a zero baseline; no new calculation or simulated price path. Read the table and its stated sample and limits in this article.

That's seven drops of 10% or more in 27 years, four of them deeper than 20%. Two stand out: after the bottoms of 2002 and 2009, it took more than three years to get back to the old high. The others recovered within about five months, except 2022, which took a little over a year.

The drops of 2010 and 2011 aren't in the table because the market hadn't gotten back to its 2007 high yet. They were part of the long climb back from 2009.

Lining up the market with the economy

The NBER dates three recessions in this period: March to November 2001, December 2007 to June 2009, and February to April 2020.

The peak came first every time. Stocks were already falling before each recession officially started, and in 2000 the lead was a full year, because the drop began as a tech bubble deflated.

The bottom usually came before the end. In 2009 and 2020, the market turned up while the economy was still shrinking and the news was still ugly. In 2001–2002, though, the bottom came long after the official end of the recession, because the market was still working off the excesses of 2000.

And a bear market doesn't need a recession. In 2022, the S&P 500 fell 23.9% on weekly closes without any NBER recession. The drivers were inflation and the fastest rate hikes in decades, not a drop in output.

Why the market leads

Stock prices reflect expected future profits. When investors see orders slowing, credit tightening or jobless claims rising, they reprice before any of that reaches the official output numbers. Recession dates, on the other hand, are set from backward-looking data like employment, income and production, and announced with a long delay.

Same thing on the way back up. By the time a recession is declared over, the market has often been rising for months, because investors are already pricing in the recovery.

What this means if you trade ES or NQ

Don't wait for the word "recession". By the time it's official, the price move is usually well underway. What matters to you is the data flow itself: payrolls, jobless claims, ISM surveys and credit spreads.

Know a correction from a bear market. Three of the seven drops stopped between 11% and 17%, with no recession, and recovered within months. The deeper ones came with either a recession or, in 2022, an inflation shock.

Expect bottoms on bad news. In 2009 and 2020, the market turned while the headlines were at their worst. A bias that waits for good news before turning bullish will often be late.

Respect recovery time. After the two deepest drops, it took more than three years to get back to the prior high. A trend filter like the 10-month moving average is one way to avoid sitting through the worst of a drop like that, at the cost of some false signals.

Where the Fed fits in

In all three recessions, the Fed was already cutting rates before the market bottomed, though in 2020 only by a few weeks. Our article on the Fed's first cut shows that the cuts of 2001 and 2007 didn't stop the declines. Policy can soften a downturn. On this history, it didn't reverse a market that was already pricing one in.

How to track it yourself

1. Plot the S&P 500 with its running all-time high. The gap between the two is the current drawdown.
2. Mark the NBER recession dates on the same chart.
3. Each month, write down the unemployment rate and initial jobless claims next to the drawdown.
4. Read a drawdown past 20% with rising unemployment as the 2001/2008 pattern, and a drawdown past 20% with low unemployment as the 2022 pattern.

Limits

• Few cycles. Three recessions and four bear markets since 1999 are case studies, not statistics.
• Weekly closes. Intraday lows were deeper. The October 2022 intraday low, for example, came after the weekly-close low we report.
• Total return. Dividends are included, which shortens recovery times compared with price-only charts.
• Recession dates come out late and can be revised. They're a historical label, not something traders knew in real time.

How we did it

• SPY weekly adjusted prices (dividends included) from Alpha Vantage (TIME_SERIES_WEEKLY_ADJUSTED), November 1999 to October 9, 2026.
• Declines are measured from a record weekly close to the lowest weekly close before a new record; only declines of 10% or more are listed.
• "SPY during the recession" is the total return from the last weekly close before the first recession month to the last weekly close of the final recession month.

Sources

• National Bureau of Economic Research. US Business Cycle Expansions and Contractions (data last updated March 14, 2023). nber.org
• Alpha Vantage. TIME_SERIES_WEEKLY_ADJUSTED (SPY). Data retrieved October 2026.

FAQ

Does the stock market fall before a recession? In all three U.S. recessions since 1999, the S&P 500 peaked before the recession started, by anywhere from less than a month to about a year.

Does the stock market bottom before the recession ends? In 2009 and 2020, yes, by three months and one month. In 2001–2002, the bottom came 11 months after the recession ended.

Can there be a bear market without a recession? Yes. In 2022, the S&P 500 fell 23.9% on weekly closes with no recession.

How long does it take the S&P 500 to recover from a crash? Since 1999, recoveries took between 12 weeks (2025) and about four years (2002–2006) from the bottom, dividends included.

How many 10% declines has the S&P 500 had since 2000? Seven, measured on weekly closes from a record high, four of them deeper than 20%.

---

Research and education only. Not financial advice. Past results do not predict future returns.

How this article was made: drafted with AI assistance from public data and the sources listed above; facts and figures checked by Gabriel Décarie and Karel Panneton.