Black Monday 1987: Causes, Crash and Recovery Timeline

On Black Monday, October 19, 1987, the S&P 500 fell 20.47% in a single day, from 282.70 to 224.84. Over the same session, the Dow Jones Industrial Average fell 508 points, or 22.6%, from 2,246.74 to 1,738.74.

A few things help read those numbers. An index point is just the unit an index is quoted in; what matters is the percentage. Of the two, the Dow follows 30 large companies and weights them by share price, while the S&P 500 follows 500 and weights them by company size, so the two rarely move by exactly the same percentage. And a 20% index decline is the combined move of its stocks, weighted the way the index weights them: some stocks fell much more that day and some fell less.

How long the recovery took depends on how you measure it. The six versions are in the recovery section below.

The crash was the end of a week-long slide. By late August 1987 the Dow had gained 44% in seven months, according to the Federal Reserve's history of the crash. In mid-October a larger-than-expected trade deficit was reported and the dollar fell. Markets dropped hard from October 14, and on Friday, October 16, the Dow lost 4.6%.

The weekend gave only a brief pause. On Monday, markets in Asia fell before New York opened, and the Fed's history notes that New Zealand's market eventually fell 60%. In New York, the Dow "crashed at the opening bell" and finished down 508 points, while the S&P 500 closed at 224.84, down from 282.70 on Friday.

In money terms: $1,000 invested at the August record was worth $839 by Friday's close and $668 by Monday's close. Monday was not the bottom, either: the index's lowest close came weeks later, on December 4, 1987, at 223.92, down 33.5% from the record, the deepest drawdown of the episode.

The next morning, October 20, the Federal Reserve said in a statement that it "affirmed today its readiness to serve as a source of liquidity to support the economic and financial system." In practice, the Fed signaled it would supply cash if the financial system needed it, and the Fed's history says it encouraged banks to keep lending on their usual terms.

To see those days from the inside, one decision at a time, try our Black Monday 1987 simulator, a free replay with pretend money.

The answer depends on two choices: which level counts as "recovered", and whether you count dividends and inflation. Each date below is the first trading day the index closed at or above its starting level.

• Back to the Friday close before the crash: one year and one day (October 20, 1988). This sounds fastest, but by that Friday the S&P 500 had already lost 16% from its August record.
• With dividends reinvested, back to the August record: about 19 months after the crash (May 17, 1989, accurate to within about a week). This is called total return: it counts the dividends companies paid during the slump as if they were put back into the index, which is roughly what an index fund that reinvests dividends does.
• By price, back to the August record: 21 months and 7 days (July 26, 1989). This is the figure most sources mean by "July 1989".
• The Dow, back to its August record: 22 months after the crash (August 24, 1989). Counted from the Dow's August 1987 peak, that is one day short of two years, which is where "two years" comes from.
• Price after inflation: around the turn of 1992, more than four years after the crash. Consumer prices rose about 4.4% a year in between, so the same number of index points bought less.
• Dividends and inflation together: late July to August 1989, about 21 to 22 months after the crash. This is the real return version.

"About 20 months", another common figure, matches total return measured from the August peak, or price measured from the December 4 low.

To check the four S&P 500 rows that use the August record, we repeated them with Robert Shiller's monthly average prices, a separate data series and a cruder method. It puts the price recovery in July 1989, total return in May 1989, real total return in August 1989 and real price in January 1992, all within about a month of the daily figures. Both checks lean on related underlying data, so treat the dividend and inflation dates as accurate to the month, not the day.

If "dividends reinvested" is new to you, our lesson on what an index fund is explains how most index funds handle dividends. To see how inflation shrinks a return, try the inflation-adjusted return calculator.

No single cause explains the October 1987 crash. Looking back, the Federal Reserve's history of the crash lists the factors regulators and economists identified afterward:

• A fast run-up. A 44% gain in the Dow in seven months raised fears of an asset bubble.
• Portfolio insurance. This was a popular strategy built to limit losses by selling automatically as prices fell. The Fed's history says it "accelerated the crash's pace as initial losses led to further rounds of selling."
• Bad news the week before. A larger-than-expected trade deficit and a falling dollar.
• Market-structure problems. Stocks, options and futures each had their own schedule for paying for trades, so some accounts could end up owing money and be forced to sell. And exchanges had no way to pause trading during a collapse.

A simplified illustration of how portfolio insurance could feed on itself (hypothetical, not a reconstruction of any real 1987 fund): a fund with $100 million in stocks follows a rule to cut its stock exposure as prices fall, and does it by selling stock-index futures. Prices dip, so it sells. If thousands of funds follow similar rules, they sell at the same moment, which pushes prices down further and triggers the next round of selling under the same rules.

The names get mixed up because 1929 had its own black days. Black Thursday was October 24, 1929, when panic selling began on about three times the normal trading volume. On Black Monday, October 28, 1929, the Dow fell nearly 13%, and on Black Tuesday, October 29, nearly 12% more. Black Monday 1987 is the October 19 crash on this page, and it is the one people usually mean today.

The bigger difference was the aftermath. According to the Federal Reserve's history, the 1987 losses "were not followed by an economic recession or a banking crisis." After 1929, the slide ran until the summer of 1932, when the Dow closed 89% below its peak. It took until November 1954 for the Dow to return to its 1929 high. The 1987 recovery, by contrast, is measured in months. For a slow bubble deflating rather than a one-day crash, the closer comparison is the dot-com bust. After peaking on March 24, 2000, the S&P 500 fell 49% to October 9, 2002, and did not close above its 2000 peak again until May 30, 2007, more than seven years later. Our AI bubble explainer covers that comparison, and our guide to how stocks performed during past wars covers other shocks. The 1929 crash, day by day, and how long its recovery really took, measured four ways, is in our post on the 1929 stock market crash and Black Tuesday. An earlier panic, in 1907, took the Dow down 48.5% over 22 months, most of it before the bank runs began; that story is in our post on the Panic of 1907.

Today's rules would interrupt a fall like 1987's. After the crash, exchanges added market-wide circuit breakers: pauses in trading across U.S. stock markets, triggered by how far the S&P 500 has fallen from the previous day's close. According to the SEC, a 7% (Level 1) or 13% (Level 2) drop before 3:25 p.m. Eastern halts trading for 15 minutes; at or after 3:25 p.m. those two levels do not halt trading. A 20% drop (Level 3) at any time halts trading for the rest of the day.

Under today's rules, a 20% fall in the S&P 500 would end trading for the day. That does not tell us where a 1987-style day would have closed, because halts change how traders behave, nor does it determine what happens in the trading sessions that follow.

For illustration only, using the closes on Friday, October 16, 2026 (the thresholds reset every day), a 20.47% fall in the S&P 500 would be about 1,568 points, and the Dow's 22.6% would be about 11,750 points. For scale, the S&P 500's largest one-day percentage fall since 1987, through October 16, 2026, was 11.98% on March 16, 2020.

Circuit breakers pause trading. They do not set a floor under prices. The historical record shows how 1987 unfolded; it cannot forecast whether or when a fall like it happens again.

• S&P 500 daily closes: Yahoo Finance, ^GSPC history. S&P 500 Total Return index from January 4, 1988: Yahoo Finance, ^SP500TR.
• Dow Jones Industrial Average, 1987 to 1992: MeasuringWorth DJIA daily series.
• Dividends before 1988: Robert Shiller's monthly S&P 500 data.
• Inflation: BLS CPI-U, all items, not seasonally adjusted (CUUR0000SA0).
• History and causes: Federal Reserve History, "Stock Market Crash of 1987". 1929 dates: Federal Reserve History, "Stock Market Crash of 1929" and EH.net, "The 1929 Stock Market Crash".
• Circuit breakers: SEC, Investor.gov, "Stock Market Circuit Breakers".

Method. A recovery date is the first trading day the index closed at or above the starting level. Total return from August 25, 1987: Shiller's dividend figure for each month is an annual rate (the trailing four-quarter total), so we divide it by 12 to get that month's dividend, spread it evenly over the month's trading days, and add it to the daily price change. On January 4, 1988 the series switches to the official S&P 500 Total Return index, rescaled so both series have the same value that day, so there is no jump at the join. Inflation-adjusted values divide each day's level by that month's CPI-U (a second version interpolating CPI between months gives dates within a few days of the first). The 2026 point figures use the October 16, 2026 closes of 7,659.44 (S&P 500) and 52,002.41 (Dow). Dot-com and 2020 figures use the same daily S&P 500 data.

Historical data shown; past performance does not indicate future results. StockCram is an educational platform, not a broker-dealer or investment adviser, and nothing on this page is investment advice.

Every post in this series measures crashes the same way; our market-history method explains how.