Socius News22.6%. That is what the Dow Jones Industrial Average lost on Monday 19 October 1987, 508 points in one session, the largest one-day decline in its history according to Federal Reserve History. Black Monday 1987 is the crash in which a hedging strategy, portfolio insurance, sold into the fall it was designed to protect against. Between $60 and $90 billion of equity assets sat under portfolio insurance programmes that week, and their models called for $20 to $30 billion of sales on the Monday alone.
Portfolio insurance was a rule, not a forecast. Its computer models set a stock-to-cash ratio for each market level and reduced the stock weight as prices fell, by selling index futures. That rule turned every decline into an order to sell more. The previous episode covered the 1929 stock market crash and its 10% margin; 1987 is the same arithmetic run by a machine.
What caused Black Monday 1987?
The Brady Report, written by the Presidential Task Force on Market Mechanisms in January 1988, starts the story on Wednesday 14 October. Two pieces of news arrived that day. The Commerce Department announced a merchandise trade deficit of $15.7 billion for August, about $1.5 billion above what the markets expected. Members of the House Ways and Means Committee filed legislation to remove tax benefits attached to the financing of corporate takeovers. The Dow fell 95 points that day, 57 points on Thursday and 108 points on Friday 16 October, a 4.6% loss and the largest one-day point drop to that date. By Friday's close the S&P 500 was down more than 9% for the week, according to Mark Carlson's 2007 history for the Federal Reserve Board.
Those three days did not cause Monday. They loaded it. Portfolio insurers sold the equivalent of about $530 million of stock on Wednesday, $965 million on Thursday and $2.1 billion on Friday, and the Brady Report notes that their models still called for far more. The market went into the weekend with a queue of mechanical sell orders waiting for the open.
The market was also expensive by the standards of the time. When the Dow reached 2,722 in August, the Brady Report says stocks were valued at levels that challenged historical precedent. Federal Reserve History puts the rise at 44% in seven months. A survey by Robert Shiller in the days after the crash found that 71.7% of individual investors and 84.3% of institutions already thought the market was overpriced before 19 October. Asked what caused the fall, respondents cited no news story. They cited the fall itself.
August 1987
Dow at 2,722
The peak, after a 44% gain in seven months. Between $60 and $90 billion of equity assets sit under portfolio insurance programmes.
14 October 1987
Trade deficit and tax bill
A $15.7 billion August trade deficit, $1.5 billion above expectations, and a bill to cut takeover tax benefits. The Dow loses 95 points.
16 October 1987
Down 108 points, 4.6%
The largest one-day point drop to that date. Portfolio insurers have sold about $3.6 billion of stock equivalent over three days and their models still call for more.
19 October 1987
Black Monday: 508 points, 22.6%
604 million shares trade. S&P 500 futures fall 29%. Portfolio insurers sell about $4 billion in stocks and futures against models that ask for $20 to $30 billion.
20 October 1987
The Fed's one-sentence statement
The Federal Reserve affirms its readiness to serve as a source of liquidity. The Dow closes up more than 100 points, the largest gain on record at the time.
1988
First circuit breakers
The SEC approves market-wide halts: one hour if the Dow falls 250 points, two hours at 400 points.
27 October 1997
First halt triggered
The circuit breaker fires for the first time. Thresholds are raised afterwards.
2012 to 2013
7%, 13%, 20%
Halts move to the S&P 500. Level 1 fires four times in March 2020.
What was portfolio insurance in 1987, and why did it sell?
Federal Reserve History defines portfolio insurance as a hedging technique used by institutional investors that employs futures and options to offset movements in prices. Carlson describes the mechanics. Computer models computed an optimal stock-to-cash ratio at each market price. When prices fell, the model told the investor to cut the stock weight. The insurer sold S&P 500 index futures in Chicago rather than the shares.
The Brady Report puts a number on the rule: the models required selling roughly 20% of holdings for each 10% decline in the market. Nothing in the rule asked whether there was a buyer.
That is the difference between a hedge and a stop. A stop on a single trade is sized against a defined risk and executed once. Portfolio insurance was a stop that re-armed itself at every new low and grew with the size of the loss. A trader who has read about revenge trading will recognise the shape: the response to a loss was more of the action that produced it.
Shiller's survey adds a detail. Only 5.5% of the institutional investors who answered said they followed an explicit portfolio insurance scheme. The strategy was concentrated, and large where it existed.
What happened on the morning of 19 October 1987?
The New York Stock Exchange opened to a large imbalance of sell orders. By 10:00, Carlson writes, 95 S&P 500 stocks representing 30% of the index value had still not opened. The futures market in Chicago had opened and was falling. With the cash index built partly on stale Friday prices, the futures contract traded at a discount to the index, 20 points by Monday afternoon according to the Brady Report.
Index arbitrage carried the pressure across. Arbitrageurs bought the cheap futures and sold the underlying stocks in New York, which pushed the cash market down, which triggered the next portfolio insurance sale in Chicago. The Brady Report describes the loop in one sentence: selling pressure in the futures market was transmitted to the stock market by the mechanism of index arbitrage.
The volumes were beyond what the system had been built for. 604 million shares changed hands on the NYSE that day, worth just under $21 billion. Three portfolio insurers alone sold just under $2 billion of stock and the equivalent of $2.8 billion in futures. By early afternoon, insurers had contributed more than $3.7 billion of selling pressure across both markets. Carlson reports that roughly 40% of non-market-maker sales in the futures market on 19 October came from portfolio insurers, and that one large institution sold $1.1 billion during the day.
Portfolio insurance sales, 14 to 19 October 1987
Stock-equivalent value sold in stocks and futures, billions of dollars
Source: Report of the Presidential Task Force on Market Mechanisms (Brady Report), January 1988
The Dow closed at 1,738. The S&P 500 lost about 20%. The S&P 500 futures contract lost 29%. From the close of Tuesday 13 October to the close of Monday 19 October, the Dow had lost almost a third, and the Brady Report values the loss across all US stocks at approximately $1.0 trillion.
Why did Tuesday 20 October matter more than Monday?
Because Monday was a price and Tuesday was a plumbing problem. Securities firms had to meet margin calls on the futures they had sold and settle the trades they had made. Their banks were being asked to lend into a market that had just lost a fifth of its value. The futures discount widened to 40 index points by Tuesday midday. The Brady Report notes that futures prices implied a Dow just above 1,400 while stocks were trading just above 1,700.
Before the open, the Federal Reserve issued a statement of one sentence. Chairman Alan Greenspan said: "The Federal Reserve, consistent with its responsibilities as the Nation's central bank, affirmed today its readiness to serve as a source of liquidity to support the economic and financial system." The Fed followed the words with open market operations that pushed the federal funds rate down to around 7% on Tuesday from over 7.5% on Monday, according to Carlson.
The banks lent. Carlson records lending to securities firms of $1.4 billion on 20 October, against a normal level of $200 to $400 million. Federal Reserve History adds that the ten largest New York banks nearly doubled their lending to securities firms during the week of 19 October. The Dow closed Tuesday with a net gain of more than 100 points, the largest gain on record at the time. Over two sessions it recovered 288 points, 57% of the Monday loss.
-22.6%
Dow, 19 October 1987
508 points, the largest one-day decline in its history (Federal Reserve History)
$60 to $90bn
Equity under portfolio insurance
Models called for $20 to $30 billion of sales on 19 October (Brady Report)
604m
NYSE shares traded on 19 October
Just under $21 billion; record volume overwhelmed systems (Carlson, 2007)
7%
Fed funds rate on 20 October
Down from over 7.5% on Monday (Carlson, 2007)
How long did the market take to recover from Black Monday 1987?
Less than two years, according to Federal Reserve History, for the market to pass its pre-crash high. The Dow started 1987 at 1,897 and closed the year at 1,939, Britannica reports. A 22.6% day did not produce a negative year.
That is the sharpest contrast with the 1929 stock market crash, where the drawdown ran to 89% and the peak was not seen again for 25 years. The arithmetic behind both is the same. A 22.6% loss needs a 29.2% gain to recover. An 89% loss needs 809%. The size of the drawdown, not the size of the headline, sets the time to recover.
The crash also travelled. In the most severe case, New Zealand's market fell 60%, Federal Reserve History records. Australia's fell more than 40% according to Britannica.
What did the circuit breakers change, and what do they look like today?
The Brady Report recommended that circuit breaker mechanisms, such as price limits and coordinated trading halts, be formulated and implemented in advance. The SEC approved the first market-wide halts in 1988, the NYSE working group's history records: a one-hour halt if the Dow fell 250 points from the previous close, and a two-hour halt at 400 points.
The first trigger came on 27 October 1997. The thresholds were raised to 350 and 550 points, then in 1998 replaced by percentages of 10%, 20% and 30%. After the 2010 flash crash the reference index moved to the S&P 500. The SEC's investor bulletin describes the current levels: a 7% decline halts trading for 15 minutes, 13% halts it for another 15 minutes, and 20% closes the market for the rest of the day. Level 1 fired on 27 October 1997 and four times in March 2020, on the 9th, 12th, 16th and 18th.
A market-wide circuit breaker is a daily loss limit for an entire exchange. It does not decide whether prices were right. It stops the machine so that humans can look at the number. Prop firms apply the same idea to a single account.
What your journal would have shown
The following figures are illustrative demo data, not a historical account.
Take a demo account that ran a portfolio insurance rule on $10 million of stock in October 1987, selling futures worth 20% of holdings for each 10% fall in the index. A journal that logged the rule as trades would show one line on 14 October, a larger line on 15 October and a larger one again on the 16th. By Friday's close, the demo account had sold at three successively lower prices and held the most futures short at the lowest level of the week.
Now read Monday in R. If the account defined 1R as a 2% adverse move on its remaining stock, the 22.6% day prints as roughly -11.3R on the unhedged part, offset by whatever the Friday futures sales covered. The average fill sits well below the level at which the rule was switched on, because a rule that sells after each decline always sells after.
The same journal shows expectancy by day of the week, and it shows the size of each position relative to the loss it was meant to cap. Those two views are what R-multiples and trading expectancy exist for. Socius Trades computes them from the trades your cTrader or MetaTrader account has already executed, on a plan that starts at Free. It never places, closes or modifies a trade. We look. We never touch.
“Portfolio insurance did not fail to predict October 1987. It never tried. It sold because the price fell, and the price fell because it sold.”
Trading involves risk of loss. Socius Trades is an analytics tool, not investment advice.