Volatility - SEC Report On August 24th Flash Crash

Equity Market Volatility – SEC Report On August 24th Flash Crash

On Monday, August 24, 2015, the U.S. equity markets and equity-related futures markets experienced unusual price volatility, particularly during the period surrounding the 9:30 a.m. E.T.2 start of regular trading hours for the equity markets.

Prior to 9:30, the most actively traded equity product – the SPDR S&P 500 ETF Trust (“SPY”) – declined to more than 5% below its closing price on the previous trading day (Friday, August 21, 2015). The most actively traded equity-related futures contract – the E-Mini S&P 500 (“E-Mini”) – declined to its limit down price of 5% below the previous trading day’s closing price and was paused for trading from 9:25 to 9:30.
At 9:30, SPY opened for regular trading hours at 5.2% below its previous day’s close and then further declined to a daily low of 7.8% by 9:35. By 9:40, SPY recovered past its opening price and eventually closed down 4.2%. SPY’s decline from previous day close to August 24 open was the second largest in the last decade, while SPY’s decline from previous day close to August 24 daily low was the 10th largest in the last decade.
From 9:30 to 9:45, more than 20% of S&P 500 companies and more than 40% of NASDAQ-100 companies reached daily lows that were 10% or more below their previous day’s closing price.

This Research Note assesses the operation of U.S. equity markets under the stressed conditions of August 24. In recent years, the SEC and self-regulatory organizations (“SROs”) have implemented several regulatory initiatives to address transitory price volatility. Among other things, August 24 provides a useful opportunity to evaluate the practical operation of these initiatives.

This Research Note provides empirical data and other information to help assess trading on August 24, including several issues that have been debated among market participants and in the media. These issues include the opening process at primary listing exchanges, the triggering of trading pauses under the National Market System Plan to Address Extraordinary Market Volatility (commonly known as the Limit Up-Limit Down, or “LULD”, Plan), and the effects of market volatility on trading in exchange-traded products (“ETPs”).

Section I describes the data used in the paper and provides summary statistics about the data.

Section II provides an overview of trading on August 24 by examining two broad indices – the S&P 500 and NASDAQ-100 – and their related products.

Section III compares trading in a large dataset of corporate stocks and ETPs on August 24 with trading during previous control periods.

Section IV addresses the opening process at the primary listing exchanges.

Section V examines the operation of the LULD Plan and associated SRO rules and practices.

Finally, Section VI focuses on the widely varying nature of trading in ETPs on August 24, with a minority experiencing extreme volatility and a large number of LULD halts.

The following are some key data points in the paper:

Price Declines (Section III.A):

Corporate stocks (“Corporates”) with the largest capitalization were particularly affected on August 24.4 Of the 41 Corporates in the Very Large capitalization bin (which alone represent nearly a third of Corporates market capitalization), more than half (21) declined by 10% or more on August 24. These included 5 of the largest 10. In comparison, only 30% of the more than 4,000 Corporates in the Large, Mid, and Small capitalization bins declined by 10% or more.
The largest capitalization ETPs experienced declines that were similar to those of smaller ETPs. Of the 50 largest capitalization ETPs, 20 (40.0%) declined by 10% or more, while 36.5% of more than 1,300 other ETPs also declined by 10% of more.
Regulation SHO short sale restrictions (“SSRs”) were triggered on August 24 in more than 2,000 securities, which is the second largest number since SSRs were implemented in 2011. SSRs were triggered in 108 constituents representing more than 37% of the market capitalization of the S&P 500, and in 42 constituents representing more than 50% of the market capitalization of the NASDAQ-100. When SSRs are triggered, short sale orders in that security generally are subject to a price test that requires the orders to be executed at prices greater than the national best bid for the security.

Trading Metrics (Sections III.B and III.C):

Trading metrics for control periods indicate that the opening 15 minutes of regular trading hours typically are the least liquid portion of the trading day, with wider spreads, less quoted depth, and higher volatility. These patterns were observed on August 24, but with much larger trading volume than normal.
Trading volume in the minutes following the open was much higher than in control periods, particularly for Very Large Corporates (more than 400% higher) and for nearly all market capitalization bins of ETPs (ranging from more than 400% higher for Large ETPs to more than 800% higher for Small ETPs). The fact that prices also declined most substantially in the Very Large Corporates and in ETPs indicates that much of this volume surge after the opening was initiated by sell orders that were relatively insensitive to the previous day’s prices.
Quoted depth (inside through 19 cents away), which typically is at its daily lows in the opening minutes of a trading day, was much lower in the opening minutes on August 24 than in control periods, particularly for Very Large Corporates (more than 70% reduction) and for ETPs (more than 90% reduction). Consequently, the surge in selling in the minutes following the open on August 24 interacted with much lower than normal levels of displayed liquidity.

Opening Process and Market-Wide Circuit Breakers (Sections II and IV):

NASDAQ and NYSE Arca open their listed equities electronically at 9:30. The NYSE incorporates a manual element in its opening process, particularly on volatile days. On August 24, many NYSE-listed stocks opened for trading later than 9:30 on the NYSE (though NYSE-listed stocks were traded at other exchanges and off-exchange venues before and after 9:30). By 9:35, for example, the NYSE had opened 38% of its listed S&P 500 companies representing 53% of such companies’ market capitalization. By 9:45, these figures increased to 86% of NYSE-listed S&P 500 companies representing 91% of such companies’ market capitalization.
The S&P 500 Index (“SPX”), as calculated and disseminated by S&P Dow Jones Indices LLP (“S&P DJI”), declined on August 24 by only 5.2% from its previous day’s close. Until approximately 9:42, SPX remained substantially higher than the prices of the SPY (7.8% decline), E-Mini (7% decline), and the net asset value (“NAV”) of SPY (8% decline, as calculated with reference to consolidated real-time trade prices). S&P DJI generally uses last sale prices from only the primary listing market to calculate its equity indexes. Until NYSE-listed constituents of the S&P 500 index were opened on the NYSE, the SPX disseminated by S&P DJI reflected NYSE closing prices from August 21. The use of these previous day closing prices to calculate SPX in the opening minutes of August 24 likely caused its decline to be less than S&P 500 related products that reflected real-time trade prices.
The SPX is referenced in SRO rules to determine whether market-wide circuit breakers are triggered. Because the SPX did not decline by 7% (the first level trigger) on August 24, the market-wide circuit breakers, which would have implemented a 15-minute trading pause in all equities and related products, were not triggered on August 24.
Six of the 41 Very Large Corporates reached daily lows of more than 20% below their previous closing price soon after the opening cross on their primary listing exchange. Much of the decline occurred in a very short period of continuous trading after the opening cross. The rapid price decline was accompanied by a volume surge after the opening cross that substantially exceeded the volume executed prior to and in the opening cross itself. An LULD pause was triggered in one of the six Corporates.