Here's What We Learned From the Official Report on the 'Flash Crash' in U.S. Treasuries
On Oct. 15, Wall Street watched in collective shock as the yield on the benchmark 10-year U.S. Treasury plunged before careering upward again on seemingly little news.
U.S Treasuries are meant to be one of the most liquid markets in the world, meaning they should in theory be impervious to big jumps in their price. Market participants have blamed a number of culprits for the day's wild swings. These include new regulations that make it more difficult or more expensive for large banks to make markets in a number of fixed-income securities, the rise of high-speed electronic trading in the U.S. Treasury market, and heavy one-way positioning by big investors.
On Monday, the U.S. government released its report on the day's dramatic swings. Here's what we learned about the day's events.
• Investors had been heading for the exits — In the weeks leading up to Oct. 15, many big investors had been betting that interest rates were going to increase as the global economy improved. "Short" positions in eurodollar futures had reached a record level by the end of September, meaning many funds were positioned for an increase in rates. As the economic outlook began to deteriorate in early October and interest rates dipped lower, these investors were forced to unwind their short positions or offset them through hedging. "Much of this unwind took place in the two weeks ahead of October 15," according to the report, "but the change was also significant on October 15 itself." The report also mentions some large asset managers that had been betting on the continuation of low volatility in markets. The combination of late investors still rushing to adjust their interest rate or volatility positions (and potentially already nursing losses) probably made them especially sensitive to any new data that appeared to confirm a dimming outlook for global growth. That data would come on the morning of Oct. 15, when U.S. retail sales figures flashed across screens at 8.30 a.m.
• Computer-driven trading played a role — As U.S. retail sales hit trading screens on the morning of Oct. 15, they triggered a frenzy of activity in U.S. Treasuries, particularly by investors known as principal trading firms (and better known as high-frequency trading firms), which typically use computer-generated proprietary trading strategies (read: algorithms) to make their money. These so-called PTFs accounted for more than 50 percent of the total trading volume in both cash and futures markets for U.S. Treasuries on Oct. 15. That proportion is not unusual, according to the report, but what is remarkable is their level of activity on the day. As trading began to heat up, these algorithms essentially fed on each other, causing the amount of "self-trading" undertaken between different arms of the same PTF firms to increase. For 10-year U.S. Treasuries, the amount of self-trading reached 14.9 percent for cash and 11.5 percent for futures. As order matching engines struggled to handle all the buy/sell messages now flooding the market, the speed at which they are able to execute orders (known as latency) also began to slow.
• Investors and dealers reacted swiftly — As the number of would-be buyers began to dramatically outweigh would-be sellers and the flurry of orders began to overwhelm trading machines, market participants responded to the changing market conditions by retrenching. PTFs dramatically reduced the amount of orders they left standing in both the cash and futures markets, while bank dealers cut back more modestly. As reported by Bloomberg News at the time, some simply pulled the plugs on their market-making machines. Market depth, as measured by the amount of quotes outstanding, plummeted and transaction costs increased as the yield on the 10-year U.S. Treasury plunged and then recovered in a matter of minutes.
• Liquidity is still a mystery — Despite the the 72-page report into the matter, much of the day's events remains a mystery and U.S. regulators did not pinpoint a single cause. Meanwhile, the degree to which liquidity, or ease of trading, in the vast U.S. Treasury market has deteriorated remains the ultimate enigma. The report notes that many of the usual measures of liquidity, such as the difference between buy and sell prices in the securities (known as bid/ask spreads), remain robust. The report also notes, however, that these measures may need to be augmented. "More traditional liquidity metrics continue to look favorable ... but may provide only a partial view into liquidity as a whole," the report said. "Average bid-ask spreads and market depth, though often indicative of the general market conditions, may need to be complemented by other measures in light of these changes to obtain a more meaningful picture of the state of market liquidity in the current structure.