Rise of the Machines
“It was the sound of well-educated young white men baying for money on the bond market,” he wrote. Other than in isolated pockets, that trading floor is gone. It’s been replaced by a world where algorithms analyze markets, place orders based on market conditions, and control billions worth of transactions completed faster than we can blink.
This setup has radically changed how people in the industry communicate, according to Jeffrey Harris, Gary D. Cohn Goldman Sachs Chair in Finance. Harris and co-author Mohsen Saad of the University of Sharjah recently explored how high-frequency traders harness information from electronic messages.
They studied whether such traders could accurately forecast short-term market conditions using state-of-the-art electronic trading methods in lieu of the face-to-face atmosphere of the floor. High frequency trading accounted for 73 percent of all trades taking place, the former head of NASDAQ said in 2011, according to Investopedia.
Their idea was to find out whether “someone, from looking at the order flow or the activity in a market place—without being face-to-face or learning anything about the person—can figure out anything about future market conditions based on what’s happening now,” Harris said.
“The bottom line of the paper is that it shows that you can, in fact, do that,” he said. “Without getting inside information, you can actually glean [information] from how people are trading or canceling or sending orders to the market… [and] make some assessment of what market conditions will be.”
Harris and Saad studied message traffic, orders, and cancellations run through NASDAQ’s order management systems. They based their study on NASDAQ data coming in at lightning-quick 15-second intervals during a one-month period.
Having the ability to make informed decisions in a split second quickly helps alleviate some of the inherent risks associated in the financial markets.
Harris cautioned that the study makes a connection between message traffic and future conditions, but it did not test whether a trading strategy could be implemented to take advantage of electronic messaging.
“We document evidence that message traffic at, and nearby, the inside quotes predicts upcoming price and quoted depth changes as much as 75 seconds in advance,” they wrote.
Knowing how information can be processed is critical in an industry where buyers and sellers are constantly finding out what direction markets or individual stocks are headed.
That flow of information is particularly important when time is measured in nanoseconds, and the numbers of transactions are so high a penny can translate into millions of dollars later.
“These are firms that buy and [sell] almost instantaneously. They don’t hold positions overnight,” said Christopher Singleton, managing director at Kanawha Capital Management, LLC, a wealth management firm in Richmond, VA. “Potentially, [high-frequency trading] can add to the volume of trading, but it certainly adds to the volatility.”
The caveat with the study, Harris said, is that the data is from 2003—which means those 15-second intervals and a simple assessment of message traffic are remnants of the past. While NASDAQ systems handled most all trades in 2003, trades today might be executed on more than a dozen systems, and at much faster speeds.
“One of [the] things we liked about the data we had is that it was a consolidated book, where NASDAQ did most of the trading in these stocks,” he said. “A high-frequency trader would now be looking at [information coming in] in seconds or milliseconds to figure the same things out.”
Today, a similar assessment would require looking at messaging from about 20 dark pools and about 17 decentralized exchanges, he said.
“It’s a much more complicated exercise now, and a much faster exercise,” Harris said. “What we wanted to do is see if we could demonstrate some ability to make this prediction.”
This isn’t the first time Harris, former chief economist at the U.S. Commodity Futures Trading Commission (CFTC,) has written about electronic trading.
In the mid-1990s, his examination of collusive activity among market traders led to a major restructuring of the NASDAQ.
This ability to project his research into urgent market situations continued. In 2005, Harris presented his research on electronic trading to the CFTC just as futures trading pits were giving way to electronic trading.
Harris says in recent years there has been an acceleration in technology applications within the financial industry, which has helped create a market where about 6 billion shares are being traded daily on U.S. exchanges.
“Ultimately, yelling across the floor is an outdated mode of communication in these markets,” Harris said.
“The Sound of Silence,” Jeffrey Harris and Mohsen Saad, was published in Financial Review in 2014.