(Bloomberg) — High-frequency trading in the U.S. Treasury market makes it harder for investors to buy and sell bonds when news events cause prices to move, according to the Bank of Canada.
Researchers studied how liquidity changed as algorithmic- trading programs, which are much faster than human beings, reacted to economic indicators including the unemployment rate, home sales and inflation. They examined trades on ICAP Plcs BrokerTec trading system from January 2006 to December 2011, putting their findings in a paper released this month. The $12.3 trillion Treasury market is the worlds largest debt market.
High-frequency trades and orders tend to have a negative impact on liquidity around public-information arrival, researchers George Jiang, Ingrid Lo and Giorgio Valente wrote in the paper. After news is announced, they significantly reduce depth, both at the best quotes and behind the best quotes.
When economic data is not being released, they improve price efficiency, the researchers wrote in a working paper. That means automated trading narrows the difference between prices to buy and sell bonds.
In 2011, Michael Spencer, the CEO of ICAP, said that high- frequency trading, or HFT, accounted for slightly less than 55 percent of the volume on BrokerTec.
Public scrutiny of Treasury trading has intensified since Oct. 15, when a sudden surge in bond prices caused the yield on 10-year notes to drop the most since 2009. The event highlighted the urgency of understanding what is happening in one of the worlds biggest markets.
Lightly Regulated
Light regulation for trading U.S. government debt has created the space for possible market manipulation. Executives at three of the largest market-making firms raised concerns earlier month. They told Bloomberg News that an electronic bait- and-switch tactic known as spoofing — already the focus of alleged manipulation at a futures exchange — needs to be investigated in cash Treasuries.
Spoofers make money by feigning interest in buying or selling at a certain price, creating the illusion of demand in an attempt to make other traders move the market. The spoofer cancels the original order and buys or sells at the new price to make a profit. The practice is sometimes called pull and hit.
The rules and responsibilities governing Treasuries are split between the U.S. Treasury Department, Securities and Exchange Commission, the Federal Reserve Bank of New York and others. Rules first enacted by Congress in 1986 have gone virtually untouched for a decade and a half, creating loopholes that allow algorithmic traders to outmaneuver less-savvy rivals.
Beyond Capacity
The Bank of Canada report didnt examine the subject of market manipulation. The researchers used tick data from BrokerTec for two-year, five-year and 10-year Treasury notes. They defined high-frequency trades as those made, canceled or altered at a speed deemed beyond manual capacity, according to the working paper.
They also used ICAPs i-Cross co-location service, which reduces the time it takes for HFT firms to execute their orders but does not have any other direct impact on market quality to identify the effect of HFT on the market, the researchers wrote.
To contact the reporter on this story: Matthew Leising in New York at mleising@bloomberg.net To contact the editors responsible for this story: Nick Baker at nbaker7@bloomberg.net Will Hadfield