Raging Against the Algo Machines
Following on from the recent high-profile algorithm-related market failures, including the Knight Capital and Facebook disasters, calls are growing to develop new methods to improve the stability of financial markets.
Automated and high-frequency trading is in the spotlight after a faulty algorithm nearly bankrupted U.S. market maker Knight Capital earlier this month, while in May trading glitches marred social networking site Facebook’s initial public offering on the Nasdaq stock exchange in New York.
“To be able to monitor the automated world, you must do it in an automated fashion in real time, or as near to real time as possible,” Justin Amos, co-founder and chief operating officer of Redkite Financial Markets, a U.K. market surveillance provider, told Markets Media.
However, with the rise in computational power and the increasing reliance by traders and investors to use machines to execute their trading philosophies and gain an advantage over the market, it appears that regulators will have to change their approach to prevent future trading mishaps.
“All market participants agree that market software needs to be more robust,” said Ted Aronson, managing principal at AJO Partners, an institutional investment manager based in Philadelphia. “We can’t tolerate the erosion of investor confidence.”
Regulators are only just waking up to the dangers. In Europe, the European Securities and Markets Authority (Esma), the pan-European regulator, is introducing new guidelines for automated trading to better monitor algorithmic trading practices such as high-frequency trading by urging firms to monitor in real time their electronic trading systems. It is believed that if the Esma rules are a success other jurisdictions across the globe will adopt the rules. Some market participants are worried, though, that the new guidelines do not go far enough.
“Real time isn’t clearly defined yet [by Esma] so some firms take the view that T+1 is close enough to real time, and for their purposes and how active they are in the markets that might be true in the short term,” Magnus Almqvist, senior product specialist at SunGard’s capital markets business, a trading and technology firm, told Markets Media in June.
“But the trend towards more efficient and more immediate surveillance is there and I am expecting firms to increasingly get into the seconds or minutes after the market event to analyze data and trigger alerts.”
In the recent cases involving Knight and Facebook, for instance, trading has been left unchecked due to the lack of regulations and this has resulted in severe problems.
A recent academic project by the Center for Innovative Financial Technology (CIFT) at the Lawrence Berkeley National Laboratory in California says that supercomputing and data intensive science could provide the help regulators need to monitor the complex markets of today.
Motivated by the U.S. ‘flash crash’ of May 2010, when the Dow Jones Industrial Average index plunged 1,000 points, almost 9%, only to recover within minutes, the researchers at Berkeley Lab said the flash crash was something of a wake-up call for markets, built on real computer networks, that were capable of unanticipated and dangerous behavior.
“There are many ways existing supercomputer computing systems are advantageous to regulation and enforcement,” said David Leinweber, director of CIFT. “They remove all of the data size and computation speed limits for these functions. The need for improved analysis, simulation and testing of market system integrity has been demonstrated repeatedly by a series of market mishaps. There is no algorithm certification of any sort today. In virtually all other complex systems, modeling and simulation play a central role. It’s not easy to do right, but with enough horsepower it becomes feasible to consider.”
Marcos Lopez de Prado, head of global quantitative research at Tudor Investment Corporation, a U.S. hedge fund which donated funds to the CIFT project, added: “Those responsible for market oversight could benefit from real-time ability to effectively monitor a complex system. Recent events, including the flash crash and other market disruptions, have highlighted the need to solve potential inadequacies in market structure and execution.”
Algorithms have become more prevalent in the spot FX market.
QB’s Algo Suite for futures market trade execution is also being co-located to HKEX.
Breaking data silos is key to deploying automation beyond 'nuisance' orders.
They can be used on quantum hardware expected to be available in 5 to 10 years.
Streaming blocks change the basis of matching and price discovery so institutions can find new liquidity.