By Nick Carey - Analysis
CHICAGO (Reuters) - In the world of automated trading, fortunes are made in less than the blink of an eye.
That wealth is generated on computer systems that can handle greater trading volumes at ever increasing speeds. These platforms often rely on algorithms — a sequence of instructions used for calculation and data processing — that can spot unseen opportunities in the market and give their users a huge advantage measured in milliseconds.
For banks such as Goldman Sachs Group Inc (GS.N), the codes are worth a fortune and this value also make them a tempting target for thieves — as appears to have happened with Sergey Aleynikov, a former computer programer at Goldman arrested by the U.S. Federal Bureau of Investigation last Friday.
“The risks of trying to steal a trading model are very high,” said David Easthope, a senior analyst at Celent, which is part of the Oliver Wyman Group. “But the potential reward is very, very high because when you look at investment banks, they’ve made higher and higher profits” from automated trading.
Goldman in particular has had a “great last few months” thanks to a trading model that has enabled it to “take the right risk at the right time,” Easthope added.
“If you happen to have built a model that makes you even 200 milliseconds faster than me, you are in a much better position,” said Sang Lee, managing director of research and advisory firm Aite Group. “You may be able to trade a thousand times before I even see what’s going on and act on it.”
It takes 300 to 400 milliseconds for human eyes to blink and it’s that advantage that encourages thievery, such as allegedly committed by Aleynikov.
“These systems have a huge value and, if they get out, then you lose your edge,” said Michael Gorham, director of the center for financial markets at the Illinois Institute of Technology.
The Goldman Sachs case is unusual because it is far more common for algorithm experts to design a system, walk away with the information in their head and build an identical system elsewhere — which can lead to civil lawsuits over intellectual property rights, but not charges of theft.
“But you don’t often hear of people actually stealing the data in question,” Gorham added.
Analysts say increased financial regulation and oversight proposed by the administration of U.S. President Barack Obama — the largest regulatory overhaul since the Great Depression if it makes it through the Congress — cannot prevent data theft of this kind.
Instead, the hedge funds, internal trading desks at banks and independent proprietary trading firms that use trading platforms are expected to beef up internal security to try to minimize the risk.
According to Professor Adam Galinsky of the Kellogg School of Management, taking the kind of risk Aleynikov is accused of is not uncommon among people given power — in this case over Goldman Sachs’ trading platform.
“Power alters the basic neurological processes in the brain and inhibits those parts of the brain that would allow a person to show restraint,” he said. “It allows them to systematically ignore the consequences of their actions.”
Galinsky referred to past examples of traders gone wild to demonstrate the effect power can have. For example, futures trader Nick Leeson, who lost $1.4 billion and brought down Barings Plc, one of Britain’s oldest banks, in early 1995. Or Jerome Kerviel, whose unauthorized trades caused a loss of 4.9 billion euros ($6.87 billion) at French bank Societe Generale (SOGN.PA).
The main difference in the latest case is that Aleynikov was not a trader, a fact that Robert Friedman, a partner in the litigation group at law firm Kelley Drye & Warren LLP said would make additional regulation redundant.
“Regulation is more geared toward transparency in the markets rather than preventing outright theft,” he said. “It is very difficult to prevent theft before it happens, but it appears Goldman Sachs had controls in place that worked.”
Others see the episode as a cautionary note, not just for Goldman, but the rest of the market as well.
“If anyone has an algorithm that’s working well, it’s very valuable,” said Iqbal Brainch, chief marketing officer at futures brokerage firm Advantage Futures — whose website touts that it is “millisecond to none.”
“This case speaks to the importance of security, the importance of not letting any of that information out — and the importance of putting steps in place to protect your algorithms.”
Reporting by Nick Carey; additional reporting by Doris Frankel and Kyle Peterson; editing by Andre Grenon