SCIENCE

Stealthy Techniques for Causing Deliberate Micro Variations in Delay of Market Data and Trades Can be Used for Nefarious Intent

The CEO of cPacket Networks has raised a red flag about the potential risk of unfair manipulations of automatic trading systems by introducing stealthy and deliberate variations of the delay of market data and trading traffic as it goes over communications networks.

cPacket has shown in its lab how a “side channel attack” can be used to create micro-variations in the delay of market data and trades.  In “high frequency trading” (HFT) every microsecond counts – often making the difference between gains and losses that can total in the millions and billions of dollars.  Therefore, by being able to manipulate specific trading activities by several microseconds, an attacker could gain unfair trading advantage.

According to Dr. Rony Kay, CEO of cPacket, a leading expert on network performance monitoring and security, “We have been able to create micro-behaviors that could be used to target specific market data and trades feeding into automatic trading systems.  We believe that such techniques pose a substantial risk of creating unfair trading, if used by the wrong people.”

Kay explained that, similar to the way a stealth bomber sneaks up on its target “under-the-radar” without being detected, well-versed hackers might be able to manipulate high frequency trading systems by introducing micro-variations to traffic in virtual private networks that cannot be detected by commonly used network monitoring tools. “In the current environment, these side channel attacks could go unnoticed until it is too late.”

cPacket has become aware of this potential problem because it is focused on providing monitoring systems that can detect discrepancies in latency and jitter of data packets flowing through high speed networks. It has described these systems to senior financial and technology executives to enable them to optimize high frequency trading (see white paper athttp://www.cpacket.com/latency). Because of its expertise in this area, cPacket believes that the same monitoring techniques can be used to ensure that high frequency traders don't lose tremendous amounts of money due to unfair activities.

“It is critical that financial executives become aware of the minutest details of their data transfers to avoid problems,” said Kay. “Automatic trading platforms that use sophisticated algorithmic trading techniques and the traders that operate them must monitor network latency with the same diligence that they monitor market prices, trends, and news. Otherwise, the variability in their network performance could drown out the sophistication of their algorithms and expose them to latency arbitrage.  This is an issue that has potentially major implications.”