Mondo Visione Worldwide Financial Markets Intelligence

FTSE Mondo Visione Exchanges Index:

Instinet Launches Cobra Algorithm

Date 23/01/2006

Instinet, a global equities execution agent, today launched Cobra, a stealth ordermanagement algorithm that opportunistically removes portions of liquidity when available, thereby leaving the stock’s spread unchanged. Performing like a seasoned trader, the algorithm will take into account a stock’s historical quoting trends to determine the best execution point, making the trade largely undetectable by the market.

“While there are certainly scenarios where people need algorithms based on ultra-sophisticated mathematical models, there are others where a trader essentially wants an algorithm to automate his normal trading behavior. Cobra does just that, as it opportunistically accesses liquidity to minimize market impact,” said Mike Plunkett, President, North America, Instinet.

Cobra, which is available immediately through Instinet’s NewportTM* and Instinet Trading Portal® front-ends, allows traders to choose among four different execution styles—passive, normal, aggressive and super aggressive. It remains virtually unnoticed by placing hidden orders in a completely randomized manner.

“Traders are struggling to meet best execution requirements in an environment of fragmented liquidity,” said Gavin Little -Gill, Research Area Director, TowerGroup. “The first generation of algorithms facilitated trading relative to benchmark goals, but TowerGroup is now seeing the advent of a second generation of algorithms that take smart order routing to the next level and one step up the value chain. Ultimately this is another tool in the toolbox and one that will facilitate the growth trajectory of algorithmic trading by the buy-side.”

The development of Cobra was prompted by the success of Instinet’s BlockPegSM algorithm, introduced in the summer of 2005. Replicating trading techniques that traders use in the OTC market, BlockPeg enables passive “pegging,” size-contingent order prices and discretionary pricing for the trading of NYSE-listed stocks. BlockPeg demonstrates the demand for algorithms whose functionality is more sophisticated than smart-routing but not as complex as the advanced mathematical models that users may struggle to understand.