To handle increased volumes as an equities firm and on its program trading desk, ABN AMRO has deployed a new rules-based platform to create and manage algorithmic-trading strategies.
The investment bank selected Apama UK, a developer of trading-strategy-management technologies based outside of London to supply the engine.
"We were looking to build an automated-trading tool completely integrated with our manual-trading system where we would give the traders a choice of automated-execution strategies or manual-trading strategies," says Gijsbert de Lange, head of Advanced Execution Services (AES) at ABN AMRO in London. The algo strategies are being used for equity trading clients and for proprietary trading as well, he says.
To support their existing program-trading infrastructure, ABN AMRO selected Apama's Event Manager; Scalability and Management Environment; and Platform Development Environment technologies. The product is scalable both in terms of being able to support expanding volumes and numbers of users, says de Lange.
According to de Lange, these are separate modules that interface to the bank's execution systems, to its graphical user interfaces and to the market data. The GUIs are visual displays so the traders can specify parameters for the algorithms.
"We've chosen for the moment to build our own GUIs," says de Lange. When the traders come up with algorithms, they work with quantitative analysts to program them.
Traders can create, manage and test their own trading strategies against live or stored data, and to monitor and react against multiple data streams, exchanges and reference data, according to the company's Web site.
Developed in a laboratory in Cambridge, England, the product is unique, says a spokesman for ABN AMRO AES, because it allows streams of data from various sources to be amalgamated and queried rather than stored in a database first and then queried from there.
After deploying a proof of concept in late 2003, ABN AMRO went live with the first version of the real-time trading platform in January 2004, and it is developing various iterations, specifically for its program-trading desk in Europe. At present, it's using the technology in Europe, but the bank plans to roll it out to its Asian hubs in the second half of the year.
While ABN AMRO intends to build the "higher end of the trading systems" in-house to capitalize on the proprietary knowledge of its traders and quantitative analysts, de Lange says, Apama offers a critical part of the architecture, which relates to sifting through various market-data feeds. As the feeds come in via the bank's normal interfaces with market data, that data is searched through based on whatever the algorithm specifies, says de Lange. "If the algorithm finds in the real-time data an event that it's looking for, it will then take an action. And the action in our context is buying or selling," he explains.
For example, if the bank was trading a program for hundreds of stocks simultaneously, the engine may look at the current order book, volumes or prices in the last three or five minutes, according to whatever rules the traders specify. "It has the ability to do that very fast," he adds.
Right now, the Apama technology is installed in the bank's program-trading environment but de Lange says, "We will start making use of the technology going forward for all of our equities trading."
Aware of the trend toward pushing out algo strategies to buy-side institutions, de Lange says ABN AMRO is running the algos for the clients at this point, though it is not giving them direct-access yet.
Once the trade is actually executing in the market, the trader may choose to retain a manual execution, opt for a complete automated execution or use some type of blended way of executing the order, de Lange says. Ivy is Editor-at-Large for Advanced Trading and Wall Street & Technology. Ivy is responsible for writing in-depth feature articles, daily blogs and news articles with a focus on automated trading in the capital markets. As an industry expert, Ivy has reported on a myriad ... View Full Bio