In today's electronic markets, billions of dollars are changing hands in milliseconds. Thanks to the professional investors, who trade seamlessly across geographical markets and asset classes. They are using the automated trading strategies based on algorithms to maximise profit. These profound changes, combined with an unprecedented bull market in stocks, have created a massive explosion in trading activity, which threatens to overburden the information systems that process client orders to buy and sell investments at banks and brokerages. With the alarm bells sounding, these financial services firms have begun to invest in new technologies to build trading platforms capable of handling a quantum increase in transactions.
Global brokerage and investment banking house CLSA is currently in the process of doing just that. It has been building its next generation trading platform throughout this year, aiming to cope with the expected surge in transactions.
"What we have seen is a huge burst of trading activity in a short period of time, which has made latency, performance and the ability to crunch an enormous amount of data critical to the trading system," says T. Rajah, CIO of CLSA. "The systems we have in place are still perfectly capable of coping with current transaction volumes and the expected increase in transaction volumes over 12 months. But clearly we need a trading platform with far better performance and scalability looking further out, if transaction volumes continue to grow along the current trajectory."
The wide discrepancy between the rate of growth in the volume of financial instruments being traded and the volume of transactions is down to a change in the way investors have been trading, Rajah says. While the growth in the volume of financial instruments being traded has been driven by more money pouring into buoyant financial markets, the growth in transaction volumes is being amplified by the growing use of algorithmic trading.
As with many investment banks, CLSA has seen its transaction volumes increase by 500 to 600 per cent in the last 12 months alone, whereas as turnover on the global stock market-namely the number of shares changing hands in stock markets-over the same period has increased by a relatively tame 50 per cent.
"Algorithms are being used on two levels. Firstly, clients are using them to manage portfolios and trade orders, and secondly financial service providers are using them to execute trade orders," Rajah says.
Based on algorithms, the new trading strategy helps investors get the best price or avoid detection by other investors when they trade a large tranche of stocks or other financial instruments. The program splits the trade into a series of smaller buy or sell orders executed over a period, based on the interpretation of real-time market data.
By chopping up big trades into lots of smaller trades, these automated systems have raised the number of transactions executed across trading platforms in the financial services industry through the roof. According to a market report on messaging in the financial services sector published by Red Hat, the volume of market and trade messages travelling over trading systems has increased by 2,500 per cent over the past five years.
With the explosion in transaction volumes, delivering bandwidth efficiency, along with improvements in performance, latency and reliability, became critical for CLSA and the need to overhaul the trading platform became apparent. The firm decided to work from the foundations up, beginning with the installation of a new enterprise messaging system (EMS). Supported by business integration vendor TIBCO, the EMS serves as the infrastructure upon which real-time information could flow among the various information systems on the trading platform.
One advantage of the new EMS, according to Rajah, is the use of publish/subscribe messaging systems to distribute information more efficiently across the network. With EMS, a trading market data feed system does not need to establish a separate connection with other related applications, like the automated trading program or a trader's terminal, in order to send the information across the network multiple times. Instead, the trading market data feed system needs only to classify the information and send it to the EMS, which will then direct the appropriate information to systems that are integrated with the EMS.
Binding systems together
Currently at CLSA, the EMS is binding together the various information systems to support the firm's trading platform. These include applications in the front office, which perform tasks such as obtaining market price data, performing complex quantitative analyses and executing transactions, to systems in the back office which manage compliance requirements, reconciliation of funds transfers, and securities settlement.
As the central nervous system of the trading platform, Rajah says the enterprise messaging system was seen as one of the most critical parts of the project and thus chosen only after a thorough assessment of the available solutions. "We wanted the fastest and most reliable available messaging engine," Rajah says. "But it was also important for us to find a messaging platform that was based on open standards and flexible, as our company has a multi-vendor environment along with in-house developed systems."
Through TIBCO's EMS toolkit or standards-based model, CLSA was looking to develop a trading platform that could be changed on the fly, says Rajah. The firm can then respond quickly to changes in business needs without any reprogramming, allowing CLSA "to focus on building the logic, not the infrastructure".
Ultimately the firm is looking to develop a trading platform that is future-proof to another quantum leap in transaction volumes and provide the backbone for further expansion, says Rajah. "Basically we want to build a platform to handle almost any eventuality going forward."
Since CLSA completed its EMS production selection in the first quarter of 2007, it began the implementation in the second quarter. There are a number of components available from the enhanced trading platform, including the infrastructure to handle market data and a system monitoring tool. Another significant phase of the project is expected to be completed by the end of the year, when the infrastructure will be able to support program trading. Rajah says he expects the whole trading platform to roll out by the end of next year.
Once completed, the new trading platform is expected to provide traders at CLSA a single view for trade execution and investment management. According to Rajah, the new system will allow traders to work more flexibly by switching seamlessly from manual trading to any number of automated algorithmic trading strategies. Traders can also receive real-time feedback and analytics showing the performance of individual stocks as well as a whole portfolio of investments.
With the provision of real-time information of each transaction and its impact, Rajah says the new trading platform will also facilitate better client servicing.
© Fairfax Business Media
Join the CIO New Zealand group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.