Next-Generation of Advanced Algorithms to be Centered on Convergence of Data
NEW YORK, NY, February 28, 2005 â According to TABB Group in a new industry report released today, âOutlook on Algorithms: New Developments in Electronic Tradingâ, the next generation of algorithms will be far more flexible, react to unexpected events, handle highly complex relationships and begin to manage part of the investment process. Underlying this new wave of algorithms is the need for high-speed order processing and low-latency connectivity driving many large sell-side firms, such as Merrill Lynch, to upgrade and expand their market data infrastructure.
âTechnology is critical to the evolution of the algorithmic space,â writes Adam Sussman, report author and TABB Group senior consultant. âThe sophisticated, single-stock algorithms now being developed need to analyze high velocity, mass volumes of real-time data, as well as historical transaction cost research (TCR) data, and run through a complex set of decision trees.â However, as global equity traders and traders of various asset classes adopt more electronic trading tools, he explains, their firmâs technology infrastructure must be ready for huge increases in bandwidth and storage requirements, or risk losing to their competition.
One of the critical shortcomings of the first generation algorithms was their inability to work in low-liquidity situations. âTo make advanced execution models more successful at trading small and mid-cap stocks, new algorithms are working with crossing networks to find hidden liquidity. This increases the algorithmâs effectiveness while lowering transaction cost,â says Sussman.
New algorithms are also being developed to extend models beyond single stocks into portfolios and baskets. These tools will further align the investment decision and the trading processes by helping the trader manage multi-dimensional risk, Sussman writes. These algorithms will formulate the appropriate trading strategy to minimize costs and maximize portfolio yield. Additionally, models will be developed to trade portfolio segments simultaneously and eventually incorporate analytics such as Value at Risk (VAR), exposure, volatility and expected returns. This will elevate the trading model from executing a trade to managing a complex transition.
âThe new wave of algorithms is really about the convergence of data,â adds Larry Tabb, CEO at TABB Group. âInformation across multiple disciplines must come together to take trading-oriented algorithms to the next level, including real-time market data, technical analysis, quantitative research, TCR and risk management.â
Key findings from the 26-page report based on conversations with leading developers and users in the electronic and algorithmic space include:
Â· UBS and Liquidnet have each introduced systems that connect retail and institutional order flow. TABB Group expects further consolidation among retail brokers, institutional broker/dealers and ATSs hoping to create similar order flow profiles.
Â· Based on projected order flow allocations and commission data from 2005, TABB estimates industry-wide equity commission revenue could drop by 8%, from US$19.3 billion to $17.8 billion, mostly due to the shift towards low-cost execution venues.
Â· Brokers increased investments in electronic trading technology from 2003 to 2005 from $350 million to $790 million for algorithms, trading portals and TCR. TABB Group sees the rate of increase leveling to $970 million by 2007 due to market saturation and harder-to-obtain market share.
Â· Next-gen liquidity-seeking algorithms will use more advanced tactics to âsniff outâ executions and reduce transaction costs. Brokers will use historical data to determine where and when liquidity is likely to appear.