So far six yet un-named asset managers, two hedge funds and two pension funds are using the new system to develop and execute trading strategies using third-party trading algorithms available to the respective firms.
Once a trader enters a parent order, Alpha Vision analyzes the firm’s historical trade database, or a generic one database provided by Portware, to find similar executions. Then it develops a trading strategy based on this data using 45 different factors in its decision-making process.
Once the trader accepts the strategy, Alpha Vision executes and monitors the strategy’s performance. After each child order transacts, that performance data is fed back into the system to hone the strategy further. If a change is necessary, the platform tweaks the strategy while duly logging those changes.
To prevent the strategy from spiraling out of control if the markets take an unexpected turn, the platform knows when to stop trading and alerts the user. The trader then can instruct the system to develop a new strategy base on the new market conditions and execute the outstanding order against it.
According to Portware CEO Alfred Eskender, Alpha Vision’s early adopters saw the platform updating its trading strategy on average 42 times compared to the typical four tweaks that a human trader might make after the commencing an execution strategy. This difference in strategy monitoring has translated into a 17-basis point improvement in executions done via the platform, he claims.
The offering might sound like a “trading desk in a box,” but Portware officials are quick to refer to it as “more of a co-pilot than a pilot.”
Although the system may not replace the trader, the performance improvements it provides just might alter some commission-sharing agreements as research brokers could use it to improve their executions and attract more direct client order flow, say vendor officials.