Pipeline Trading Systems, the electronic institutional brokerage firm, has released a streaming, at-trade analytic tool to help traders maximise alpha capture and minimise adverse selection.
Alpha Pro brings predictive analytics directly to a buy-side trader's desktop, providing actionable intelligence immediately on order arrival.
Incoming orders are analysed in light of current market conditions and historical patterns, and factors most likely to predict impact-free price movement are identified.
Taking these statistical factors into account, the service then recommends a trading strategy predicted to maximise alpha capture and minimise adverse selection.
“When high-frequency trading firms run strategies like short-term arbitrage opportunities, it tends to accelerate the execution of institutional algorithms when the market conditions are most likely to be adverse to the trade in the short term,” Henri Waelbroeck, Pipeline's director of research, told theTRADEnews.com. “If the market appears to be heading down, the institutions will buy faster and vice versa.”
Buy-side firms have information that can be used to understand short-term alpha, says Waelbroeck, such as their own trade decision processes, the historical records of portfolio managers trades and the firm's decision process which is not known to high-frequency trading firms. That information gives them a natural advantage but it tends to get lost between the portfolio manager and the trading desk.
As the trade proceeds, the Alpha Pro analytics stream mines this historical information, with market feedback from the execution process and feeds including news and real-time order flow analysis.
It looks for insights as to when trades are more likely to have alpha and provides updates to users on changes in the alpha outlook helping traders to make informed decisions on the execution schedule they would like to adopt.
Once the firm has set out an execution schedule, controlling algorithms are used to ensure the right algorithm is used to keep the strategy on track. Typically if a firm uses an algorithm that performs at a 10% participation rate on average then when the market moves against it the algorithm will perform at closer to 5% participation rate, and if the market moves with the firm the algorithm will perform at closer to 20%. Waelbroeck says that the purpose of algorithm management is to help the firm take away the market's control over execution performance, by using algorithms that will perform correctly in the current market situation.
“Our Alpha Pro clients have reduced implementation shortfall losses by typically 20-30%, and in some cases up to 65%.” he continues. “This shows it's possible to reverse adverse selection by taking a well-grounded view on short-term alpha, then using predictive algorithm control to execute this view efficiently.”