Smart algorithms could cut trading costs

Smart algorithms could cut trading costs
Smart algorithms could cut trading costs

In search of operational cost savings, portfolio managers should take a hard look at their execution and empower their equity traders to select algorithms that are optimal to the task, according to Ofir Gefen, co-head of execution, Asia Pacific at Investment Technology Group (ITG). 

Research produced by ITG, a brokerage and a provider of financial technology and analytics, shows that traders tend to pick familiar trading strategies and providers, even if they are not optimal for the type of order, security and market conditions they are faced with. 

While quantifying possible cost savings is difficult, since using a bad strategy drives the stock's momentum which cannot easily be disentangled from other factors, Gefen claims that potential for cost reduction is substantial.

The notion of identifying and utilizing optimal trading algorithms has been given too little weight among Asian investors, claims Gefen, basing his observations on data collected by ITG's analytics team and on his interactions with the region's investors. 

"Portfolio managers see trading as an afterthought, a cost centre," he says, "and that's not the right way to think about it". They have little understanding which of the many trading algorithms should be used, and which providers offer the best implementation. 

Execution traders, who have some of that knowledge, frequently simply follow portfolio managers’ directions rather than use their skills.  "If the traders don't feel they are empowered, that they are responsible for execution that adds value to the investment process, why would they put any thought into it?," asks Gefen.

While choosing an optimal trading strategy for each order is bound to generate cost savings in the long term, traders often prefer sticking to a few common algorithms and providers, especially when choices are overwhelming.  They also prefer to follow portfolio managers’ instructions rather than risk being blamed for using a different strategy if the market moves against them.  

Besides a change of attitudes, empowering traders to use their skills and judgement in executing orders, analytics is needed to identify optimal algorithms and to measure the performance of their providers. ITG provides this type of analytical tool, an "algo wheel", which Gefen says aims to optimize execution and reduce implementation costs. 

The algo wheel randomizes the brokers' selection of execution algorithms among trading strategies and providers, and measures their performance.  As the data accumulates, it becomes clear which algorithms are optimal for a given order type, country or market sentiment. 

Gefen hopes that by using technology like this, the market in Asia can develop to the point where a feedback loop can exist on the trading desk. “The next stage is that you start integrating that feedback into your trading decisions and actually have that information help drive your decisions,” he says.