Put a Lid on It: Measuring Trade Information Leakage

Traders Magazine Online News, August 2, 2017

Ben Polidore

For institutional traders, striving to achieve best execution and improve investment returns, information leakage is a constant challenge. In the current U.S. market structure, with 13 exchanges and 30+ other execution venues, it is difficult to attribute information leakage to any single venue because most modern routers create routes to many venues proximate to any price move. Since it is possible to have information leakage without an actual fill on an order —the most obvious example being lit quotes on exchanges—any of the venues used before a price move could have caused the price move.

Despite the difficulty in pinpointing the precise source of information leakage, traders are clearly convinced that it is a major problem which weighs on performance. In a recent ITG poll of more than 80 buyside traders, more than a third of respondents (37%) estimated that information leakage comprised more than half of their overall trading costs.  

According to almost half (47%) of the traders surveyed, the top source of this information leakage is not human traders but rather schedule-based algos (e.g. VWAP, TWAP).  The second most common source of leakage cited in the poll was cash desks (i.e. high-touch sales traders) at 33%. Dark trading was seen as less leaky, with far fewer respondents citing dark algos (13%) and block networks (7%) as key sources of information leakage.

The traders surveyed were almost unanimous (95%) in their view that high-frequency traders benefit from information leakage. More than half of those polled believe that sell-side risk desks can benefit from information leakage, while 43% cited other buyside firms as beneficiaries of information leakage.

Regardless of the source, and the beneficiaries, of information leakage, it is clear that it can result in significant underperformance, particularly for large orders. In general, dark pools reduce information leakage relative to alternatives such as high-touch traders, block positioning and exchange trading. However, market fragmentation creates a large surface area for potential problems.  A randomized, controlled measurement of information leakage on a venue-by-venue basis can yield important insights into the trading process and drive improvements in routing, algorithm design and selection and trading strategy.  

 

Ben Polidore is Managing Director, Algorithmic Product Management at ITG

For more information on related topics, visit the following channels:

• Insights/Viewpoints