2nd CEPR-Imperial-Plato Market Innovator (MI3) Conference 2018 – Order splitting and searching for a counterparty



2nd CEPR-Imperial-Plato Market Innovator (MI3) Conference 2018

Order splitting and searching for a counterparty

Vincent van Kervel (Pontificia Universidad Catolica de Chile)
Amy Kwan (University of Sydney)
Joakim Westerholm (University of Sydney)


Vincent Van Kernel is an assistant professor at the Pontificia Universidad Catolica de Chile. He has previously published research on “high-frequency trading around large institutional orders”, “competition for order flow with fast and slow traders” and “The impact of dark trading and visible fragmentation on market quality.”


Amy Kwan is a lecturer at the University of Sydney, where she joined in 2014. Her research areas are mostly concerned with market microstructure and corporate finance. She holds a PhD from the University of New South Wales and a Bachelor’s degree form the University of Western Australia.


Joakim is an Associate Professor at the University of Sydney. Joakim’s teaching and research interests are in the areas of Asset Pricing with focus on security market microstructure and behavioural finance topics alongside corporate finance with focus on CEO and corporate insider trading strategies and acquisition decisions.


About the paper


Large and patient traders already have strong incentives to find natural counter-parties to allow for larger orders to be processed at lower costs. The challenge is that buy- and sell-side must coordinate to not only trade, but to initially find each other.


The theoretical model put forward by Vincent, Amy and Joakim shows that searching for a counterparty can be done by means of order-splitting, where a large quantity is broken up into individual trades which are executed over a longer time.


This facilitates coordination as it indicates trading interest to the market and helps detect the presence of counterparties.


The paper confirms empirically that the presence of counterparties for trade execution can be detected in real-time, and that the behaviour of counterparties affects parent order characteristics. For instance, for a one-standard deviation increase in volume, order sizes are approximately 17% larger and have an implementation shortfall 7.6 basis points lower.


Main Takeaways


Order splitting gradually reveals trading intentions to the market. If a natural counterparty is present (i.e. a trader with a different private value) then he will also signal his presence through order splitting. As both traders observe order flows and gradually infer each other’s private values, they can start trading larger amounts without moving prices – realising larger gains from trades.


That said, if all traders in the market have similar private values, the trading pressure will quickly move the price towards this value and profitable trading opportunities will disappear.


The interesting aspect of this model is that each trader tries to signal his own private value and learn the private value of the other. If all things were equal, both traders provide liquidity to each other, and this provision gets stronger each time they trade with each other.




The interaction between strategic investors is increasing in importance with the rise of order splitting algorithms and other trading strategies.


In their paper, Vincent, Amy and Joakim present a model where two liquidity motivated traders signal and learn about each other’s private values in order to maximise gains from trades. They are able to effectively coordinate and accommodate their separate orders in a process where trades by one push back the price pressures caused by the other.


As markets evolve and complex execution strategies become increasingly automated, there is a need for more work to provide guidelines on how institutional investors should adapt.