A new service, from custodian BNY Mellon to its market-participant clients, helps these avoid about 40–50 percent of their settlement fails. In PostTrade 360° Helsinki on Wednesday, digital business leader Victor O’Laughlen shared his inside view.
Some of BNY Mellon’s large clients pay 10–20 million dollars a year for failing settlements, and then inefficiencies in the daily cash allocation comes on top. A problem, yes. But also a great case for machine learning, if BNY Mellon’s Victor O’Laughlen is to believe.
His session in the Helsinki conference went through the case – but also a number of principles for how to set up projects and solutions in the growing area.
“In my view many underestimate the impact that this area will have,” says Victor O’Laughlen.
The purpose has been to build a machine learning model that brings insights you can’t achieve with traditional coding models. Machine learning is one category of artificial intelligence.
”Feature engineering” is a key part of setting up the solution. It is important to understand the sources of data, and the drivers of fails. This takes a somewhat science-like work process, where the stories of the staff are used to generate hypotheses which are then evaluated against data.
• News from the PostTrade 360° Helsinki event is gathered here.
• The 32-page pdf magazine, which includes the agenda, can be downloaded by clicking here.
• For a 3-page breakout of the agenda section, click here.
• By the way … are we connected on LinkedIn already, among the 2,000 post-trade pros who are? Follow us here.
