Japan's Mizuho Bank has been working with Nomura Research Institute (NRI) on the use of artificial intelligence to eliminate lengthy and complex derivative contract procedures.
Traditionally, the standard derivative contract procedure at Mizuho includes the exchange of print ISDA contracts with clients via FAX and PDF and the manual archiving of all the information.
Each contract contains an enormous amount of unique information, meaning that registering it to the system and referring to past contracts requires specialised knowledge and time-consuming manual input.
The bank enlisted NRI for a now-completed proof-of-concept on how AI can make the process more efficient. The test found that Optical Character Recognition can digitize ISDA contracts and the system can define, extract, and archive the sum total of information from the digitised contract into a format that can be easily accessed and searched after the fact.
As a result, the time it takes for analysts to register ISDA contract information into the system and refer to past information is significantly decreased.
Mizuho says it will carry on working with NRI to develop a full-fledged system for all derivative contract operations.
Masahiro Ueki, SVP, legal and compliance team, Mizuho, says: "In addition to the improved operation efficiency brought by the AI implementation, we expect these exciting new systems can also maintain a high-level understanding of ISDA contract operations across the board."