National Settlement Depository (NSD), a member of the Moscow Exchange Group, has, together with the ITBF Group, has automated the process of data parsing from Form W8-BEN.
Form W8-BEN is completed by non-US resident individuals who intend to receive dividend or interest income on U.S. securities and apply a reduced withholding rate. Once a client completes the Form, he/she sends it to his/her broker that, in its turn, delivers the Form to the upstream depository that acts as withholding agent withholding and paying the tax to the U.S. Internal Revenue Service.
An upsurge in interest in non-Russian stocks has resulted in a significant increase in the number of Forms W8-BEN received for processing and in higher requirements to the quality of data extracted from them. Before, NSD's employees used to analyze data in the Forms manually, which required substantial efforts and time. The need, therefore, arose to find an effective, quick and high-quality solution that would help extract necessary data from the Form automatically. ITFB Group was invited to join the project.
The parties chose an enterprise content management system based on the ITFB EasyDoc platform and designed for data scanning and recognition management, analysis of data parsed from documents, and data storage and search management.
ITFB Group has designed an EasyDoc-based app that recognizes Form W8-BEN attributes. To that end, the designers used the existing Forms to train a recognition model, set up rules for automated data verification, and integrate the model into NSD's system. The software fully automates the processing of Form W8-BEN, from the time it is received from a client until the data parsed and verified is uploaded to NSD’s system.
Pavel Andrianov, NSD's Executive Board member and IT Director: “The new solution has helped us to implement, within a short time, the automation project and reduce the load on the operations department, leaving only very poor-quality copies for them to process manually. The beta testing of the system has demonstrated a high quality of data recognition, while the quality of data is guaranteed by multiple automated data verifications. Moreover, the solution meets the security and availability standards. Investors don’t need to worry about security of their data.”
Yekaterina Golyanskaya, project manager, ITFB Group: “The main challenge we have faced when working on Form W8-BEN pattern recognition was a high variability of document representation format despite the seemingly similar document structure. To resolve the issue, in addition to applying the classic pattern recognition technology, we have used artificial intelligence elements, such as neural network training with the help of data provided by the project owner.”