Trade finance back in office back in France Thanks to AI?

In fact, the development of artificial intelligence (AI) technology makes this organizational model less profitable, and some market players are already planning to relocate their trade finance back-office operations to France.

Banks, since the 2000s, have shifted their trade finance back-office operations to CSP.

Most international banking institutions use CSPs to process their back-office operations, especially trade finance. Made in the United States in the 1980s and adopted by French companies in the early 21st century, these autonomous entities pool the processing of banking institutional documentary credits and international guarantees on a regional or global scale.

This centralization of trade finance back-office activities has many advantages in a single place, especially thanks to the scale economy induced by the processing of a large number of operations including reduction of process costs.

This cost reduction is often reinforced by relocating the CSP to a country where labor is less expensive than in France. Poland and Romania are frequent destinations in Europe or India where labor costs are 4 to 5 times lower than in France. Some banks use CSP based in India for back-office processing of their trade finance transactions. They only capture activities that require greater expertise locally, such as the legitimacy of the transaction or permanent control by sampling. Similarly, the direct relationship with the customer, the need for a detailed understanding of the local market, remains in the hands of the business parties in France.

Made possible for digital tools, this use of CSP has also accelerated the digitization of trade finance activities.

Thanks to digital tools it is possible to split processes in different places. They allow information to be shared between different stakeholders in the process, even located thousands of kilometers away from each other. For example, CSPs based in India are working on digital versions of documentary bundles, which were obtained in paper format in France and then scanned. This distribution of activities has certainly created a very strong need for coordination between delocalized back-office and internal teams. There was a need to set up digital tools for managing and monitoring flows that would allow for the final processing of a process in a decentralized manner. This new organizational model, directly as a result of the digital revolution, has also contributed greatly to the increasing digitization of trade finance activities.

They are now entering a new phase of technological development: artificial intelligence. At several local banks, projects are underway to fully automate the analysis of documentary bundles. Once scanned, the bundles are “read” by Character Recognition Software (OCR). The extracted data will automatically fill in the corresponding fields of the transaction processing tool using an API or a robot (RPA: robot process automation). At the moment there are only two activities left in the process, performed only by humans: scanning of documentary bundles, which have not yet been dematerialized, and validation of operations after verification that the data that has been automatically dumped is correct. However, if the AI ​​results are satisfactory, manual verification will probably no longer be required and only a sample will be tested later.

The technological leap towards AI allows the transfer of trade finance value chains.

If the entry cost of such a project can be significant, the use of AI technology leads to a significant performance gain and considerable reduction in costs. Based on algorithms, these new tools are capable of handling huge amounts of transactions with an infinite number of errors. As a result, their use often results in lower costs than using a CSP, as shown in a 2016 study by Everest Group. As a result, banks have begun to consider the return of all activities of trade finance value. Chains within their local business teams, the use of more efficient digital tools should allow substantial gains in productivity so that outsourcing of back-office operations is no longer profitable. One such project was launched in 2017 on HSBC.

Thus, while AI will ultimately allow for the transfer of trade finance back-office activities, this is also not true of jobs. In fact, tasks currently performed by CSPs will be partial or fully automated. Humans will often have a role to play in the maintenance of the instrument, no matter how important. The qualifications of back-office managers will be questioned in both France and the CSP. Furthermore, a survey conducted by IBM in 2019 estimated that more than 2 million employees in France, and 120 million worldwide, would have to re-qualify in the near future due to the use of robots and AI.

Leave a Comment