The global financial landscape is increasingly plagued by the complexities of financial crime compliance. As the severity of global money laundering grows, major financial institutions are entangled in a dense web of risks and regulations.
This scenario has given rise to ‘de-risking’, a self-preservatory practice where financial institutions terminate or restrict relationships with broad client categories to minimize potential financial crimes. This blunt approach, while reducing immediate risks, inflicts significant collateral damage on the global financial system.
Consilient, which offers the first federated learning technology for financial crime detection, recently examined the power of federated learning for de-risking.
Impact of De-Risking on Correspondent Banking
De-risking has severely impacted correspondent banking, essential for global finance operations, it said.
In correspondent banking, larger banks (correspondent banks) provide various services like wire transfers and trade finance for smaller or peer institutions (respondent banks). Observations from the UK’s Financial Conduct Authority and reports by the Bank of International Settlements have confirmed a worrying 22% decline in active correspondent banks globally between 2011 and 2019.
This contraction reflects correspondent banks’ hesitance to manage the risks posed by clients of respondent banks, leading to a widespread retreat from these relationships, particularly affecting institutions and beneficiaries in developing nations.
The High Cost of De-Risking for Developing Nations
The consequences of de-risking are particularly harsh for respondent banks in less developed countries, often viewed as high-risk environments. A 2017 survey by the Caribbean Association of Banks highlighted that a significant number of Caribbean banks had lost correspondent banking relationships, severely hindering their international financial connectivity. Similar trends were observed in Africa and Angola, where a substantial reduction in foreign banking partnerships was reported, undermining economic growth and financial inclusion in these regions.
Introducing Federated Learning: A New Approach to Risk Management
Federated Learning offers a groundbreaking method for training machine learning models across multiple decentralized sites without moving sensitive data. This approach leverages local data to develop a unified model that encapsulates diverse insights, enabling nuanced risk management strategies. By keeping data local and mobilizing models, Federated Learning addresses privacy concerns and enhances data security, making it a formidable tool against financial crime, Consilient explained.
How Federated Learning Combats De-Risking
Federated Learning enables correspondent banks in high-risk areas to enhance their risk management capabilities, it said. This technology provides banks with better control and assurance over the financial crime risks by allowing them to directly engage with and refine the shared models. Such involvement ensures that the models align well with the banks’ regulatory needs and risk tolerance, fostering more precise and robust financial crime detection.
Benefits of Adopting Federated Learning
Adopting Federated Learning can reverse the adverse effects of de-risking. For correspondent banks, it opens up avenues for business with high-risk jurisdictions, thereby diversifying their portfolios and boosting revenue. This method also supports economic growth and financial inclusion by enabling legitimate businesses, particularly in smaller economies reliant on sectors like tourism and trade. Moreover, by maintaining transparency and combating the shift towards unregulated financial channels, Federated Learning upholds the integrity of the global financial system.
Read the story here.
Keep up with all the latest FinTech news here
Copyright © 2024 FinTech Global
Copyright © 2018 RegTech Analyst
