3 Applications of Machine Learning in Banking Risk Management
Artificial intelligence has been finally recognised as the technology that can transform banks’ critical functions.
From chatbots to credit underwriting to stock market predictions, there is no shortage of use cases of machine learning in banking.
Despite the fact that risk management has always been at the top of banks’ agenda, many processes are still plagued with inefficiencies that are continuously draining bank resources. In this article, Andrey Koptelov discusses how banks can apply machine learning to streamline regulatory risk management and advance their fraud detection methods.
Streamlining regulatory change management
In the banking context, risk management and regulatory compliance are closely aligned. Banking employees have to manually monitor updates of thousands of regulatory documents, which involves visiting the websites of regulatory authorities and sifting through countless policy documents. This process is not only extremely resource-intensive but also error-prone and overall ineffective.
Machine learning can be used to automate a major part of the regulatory change management process. For example, Compliance.ai, a Silicon Valley startup founded in 2016, provides an ML-driven platform that helps banks keep up with regulatory changes. Using NLP and machine learning, the Compliance.ai platform automatically sources all relevant regulatory content from financial authorities, whitepapers, and news media. Importantly, when significant regulatory changes have taken place, the tool immediately alerts compliance officers.
Bank of Marin, a commercial bank that primarily operates in the Bay Area and has over $2 billion in assets, turned to compliance.ai to streamline its regulatory change management processes. Now Bank of Marin employees has access to all relevant regulatory content in one place, which significantly simplifies regulatory change management. As reported by compliance.ai, the Bank of Marin got returns on its investments in a short time by decreasing the number of resources for compliance.
Optimising stress testing
After the 2008 financial market crash, financial authorities substantially strengthened reporting requirements to ensure that banks can withstand significant economic downturns. This is why financial institutions need to routinely define and report their solvency to stay compliant, and banks with $50 billion or more in assets have to have their risk management teams conduct stress tests.
While undoubtedly important, these risk assessment processes often require hundreds of field experts and inordinate amounts of hours to complete. Moreover, finding relationships between various economic variables in relationship to banks’ performance is an increasingly complex and resource-intensive task.
Machine learning can help compliance experts to identify which exact combinations of values can cause risks, increasing reporting accuracy and decreasing the time it takes to complete these tests. This is exactly why Citi, one of the world's largest financial institutions that operate in 98 countries, joined forces with Symphony AyasdiAI to develop an ML-driven risk forecasting model.
Citi’s ML-driven stress testing solution
Prior to machine learning adoption, Citi had a hard time passing the annual Comprehensive Capital Analysis and Review (CCAR) conducted by the US Federal Reserve. CCAR requires a bank to submit its annual capital plans to the Fed to prove the bank's ability to deal with severe economic shock. Given the sheer number of economic variables at hand, Citi’s increasingly manual modelling approach took too long to complete, leaving business leads little time to understand the logic behind the models. As a result, Citi couldn't confidently defend its models in an annual submission to the Fed.
To create an accurate revenue forecast model, Ayasdi started with enriching the macroeconomic variables stipulated by the Federal Reserve. Then, Ayasdi used its proprietary machine-learning software to reveal how exactly these variables impact the revenues of each business unit. This allowed the company to define which unit-specific variables have the most impact on revenues. The detected variables were then used to create a custom ML-powered model that can accurately predict business units’ revenues under stressful economic conditions. As a result, Citi has managed to make this compliance process three times faster and less resource-consuming.
Enhancing fraud detection
Since the beginning of banking, fraud in its various shapes and forms has always been a persistent problem for many financial institutions. While many banks use sophisticated fraud detection systems, the rule-based nature of these solutions leads to a high probability of false positives. Importantly, fraudsters also keep innovating, which makes banks’ fraud detection systems grow obsolete.
Machine learning models, on the other hand, can learn and evolve together with the fraudsters. In very simple terms, a machine learning model can detect unfamiliar deviations from the normal pattern, notify the human employee about it, and, based on human feedback, learn if this kind of deviation is the new acceptable pattern or a case of fraud.
Just a few months ago, IBM launched a new generation mainframe that allows financial institutions to conduct fraud analysis of 100% of their transactions in real-time. To compare, around 10% of transactions could go through an AI-based fraud detection engine. In a nutshell, this means that banks that use the z16 mainframe can significantly reduce the number of false positives and increase customer satisfaction.
Risk management is a natural playground for machine learning. With the amount of data that banks have accumulated in the past decades, it’s only right to use the technology for its analysis. Machine learning models can drastically increase the accuracy of forecasts, decrease operational costs, and make risk management more effective overall.