The banking industry is undergoing an enormous shift driven by advancements in data analytics and artificial intelligence (AI). These technological innovations are fundamentally changing how financial institutions manage risk, enhance operational efficiency, and ensure compliance with regulatory standards. The integration of AI and machine learning models is revolutionizing risk management by providing deeper insights and predictive capabilities that help mitigate potential threats and capitalize on opportunities.
Fabian Serrato, CRO Retail and SMB in Citibanamex and previous Senior Vice President of Model & Data Risk Management in Scotiabank is leading the charge in this evolving field. Fabian’s responsibilities during his carrer have included ensuring regulatory compliance across various risk models, enhancing capital efficiency, and nurturing strategic partnerships to align with enterprise priorities. His passion for AI stems from a belief in its transformative potential for banking and risk management, driving his efforts to integrate advanced technologies into the bank’s risk management practices.
In the entities where Fabian has worked, these efforts are central to their strategy of becoming data-driven organizations. These institutions have applied Fabian’s expertise to modernize their end-to-end modeling environments and enhance data-driven decision-making processes. By embracing continuous learning and cultivating innovation, they aim to stay ahead of regulatory changes and industry trends, ensuring robust risk management and sustained growth in an increasingly complex financial sector. Let’s explore Fabian’s journey of transformative leadership in the banking industry:
Advocacy for Portfolio Management and AI Risk Awareness and Defense
Fabián provides strategic direction, leadership, and oversight for the management of product risk across Consumer Bank portfolios, encompassing both secured and unsecured products. His role involves developing and managing credit and collection risk strategies to ensure alignment with the organization’s risk appetite and compliance with regulatory standards.
He is responsible for implementing predictive models for risk and loss analysis, which enhances the organization’s ability to forecast and manage potential threats. Fabián leverages innovative approaches to advance risk management practices and ensures that all activities comply with established control and governance frameworks. His efforts focus on optimizing risk controls, improving capital efficiencies, and fostering strategic partnerships to support the organization’s overall objectives.
In previous roles Fabian provided strategic direction, leadership, and oversight for the management of Global Risk Models, Data & AI Risk. His role ensures that business strategies, plans, and priorities meet business expectations while remaining in compliance with governing regulations, internal policies, and procedures.
He builds strategic partnerships with business stakeholders to accelerate the deployment of innovative risk model solutions and to provide quantifiable business impact. He also promotes Data and AI risk awarenessé
Fabian is very passionate about AI because he strongly believes that it will continue to impact the way they work and how people do banking. He is a firm believer that it will revolutionize risk management practices.
Growth Through Diverse Roles in Risk Management
Fabián studied Economics and holds a master’s degree in Econometrics. His experience and academic background fostered a passion for risk models and analytics. Early in his banking career, he realized the importance of these disciplines in credit decisions and the quality of portfolios, and he has focused on growing in these areas throughout his various roles.
He led the risk teams in Colombia, Peru, and Mexico, experiences that greatly enriched his career as he navigated a diversity of economic cycles. He then moved to Toronto to lead the Unsecured Risk Portfolio in Canada. Prior to his role at Citibanamex, he served as the SVP for Retail and Small Business Risk for International in Scotiabank, a role that pushed his professional boundaries as he led the international credit division through the pandemic and post-pandemic stages.
More specifically to his current domain, the Gen AI use cases underway are a great current example because applications to improve processes are numerous throughout the banks he worked, but there are synergies among them. Many groups want to convert the coding languages of their programs; many want to document their code or processes; many want to summarize large documents, etc.
By creating a centralized program to work on these use cases, various groups are brought together to work on the solutions. Some of these groups may not have much interaction otherwise. It’s been a great opportunity to break down silos and share best practices to address challenges.
Harnessing the Power of Data and Analytics
The growth in data assets and the knowledge provided by analytics has always been on the rise and will continue to do so. There will always be more potential in data and analytics. More data sources, better ways of storing it, better ways of combining it, better tools to analyze it, and better methods of sharing insights.
Fabián’s approach is one of challenging his teams to determine how they can best leverage the latest tools to accelerate their processes, especially repetitive ones, so that they may spend more time deriving business and analytical insights rather than solving technical challenges. Through this process of continuous improvement, they enable data-driven decision-making in risk management.
The growth in data assets and the knowledge provided by analytics has always been on the rise and will continue to do so. There will always be more potential in data and analytics. More data sources, better ways of storing it, better ways of combining it, better tools to analyze it, and better methods of sharing insights.
Cultivating a Holistic View through Expertise Fusion
Collaboration and innovation are not just important within Fabián’s team but with all their partners. He sees a partnership as more than a collaboration; it’s a fusion of the best minds dedicated to creating innovative tools to support the financial industry’s vision. Together, they combine their expertise and tools to provide a holistic view with better insights and indicators.
Fabián is a big promoter of innovation and his team regularly structures internal innovation competitions. They heavily reward their talent for coming up with “out of the box” solutions. These efforts are a great way to build their pipeline and focus on what they need to do next.
Data-Driven Decision-Making in Banking Institutions
Financial institutions where Fabian has worked extensively leverage analytics in their risk management strategies to drive the bank as a data-driven decision-making institution, especially in risk management. The examples are numerous.
Their models and analytics support efforts to maximize the Risk-Adjusted Return (RAR) under the constraints of losses or capital allocation. For instance, with their current emphasis on client primacy, they are integrating their tools to better understand the profitability of priority segments, including Expected Revenues and Losses, among others.
They also run stress testing models to simulate impacts on the balance sheet of the bank and determine how business lines react to various shocks and crisis scenarios. As a result, based on the probability of those scenarios, the bank can anticipate their likelihood, design strategies to mitigate losses, and create a plan to execute those strategies.
For a Collections Strategy, they identified a high probability of defaulting on customers. Analytics enables them to segment those customers and determine treatments for the collections team (Loss Mitigation Tools, scripts, assignments to specific groups of collectors, etc.). As a result, collection processes were automated, and the implementation of strategies was improved by being able to better discriminate high-risk clusters.
Modernizing End-to-End Modeling Environments
In his view, the expansion of data assets and the insights offered by analytics are on a continuous upward trajectory. There is an increase in data sources, more efficient methods of storage, improved access and integration capabilities, enhanced analytical tools, and more user-friendly ways to generate and share insights.
In the same vein, model development is also evolving; there are better development tools, newer and simpler coding methods, new types of models, and improvements in model platforms as back-end systems become more receptive to machine learning and AI models.
In view of all these developments, Fabian and his team are in the process of modernizing their end-to-end modeling environment. This modernization spans from the data marts to their tools and platforms and all the way to deployment.
The shift to the cloud is a prime example of this modernization. It will equip them with a more detailed understanding of their client base. The ability to consider the full value of a client by combining insights on all the services they use and both unstructured and structured data presents a significant opportunity.
AI is poised to play a crucial role in revolutionizing risk management practices. Generative AI will transform the way they code, interpret, and generate insights. It will automate existing processes and assist in day-to-day operations, completing many tasks almost instantly.
Embracing Continuous Learning
Fabián’s professional development encompasses studying, attending conferences, and learning from his team. He maintains his knowledge of economics by reading various publications and staying up to date with economic news. He attends several conferences, many of which have been AI-focused recently but also include conferences on data and model risk management.
Another significant source of information for him is his team, which boasts a great diversity of expertise and interests. This diversity is supported by robust platforms and forums that facilitate the sharing of the latest developments with each other.
Staying Ahead of the Curve
He believes one should not hesitate to employ new techniques. It is beneficial to have the courage to be an early adopter rather than adopting a too-conservative approach. As a beta user, one can learn through practical applications. To mitigate risks, it is advisable to create sandboxes where one can experiment and even fail. It is important to leverage the current global AI boom.
While these machine learning techniques for AI are not new, we now possess the computational power to operate them. Being ahead of the curve and accelerating analytics can lead to the transformation of one’s organization, and leaders are likely to be supportive of such initiatives.
The Evolution of Models in Decision-Making
Modeling and data risk are two areas currently undergoing significant changes. With the increasing importance of models in decision-making, the regulator in Canada, OSFI (Office of the Superintendent of Financial Institutions), has updated their guidelines (E23), which are having a substantial impact on the model landscape in the banking industry.
The alterations in the definition of a model alone have a drastic impact. The scope has been expanded to now include more model types (including machine learning, qualitative models, and expert judgment models) and more business areas (including wealth, insurance, and pricing), leading to a potential 50% growth in model inventories.
These adjustments to regulatory guidelines significantly increase the governance required around data standards. The new guidelines focus on the end-to-end model lifecycle, which begins with data lineage and fit-for-use.
With these changes, roles and responsibilities have been greatly expanded as well. This applies not only to new model owners and users but also to many teams already part of the regulatory compliance process for models. Several development teams have more steps to incorporate into their processes; the validation teams have more models to review, and the governance team is establishing processes to guide all the teams to ensure compliance.
The Future of Banking
In addition to AI transforming risk management and the way operations are conducted, there is an expectation to see more automation and open banking in the future. Tools are becoming more advanced and platforms are integrating these tools. Whether it’s data acquisition, model development (including documentation), validation, implementation, or governance, the processes are becoming streamlined.
Open banking has been a topic of discussion for years, but the federal government recently allocated some funds to begin preparing for oversight of a new framework and system. With the changing nature of banking that this will bring about and the increased competition, risk management will play a significant role not just in protecting individual financial institutions but also the system.
