FUTURE OF RISK MODELING

It is said that we are entering the fifth industrial revolution: the age of artificial intelligence. The ability of computers to start performing human tasks (called artificial intelligence, AI) and the wider use of complex algorithms that detect nonlinear relationships and self-learn (called machine learning) are starting to mature from experimentation to production and, in turn, revolutionizing many aspects of the financial services industry.

The uptake of these technologies for process automation and in digital customer journeys is growing exponentially in many industries, yet we are observing a more conservative and slower uptake in financial risk management. In an era where so much information is available on the use of AI and machine learning, financial organizations are cautious about its pertinence in regulated areas that expect compliance and transparency in decision-making.

At the same time, the digital revolution is occurring against a backdrop of an increasingly uncertain world. Volatility is at an all-time high. The risk management function is contending with new types of risks every day. Organizations around the world are dealing with myriad risks such as a haphazard recovery from the COVID-19 pandemic, rising inflation, cumulating geopolitical risks, and the impacts of climate change.

With this book, we want to highlight the strengths and weaknesses of AI and machine learning and explain how both can be effectively applied to everyday risk management problems, as well as efficiently evaluating the impacts of shocks under uncertainty, such as global pandemics and changes in the climate. Throughout the text, we aim to clarify misconceptions about the use of AI and machine learning using clear explanations, while offering practical advice for implementing the technologies into an organization’s risk management framework.

With the right controls, AI and machine learning can deliver tangible benefits and become useful tools in the toolkit of the risk function. It can improve the accuracy and speed of risk assessments compared to human-led or other traditional methods of decision-making, and at the same time introduce new ways of work in risk management through increased automation. The rewards for innovation are not without risks of their own, and these technologies are largely underregulated today (although, that will change in the future). In this book, we also highlight the barriers that organizations face in using AI and machine learning and provide ways to overcome them.

The book is structured to introduce AI and machine learning in the context of financial risk modeling, including the onboarding and preparation of diverse datasets. Throughout the book, we provide real-world risk management applications. The book contains dedicated material on model implementation, explainability, and addressing bias and fairness. It also provides details on extending model governance frameworks to AI and machine learning, the use of optimization in machine learning, and how AI and machine learning can help risk managers better assess and address new types of risks like climate change.

With the transformational advances in AI and machine learning, together with the radical speed of new development, as an industry, we are only scratching the surface in its practical application in financial risk management. With this book we aim to enable organizations to continue putting in place the right frameworks and infrastructure to enable modern technologies, and more importantly, build proficiency and capacity in AI and machine learning.