Data Science & Decision Optimization in Banking & Financial Services
How to offer the right product to the right customer at the right price through the right channel with the right incentives at the right time? How to find the weights of a set of elements in a given set to maximize a given financial objective subject to the constraints of each element? How to compete on price and product in a game environment? Or how to manage risk and capital to navigate complex regulatory environments? These decision optimization problems are at the heart of retail banking and finance.
This newly offered, yet often overlooked, course will pave your way with the art and science of decision optimization to succeed in consumer banking, commercial banking, corporate banking, and wealth management.
Quantitative decision making in banking and finance involves a blend of multidisciplinary theories, method and techniques, from financial accounting to economic theories, from traditional statistical and econometric methods to machine learning and artificial intelligence techniques, from estimation optimization algorithms to decision optimization procedures, and from quantitative approaches to business acumen. This course equips students with the ability to integrate and apply these diverse methodologies to formulate decision problems, develop robust quantitative models, and ultimately optimize decisions to tackle real-world financial challenges.
While decision optimization touches almost all areas of banking and finance, this course currently focuses on business structure, asset portfolio, capital allocation, marketing campaign, price competition, risk management, operational processes, and treasury liquidity. Products covered include deposits, banking, credit, loans, mortgages, and assets in consumer, commercial and corporate banking, as well as wealth management.
In addition to rigorous academic training, students gain practical experience through real-world case studies and class projects with the support of financial industry professionals.
This course is divided into the following five Parts.
Part One: Introduction
- This part serves as a foundation for understanding decision making in the banking and financial services industry. It first discusses the systems of equations implied in bank financial statements, P&L equations, and the risk dynamics of different product lines. It then reviews key concepts of decision science concepts, including expected utility, risk-neutral equivalence, stochastic dominance, risk-return tradeoffs, Bayesian decision theory, decision making in game environments, and Bellman dynamics
Part Two: Modeling and Optimization Techniques
- This course provides a focused review of the methods as needed. Machine learning techniques include ensemble learning, neural networks, reinforcement learning, sentiment analysis, and natural language processing. Decision optimization techniques include linear programming, nonlinear programming, and dynamic programming, stochastic optimization, robust optimization, and scenario-based optimization such as Minimax and Maximin.
Part Three: Business Strategy Optimization
- This part covers hoe to select a strategy or combination of strategies from a viable set to maximize a given financial objective subject to a set of conditions. A typical example is offering the right product to the right customer at the right price through the right channel with the right incentive at the right time. Key areas covered include marketing campaigns, risk and capital management, interest rate and liquidity management, and bank-wide strategy.
Part Four: Asset and Business Portfolio Optimization
- Starting from the mean-variance framework, this part explores how to find weights or quantities of elements in a given set to maximize a given financial objective subject to a set of conditions. Key areas covered include capital structure and allocation optimization, asset portfolio optimization, and business portfolio optimization.
Part Five: Other Types of Decision Making
- Price competition in a game environment with Nash equilibrium approach, dynamic optimization using methods such as Bellman equation.
Class Projects
- To help students master risk management concepts, methods, and techniques, students are expected to work in groups on well-designed, real-world financial risk management cases.
- Hands-on class projects will account for the major part of a student’s final grade.
Instructor
- Hengzhong Liu, Executive in Residence, former managing director and SVP at Bank of America and Citigroup. His research and publications have focused on financial markets, industrial competition, and economic reform and development in China.
Prerequisites
- College-level calculus, Linear Algebra (e.g., MATH 216, 218 0r 211), probability and statistics (e.g., MATH/STA 230, MATH 340/STA 231 or Math 238L/EGR 238L).
- Basic programming skills in Python, R or SAS
- While prior knowledge in finance is not required, some basic understanding of economics, finance, and financial institution is preferred. Intellectual curiosity and independent study habits are strongly desired.