The JOIM Conference Series showcases high quality presentations and a platform for interactive discussions of current topics in the investment management arena. Prevalent throughout both activities is the highest quality material suitable for academics and practitioners.
October 5 – 6, 2020
Scheller College of Business, Georgia Tech
Applications of Data Science in Investment Management
The main focus of the JOIM conference will be on AI in investment management. Industry is rapidly incorporating machine learning methods into investment management as a complement to standard econometric approaches. This conference will feature papers that are early exemplars of this methodological shift, covering supervised learning, unsupervised learning, and reinforcement learning.
This event will be held at the Scheller College of Business, Georgia Tech campus. We will practice social distancing and follow all recommended local safety guidelines.
Sudheer Chava, Georgia Tech
Tucker Balch, JP Morgan
Sanjiv Das, Santa Clara University
Dynamic Goals-Based Wealth Management Using Reinforcement Learning
Time-Series Variation in Factor Premia: the Influence of the Business Cycle
heterogeneity can be exploited to motivate dynamic rotation strategies among established factors: size, value, quality, low volatility and momentum. A timely and realistic identification of business cycle regimes, using leading economic indicators and global risk appetite, can be used to construct long-only factor rotation strategies with information ratios nearly 70% higher than static multifactor strategies. Results are statistically and economically significant across regions and market segments, also after accounting for transaction costs, capacity and turnover.
Vasant Dhar, New York University
The Stability of Machine Learning Models in Finance
I report results from a controlled study that measures model variance as a function of (1) the inherent predictability of a problem and (2) the frequency of the occurrence of the class of interest. The results provide important guidelines for what we should expect from machine learning methods for the range of problems that vary across different levels of predictability and base rates.
The results are of general scientific interest, and explain when and why we trust machine learning models in various domains including vision, language, finance, and healthcare.
Marty Flannagan, Invesco
Bryan Kelly, Yale University and AQR Capital Management
Text Data in Asset Management
Ananth Madhavan, BlackRock
Model Follower Identification Using Modern Machine Learning
Deep Srivastav and Jennifer Ball, Franklin Templeton
Data Science in Retirement
The event is co-sponsored with the Scheller College of Business, Georgia Tech, Invesco, BlackRock and Franklin Templeton.
Risk Managers, Portfolio Managers, Pension Managers, Senior Executives of Financial Firms such as Plan Sponsors and Endowments and Academics would all benefit from attending these conferences.