B1075 - FINANCIAL INTELLIGENCE: ADVANCED TOPICS IN FINANCIAL KNOWLEDGE GRAPHS

Conoscenze e abilità da conseguire

The aim is to implement the knowledge achieved in the advanced quantitative methods by working on a business problem that is proposed in cooperation with a local industry partner. The course includes an off- and onsite introduction to the theoretical background needed to work out the proposed case study, and a phase of team work.

Contenuti

This course is reserved to the QF master students who are declared winners of the dedicated call for applications.

The increasing complexity of financial markets has turned technology into the key to extracting information hidden in the huge volume of noisy background data. The construction of reliable and robust investment portfolios is the main goal of fund managers in banks and asset management companies, both for investment funds and private clients. Such portfolios need to incorporate company market views into forward-looking market scenarios, ensure the right degree of diversification avoiding excessive concentration and respond smoothly to investing environment shifts. Artificial Intelligence (AI), specifically Machine and Deep Learning methods applied to asset allocation, allows to change completely the classic ‘portfolio theory’ mindset, that is still prevailing in most of the industry worldwide and has got insurmountable and serious shortcomings that have been affecting the effectiveness of the asset portfolio construction process since its foundation.

In this intensive program we will build the full asset portfolio construction process in an AI modeling environment, with precise effort on defining the dynamic probabilistic structure of ‘market regimes’ and the definitions of input data for portfolios that adapt to future scenarios in an optimal way. In this journey we will revise and employ many of the AI most powerful tools, mainly ensemble random forests, neural networks and hidden Markov models in a multivariate setting. The participants will work on this project in a professional way, with the aim of building a fully operational process in the reality of the financial industry, blending theory with practice on a top financial subject of today’s markets.

Intensive Program Schedule:

Dott. Maurizio Morini - Aletti Bank

• Introduction to the course and problem assignment

Prof. Umberto Cherubini

• General Portfolio Management Issues
• Asset Prices and Risk Factor Dynamics

Prof. Giovanni Della Lunga

• Introduction to ML techniques
• General Issues in Data Analysis
  

Prof. Pietro Rossi

• ML Techniques Applied to Asset Management
• Computational Issues in Machine Learning

Dott. Maurizio Morini

• A Portfolio Problem: Asset Classes and Risk Factors
• Machine Learning Techniques for Market Regimes Identification
• Identification of Portfolios for Target Investors

Dott. Maurizio Morini

• Discussion and evaluation of student teams work.
  

Testi/Bibliografia

Marcos M. Lòpez de Prado, Machine Learning for Asset Managers, Cambridge University Press, 2020

Marcos M. Lòpez de Prado, Advances in Financial Machine Learning, John Wiley, 2018

Metodi didattici

This Intensive Programme is a unique opportunity to take part in an international event, to deepen your knowledge on current challenges for Artificial Intelligence applications in financial markets, to work in contact with professionals from important financial companies and to experiment professional group work in a multinational team.

The course is organized in the form of a workshop including both lectures and tutorials offered by local teachers and distinguished guest experts. The classes will provide students both a theoretical and methodological background to understand the application of the main AI techniques to the problem of dynamic portfolio management both for proprietary and client portfolios. This intensive program also aims to provide students critical skills to be able to understand concepts and tools to analyze AI applications also from the point of view of risk management.

Intercultural learning will be an integral part of the onsite local course experience. In addition to the diversity of the case group, consisting of students that will represent each of the four academic institutions.

Modalità di verifica e valutazione dell'apprendimento

Students will be evaluated based on an individual test, concerning the topics covered in the lectures, and the team work on the assignment proposed at the beginning of the program. 

Both the individual exam and the the team work will be evaluated on a score ranging up to 30. The final score will consist of the weighted mean of the individual scores (with weight 30%) and the teamwork score (70%).

Attendance is mandatory, at least 75% of all classes is necessary and will be verified.

The exam will be evaluated on a score ranging up to 30.

Strumenti a supporto della didattica

Slides for the lectures and computer labs.

Orario di ricevimento

Consulta il sito web di Umberto Cherubini