Learning non-linear relationships in the cross section: a real-world application of Machine Learning to factor investing with Credit Suisse's HOLT dataset - 3 de febrero de 2021 - TecnoWebinars.com

Institutional investors and active portfolio managers have been using factors to explain and forecast market returns. Empirical evidence has shown how these could help increase diversification, achieve excess returns and manage risk properly. Among these factors we typically include Style factors such as Quality, Size, Value, Momentum, Low Volatility, Dividend Yield, CFROI, …

Highlights

Joined by Credit Suisse’s equity investment strategist Ricardo Pachon Cortes, during this webinar we will effectively apply Machine Learning models to a broad range of factors from the HOLT dataset. We investigate the relationship between these, and market returns and include them within a typical investment process. During the webinar, the key sessions will focus on:

           - Implement Machine Learning models to automatically select relevant factors and to explain the relationship between factors and market returns.

           - Optimize the hyperparameters of the models for performance improvement.

           - Develop Machine Learning-based investment strategies for factor allocation.

           - Interpret the behavior of Machine Learning models.

Who Should Attend

Quants and investment professionals with an interest in designing and implementing automated processes that leverage Machine Learning models to select the "most performing" investment factors.

Agenda

15.00:Equity Factor Investing

           - Introduction to Equity factors as a source of alpha

           - Evidence in the literature of market timing

            - HOLT factors and data

15.20 :Factor prediction using Machine Learning

           - ML single-stock prediction models

           - Learning non-linear patterns interactively with MATLAB Apps

           - Model's performance improvement with automated hyperparameters tuning

           - Aggregation of predictions

16.00: Strategy construction and interpretability

          - ML-powered strategies for factor allocation

           - Interpretability of ML models: Global and local behaviour

16.40 : Q&A