TSAA-SF May Monthly Virtual Gathering
May 5, 2026 (Tue) at 5:00 PM PT
Zoom
TITLE:
Machine Learning, Regression, and Trading Strategies:
Do Complex AI-based Models Always Win?
Speaker:
Davide Pandini holds a master’s degree in Electronics Engineering (Summa cum Laude) from the University of Bologna (Italy), a PhD in Electronics and Telecommunications from the University of Padova (Italy), and a PhD in Electrical and Computer Engineering from Carnegie Mellon University, Pittsburgh (U.S.). He joined STMicroelectronics in Agrate Brianza (Italy) in 1995, where he is a Technical Director and Fellow of STMicroelectronics Technical Staff. Since June 2015, he has been the chairman of the ST Italy Technical Staff.
Dr. Pandini holds the CMT, MFTA and CFTe designations, and is a professional member (CSTA) of SIAT (Società Italiana di Analisi Tecnica), a member society of IFTA. In 2021, he was the recipient of the XII SIAT Technical Analyst Award in the Open category, and the winner of the prestigious IFTA 2021 John Brooks Memorial Award.
He was a speaker at IFTA 2023 and 2024 annual conferences, at Bogu Investment Forum 2024 in China, at the Open Source Quantitative Finance 2025 conference in Chicago (U.S.), and at several SIAT events. His work has been published in the Journal of Technical Analysis (CMT Association) and in the IFTA Journal. Since October 2024 he has been a Member of the SIAT Scientific Committee, and since October 2025 of the IFTA Board of Directors.
Dr. Pandini served as volunteer at the Universal Exhibition Expo2015—Feeding the Planet, Energy for Life—in Milano (Italy).
DESCRIPTION OF THE TALK:
Artificial intelligence and machine learning are often assumed to outperform traditional statistical models in market forecasting. But do more complex models actually produce better trading results?
This presentation examines that question through a data-driven comparison of multiple linear regression, XGBoost, support vector machines, artificial neural networks, and deep neural networks applied to the S&P 500. Using a standardized trading framework, the study evaluates not only forecast accuracy, but also real-world trading performance through return, risk-adjusted return, drawdown, and the impact of transaction costs.
By comparing model-driven strategies with a traditional buy-and-hold benchmark, the talk highlights an important distinction between models that look strong in prediction and models that remain effective in practice. The results show that while some AI-based methods can improve forecasting metrics, greater complexity does not automatically translate into better trading outcomes. In several cases, simpler and more interpretable statistical approaches remain highly competitive, especially once transaction costs and practical trading constraints are taken into account.
For technical analysts, traders, and investment professionals, this presentation offers a practical framework for judging when advanced AI tools add value and when simpler statistical models may be the more robust choice.
Ultimately, we seek to answer a fundamental question: Does complexity always lead to better trading and investment decisions, or can traditional statistical models still hold their ground in financial forecasting?
Please note: these monthly gatherings will now be members-only. If you're not already a dues paying members, please join us!
Please join us for our next TSAA-SF monthly gathering of our membership. At these events we celebrate our Technical Analysis Heritage. Each month the members meet and consider the many aspects of T.A. or visual financial data analysis. The programs will profile the charting methods and techniques of our members and guest presenters.
Mike Jones, Nick Chong, and Bruce Fraser are our co-hosts at these meetings.
Zoom link will be sent out to registered attendees. This event is open to dues paying members only.
Best regards,
TSAA-SF
