Guido Germano

Biography

Guido Germano is Professor of Computational Science at University College London, where he has been Director of the MSc Computational Finance since its foundation in 2015. He is also affiliated with the Systemic Risk Centre, London School of Economics. He studied at the University of Pisa obtaining a Laurea in theoretical physical chemistry (1994, thesis at MPIP Mainz) and a PhD in quantum-classical molecular simulation (1998), which included a stay in the second year with his co-supervisor at the University of California, San Francisco. After postdocs in theoretical soft condensed matter physics at Scuola Normale Superiore, Pisa (1998), the University of Bristol (1998-2000) and the University of Bielefeld (2000-2002), he held faculty positions at the University of Marburg until he joined UCL in 2013. He gradually shifted his research from large-scale molecular dynamics simulations on massively parallel computers to stochastic processes in statistical physics and financial mathematics, a field he became interested in during an assistantship at the University of Bonn (Statistics Section, Faculty of Law and Economics, 1994-1995). He was also a fellow at Università del Piemonte Orientale, Novara (Department of Economic Sciences and Quantitative Methods, 2009-2012). He briefly returned to Pisa as a fellow at Scuola Normale Superiore (Quantitative Finance Group, 2012-2013) and as Visiting Professor at Scuola Superiore Sant'Anna (Institute of Economics, 2026).

Workshop contributions

Applications of machine learning in finance

The talk covers recent applications of machine learning in finance I have been working on in the last ten years with my PhD students: 1. The use of large language models to assess sentiment in financial news, correlate it with returns, and deploy it in sentiment-based trading strategies; 2. The use of reinforcement learning and sentiment-augmented reinforcement learning in portfolio allocation. 3. The use of deep neural networks as surrogates to speed up the pricing with stochastic models; 4. The use of deep neural networks for model calibration, i.e. the forward-looking estimation of model parameters from the market prices of European options; 5. The use of various machine learning techniques (logistic regression, support vector machines, neural networks, Bayesian regularisation, k-nearest neighbours, etc.) for credit scoring. Not all these approaches work equally well or have an edge with respect to traditional methods just because they are based on machine learning: the first has the most spectacular results, the fifth the least, the third and fourth depend on the model.