Big Data and Machine Learning in Investing

Speaker

Garthee Ganeshapillai, Stefano Pasquali, Alex Remorov
BlackRock (CSAIL Alliances Member Company)
Abstract: We outline how we use Big Data and Machine Learning in our everyday investing process at BlackRock. On the Data side, we strive to incorporate a vast amount of granular information – while making the availability of this information fast and scalable for downstream models. We will give a few examples of what challenges we encounter in our data process and how we address them. On the Machine Learning side, we will discuss what makes the financial domain so special when applying ML, and which situations are more favorable for applying machine learning models. We will also present a few live applications of Machine Learning in investing, including liquidity and trading modelling, managing risk, as well as extracting alpha insights.



Participants:

Ganeshapillai Gartheeban, PhD, Director, is a member of the Global Equity Research team within BlackRock's Systematic Active Equity group where he focuses on extracting patterns from large scale heterogeneous datasets.

Prior to joining the firm in October 2014, he was doing in his PhD at Massachusetts Institute of Technology, where he worked on developing machine learning algorithms for various problems in medicine, systemic risk in financial systems, and sports. Primary motivation of his research is to discover novel approaches to automatically recognize patterns in large datasets and develop tools to answer questions that affect people. His PhD thesis was on learning cross-sectional connections in financial time series, and was supervised by Professor John Guttag and Professor Andrew Lo.



Stefano Pasquali, Managing Director, is the Head of Liquidity Research Group at BlackRock Solutions. As Head of Liquidity Research, Mr. Pasquali is responsible for market liquidity modelling both at the security and portfolio level, as well as estimating portfolio liquidity risk profiles. His responsibilities include defining cross asset class models, leveraging available trade data and developing innovative machine learning based approaches to better estimate market liquidity. Mr. Pasquali is heavily involved in developing methodologies to estimate funding liquidity and better estimate funds flows. These models include: the cost of position or portfolio liquidation, time to liquidation, redemption estimation, and investor behavior modelling utilizing a big data approach.

Previous to Blackrock, Mr. Pasquali oversaw research and product development for Bloomberg's liquidity solution, introducing a big data approach to their financial analytics. His team designed and implemented models to estimate liquidity and risk across different asset classes with a particular focus on OTC markets. Before this he led business development and research for fixed income evaluated pricing.

Mr. Pasquali has more than 15 years of experience examining and implementing innovative approaches to calculating risk and market impact. He regularly speaks at industry events about the complexity and challenges of liquidity evaluation? particularly in the OTC marketplace. His approach to risk and liquidity evaluation is strongly influenced by over 20 years of experience working with big data, data mining, machine learning and data base management.

Prior to moving to New York in 2010, Mr. Pasquali held senior positions at several European banks and asset management firms where he oversaw risk management, portfolio risk analysis, model development and risk management committees. These accomplishments include the construction of a risk management process for a global asset management firm with over 100 Billion AUM. This involved driving projects from data acquisition and normalization to model development and portfolio management support.

Mr. Pasquali, a strong believer in academic contribution to the industry, has engaged in various conversations and collaborations with universities from the US, UK, and Italy. He also participates as a supervisor in the Experiential Learning Program and Master of Quantitative Finance Program based at Rutgers University, along with tutoring students in research activities.

Before his career in finance, Mr. Pasquali was a researcher in Theoretical and Computational Physics (in particular Monte Carlo Simulation, Solid State physics, Environment Science, Acoustic Optimization). Originally from Carrara (Tuscany, Italy), he grew up in Parma. Mr. Pasquali is a graduate of Parma University and holds a master’s degree in Theoretical Physics, as well as research fellowships in Computational Physics at Parma University and Reading University (UK).

Stefano lives in New York since 2010. In his spare time, he tries to devote to his passions which are music, traveling and spending more time as possible to the sea and sailing boat.

Alex Remorov, PhD, is a Vice President at BlackRock's Systematic Active Equities (SAE). In this role, Alex builds systematic alpha strategies for hedge funds and long-only portfolios by leveraging machine learning, alternative data, and investment intuition.

Alex earned a BSc in Mathematics and Statistics from the University of Toronto and a PhD in Operations Research from MIT. He has carried out academic research on systematic trading strategies, behavioural biases, and investor decision-making. During this time Alex also did short stints at Manulife Investment Management in Toronto, as well as at Goldman Sachs in New York and London.