Assistant Professor Halis Sak's research on explainable machine learning in finance

Over the years, top journals have published hundreds of characteristics to explain why some assets or portfolios provide higher expected returns. And recently there are papers showing that machine learning models perform much better than leading regression-based strategies from empirical asset pricing literature. The research of Assistant Professor Halis Sak of our school aims to explain how machine learning models achieve better performance. In his first working paper (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=3202277), he and his co-authors attribute this to machine learning portfolio’s alternating exposure between investor arbitrage constraint and firm financial constraint characteristics, the timing of which aligns with credit contraction and expansion states.

Before joining to the School, Assistant Professor Halis Sak worked as a Visiting Assistant Professor of Finance at the department of finance of HKUST where he taught FinTech courses for MSc and PhD students for two years. His academic work appeared in top rated journals like Review of Finance (FT50).