Multi-(Horizon) Factor Investing with AI

Abstract

Can the backbone technology behind ChatGPT create and manage portfolios? We apply this tech-engine, adapted for finance applications, to multi-factor investing by a long-horizon investor who uses bigger that traditionally used data and takes into consideration long-term versus short-term volatility, liquidity and trading costs trade offs while maximizing expected portfolio returns. The answer is yes, as we are able to actively time factors' premium realizations while dynamically re-balancing and diversifying between factors. Moreover, the long horizon perspective is critical, as it allows for more patient trading and re-balancing needs, more strategic factor timing, and a different set of fundamental signals to rely on.