Enhancing relative value trading in the bond market
In this contribution we discuss relative value, specifically mean reversion, in the bond market.
First, we introduce the story behind this research and its importance to asset managers in the bond market. Then, we explain what is mean reversion and discuss its inherent steps namely: (1) identifying relevant pairs of bonds; (2) detecting anomalies; (3) when to open position; and (4) when to close position. Finally, we will present lessons learned from applying machine learning to tackle all those steps in one go.
Stéphane Collot has been working as a Data Scientist in ING for 4 years. He has two Master degrees from Mines de Saint-Etienne (France) specialized in computer science and from Georgia Tech (Atlanta USA) specialized in big data/machine learning.