High-speed detection of emergent market clustering via an unsupervised parallel genetic algorithm

We implement a master-slave parallel genetic algorithm (PGA) with a bespoke log-likelihood fitness function to identify emergent clusters within price evolutions. We use graphics processing units (GPUs) to implement a PGA and visualise the results using disjoint minimal spanning trees (MSTs). We demonstrate that our GPU PGA, implemented on a commercially available general purpose GPU, is able to recover stock clusters in sub-second speed, significantly faster than a test case implementation of a comparable serial genetic algorithm, based on a subset of stocks in the South African market. This, combined with fast on-line intraday correlation matrix estimation from high frequency data for cluster identification, offers cost-effective, near-real-time risk assessment for financial practitioners.