GCASR 2016‎ > ‎Presentations‎ > ‎

Technical and Quantitative Challenges in US Equities Market Making

Technical and Quantitative Challenges in US Equities Market Making

Carson CookColin Clarke and Rob Lopez (KCG Holdings, Inc.)



Abstract:
To efficiently make markets in US stocks requires solving challenging problems in the quantitative, data, and computing spheres. We employ various machine learning techniques to model the fair value of stocks, and train these models using a multi petabyte store of historical market data. We do this to build and simulate electronic trading models that provide liquidity in many thousands of equities which trade on tens of exchanges in several physical locations. This liquidity helps to communicate to the market fair asset prices and lowers the cost of trading for people who need liquidity. In this talk we highlight a few interesting problems that we confront in our industry, including statistical modeling, managing large datasets, and distributed computation.

Bio:
Carson Cook has been a developer at KCG since 2011.
Colin Clarke is currently a trader at KCG.
Rob Lopez has been a trader at KCG since 2010. Before that, Rob was VP, Head of Quantitative Strategies at TD Securities.
Comments