<<< Previous speaker next speaker >>>

Jonathan Felch, Quantitative portfolio manager at E.H. Smith Jacobs

 Jonathan  Felch

Jonathan Felch has worked as a programmer, a project manager, an enterprise architect, a high-tech venture capitalist, a desk quant for quantitative trading strategies, a developer of algorithmic trading strategies, and a quantitative trader (and also a political operative and lobbyist a long, long time ago). But mostly he loves programming computers and pushing his technical skills into new areas.

Most recently he has been working on projects that bring together three themes: DSL in finance, distributed caching, small-code base polyglot software ecosystems. Specifically he has lead teams utilizing functional programming languages, meta-programming techniques and DSL for quantitative financial analysis. They have built distributed caching and shared memory solutions for the management of large amounts of data, including the short term storage of calculated and partially calculated time series analytics. He mostly works in C++, C#, Java, Scala, and Groovy.

Currently is a quantitative portfolio manager at E.H. Smith Jacobs and manager of their algorithmic trading and market data infrastructures. Previously he has held positions at Credit Suisse, Lehman Brothers, Paloma Partners, and a Goldman Sachs / Boston Consulting Group / General Atlantic Partners joint venture dedicated to incubating and carving out high tech assets in Global 2000 corporations.

Presentation: "Groovy on the Trading Desk"

Time: Thursday 15:00 - 16:00

Location: Olympic

Abstract: In 2007, a trend began emerging in the best quantitative trading desks: Groups were combining a high performance language --typically Java, C#, C++, or C -- with one or more scripting or functional programming language -- most notably F#, Python, Lua, or R Language. The proprietary trading desk at a major investment bank was responsible for a large existing JEE infrastructure focused on cross-asset class trading across credit markets, fixed income products, equity and volatility markets and reviewed and prototyped enhancements leveraging a similar evolution using Groovy, Scala, JRuby, R Language, and BeanShell. The team ultimately standardized on Groovy, which ended up being a far more strategic investment than was originally envisioned. This talk will focus on what worked and what didn't -- features of the language that yielded real benefit, features of the language that introduced problems, use-cases where Groovy was a good fit and use-cases where it wasn't, investments that scaled well and investments that didn't -- with an eye to both contributing to Groovy's success in the enterprise Java world and as a starting point for domain specific and new language development.