Erik Rigtorp


Blog

08 Jun 2013
Estimating order queue position
22 Nov 2012
Designing a high performance market data feed handler
01 Jan 2011
Efficient rolling statistics with NumPy
20 Nov 2009
Erlang Latency Guide

Education

Stanford University 2007 - 2009

Electrical Engineering

The Royal Institute of Technology 2004 - 2009

Master of Science in Electrical Engineering

Awards & Scholarships

  • The Royal Institute of Technology’s General Scholarship, 2007-2009, Awarded for academic achievements.
  • John and Karin Engblom’s Scholarship, 2007, Awarded for studies at Stanford University.
  • Henrik Göransson’s Sandviken Scholarship, 2007, Awarded for academic achievements.

Projects

Spartan
High-Frequency trading components
NanoMQ
Ultra low latency messaging kernel
ipc-bench
Latency benchmarks of Unix IPC mechanisms

Experience

CTO 2013-present

Pan Capital LLC, Fort Lauderdale, Florida, USA

  • Outsourced infrastructure moving to a managed solution reducing expenses by 40\% and significantly reducing staff requirements.
  • Re-architectured trading system reducing the number of colocation sites from 6 to 2 while retaining capabilities and performance.

Quantitative Trader 2011-2013

Buttonwood Group Trading LLC, San Francisco, California, USA

  • Created a statistical prediction model based on order book and cross asset information that generated 30% of FX arbitrage PnL.
  • Traded and researched ETF market making and relative value strategies.
  • Developed a low latency feed handler for CME FIX/FAST in 2 weeks saving $75000 in annual licensing costs. Implemented as a code generator in Python that generated a C++ decoder from a FAST template, achieved lower latency than commercially available solutions.
  • Devised methods for order queue position estimation, order stacking and routing.
  • Developed NASDAQ, BATS and Arca feed handlers and exchange connectors in C++. Attained microsecond latency by using cache efficient algorithms and avoiding CPU cache pollution.