Systematic Alpha Management LLC (SAM) is a New York-based firm specializing in trading systematic, short-term quantitative strategies, using fully-automated, around-the-clock electronic execution in a wide range of futures markets and proprietary spreads. SAM’s investment programs aim to generate consistent positive returns with zero-to-negative correlation to any major equity, bond, currency and broad hedge fund or CTA index.

Research.

Every systematic trading company doing research follows a certain unifying set of principles on which the research is based. Typically, these principles are related to the background of the founders and key personnel. Systematic Alpha Management is not an exception in this regard.

The background of one of its principles and key researchers is in theoretical physics and applied mathematics.  Alexei Chekhlov, has spent many years working in both fields,  theoretical physics and applied mathematics.

He graduated from Moscow Institute of Physics and Technology, a Russian version of MIT. His academic experience includes post-graduate work at Landau Institute for Theoretical Physics and his work at Princeton University, where he was focused on developing theoretical models of fluid turbulence. For the last 8 years Alexei has been also teaching mathematics of finance at the Mathematics Department of Columbia University in New York.

Our approach to studying the financial data and proposing and testing the statistical hypothesis is rooted in desire to have hypothesis which have simple “physical meaning”, with minimal number of parameters, and reliable statistical in-sample and particularly, out-of- sample testing. The robustness of the statistical results is absolutely key in the ever-changing financial and political environment.

Over the years of financial data research, some of the tools Systematic Alpha Management has developed and uses are:

  • A proprietary backtesting environment which on a given financial data (given data resolution and type such as trade, BBO, equidistant bars) can test a rather generally formulated trading strategy. The testing can be done in-sample, or walk-forward out-of- sample. The arising complexity of such optimization problem is resolved through massive parallelization of computations across many cores on a given server or even across clusters of available servers.
  • Efficient portfolio optimizers based on trading equity drawdown being used as the main risk measure.
  • Trading robots environment that can execute any of the above tested strategies across various FCMs, connected to the middle-office and back-office software to automate and simplify end-of- the-day processing with minimal human input.
  • Unified futures database with various types and resolutions of data going back as long as the dada exists.

Research Notes: