Systematic Behavorial Global Macro
Minimum Investment |
$ 5,000,000
|
Management Fee |
2.00% |
Performance Fee |
20.00% |
Summary
Saleem Mahjub, Portfolio Manager
Saleem began his trading career at Tradelink in 2005 after graduating from the University of Illinois with a degree in mechanical engineering, and first found success in the Chicago prop trading world. Relying on his background of engineering and software development, he started creating the framework of the quantitative models that would become the Systematic Behavioral Global Macro strategy. His belief in a disciplined and systematic approach have been reaffirmed through various market cycles, including the 2008-2009 Financial crisis, the zero interest rate policy (ZIRP) markets of the 20-teens, and the market volatility induced in 2020 by COVID lockdowns. With more than a decade of experience as a systematic trader, spanning numerous market regimes, Saleem has developed a deep insight into the discovery of alpha, realistic back testing procedures, robust portfolio creation and efficient risk management. His consistent risk-adjusted results combined with his approach to risk management have made him a valuable asset to the ARB team.
Statistical Pattern Recognition
> The SBGM strategy, using as much historical data as possible, looks for anomalies that indicate a herd mentality is forming in a given underlying market.
> Numerous sub-strategies (currently 13), each looking for these outlier herd indicators in their own way, are fed into a decision engine that uses the composite results to generate potential trade signals.
> The trade decision engine uses the input from the sub-strategies to generate a consolidated trade signal. Inherent in that signal is direction, strength (conviction), and sizing (risk).
> Individual trades are all short-term in nature, ranging from minutes to a few days, and in rare cases more than weeks.
> During the modeling phase, the stability and quality of the underlying sub-strategies are analyzed over a long period of time, with the goal being to optimize each signal schema with respect to various performance metrics like sharpe, sortino, max drawdown, etc.