Research ~ Global Quantitative Financial Research by Mark Brown

Sophisticated systems elegantly programmed yet simplistic and not beyond the comprehension of ordinary human logic.  Developed using proprietary custom analysis software and hardware created to deep data mine for the highest probability reoccurring market anomalies.  The human intellect is capable of understanding the results of deep data mining market research.  However it would be impossible for a human mind to ever discover or conceive these replicating anomalies from observation alone.  

Expertise in computer modeling and exhaustive knowledge distinguishes the typical empirical derived system from a deep data mined model.  Quantitative computer analysis combined with data cleansing allows the true nature of the subject market to be revealed.  Enormous financial commitment is required with no guarantee that replicable, let alone tradable market anomalies can be found or exploited.  Yet to date there have been no liquid markets which have not yielded profitable models in multitude.       

Myths abound that mechanical systems require continual adjustment to adapt to varying market conditions.  Properly designed trading models will enjoy many years if not decades of continual profitability.  Again this is what separates empirically designed models from superior data mined modeling.  Much of the success of discovery can be directly attributed to acquiring the knowledge to properly filter data.  This technique alone can be credited with a noteworthy percentage of the profits derived from this method of modeling.  

Experience and capital commitment play a large role in the longevity of modeling.  These types of models are the direct results of countless man-hours and millions of dollars spent in research spanning well over a decade.  Yet all that effort could have easily been wasted had the development process not been pursued with a feverish tenacity as it is to this day. 

Great discovery's sometimes come about when researchers stand aside and let the process reveal the truth of its final analysis.  Data mining and complex analysis are difficult enough without tainting the whole process with human preconception and emotions.  Thus human intervention has proven to be the least desirable tool in the modeling research inventory.

A true story : In pursuit of the Holy Grail.