Principals

The founder of the Swarm, Nathan Mirsky, started his adulthood as an aeronautical engineer. He has managed enterprise-level programming projects for the Toronto Stock Exchange and raised capital for a chain of private medical clinics which he then led to IPO. His interest in Emergence, Marketing and, of course, Stock Trading helped lead to the creation of the engine and methodology.

The chief scientist, Satya Kumar, was initially trained for a career in pure Mathematics. Fifteen years ago, he found the available software for performing calculations and explaining Mathematics concepts inadequate and set out to write his own. Encouraged by the enthusiastic response from his students and colleagues and his successful freelance work as a programmer, he switched careers. His projects include a Graphing/Financial/Statistical calculator, online software for tracking fleets of vehicles, interfaces for RF tuners and numerous Windows applications and websites. He met Nathan when they worked for a common client who was customizing an interface for encrypting e-mail with PGP. He shares Nathan's interest in Emergence and the Stock Market.

Oleg Granitski, our CTO, holds an MSc. in Computer Science. His thesis entailed the use of a combination of programs for stock-picking using Technical Analysis, Neural Nets, Artificial Intelligence and Machine Learning. He worked as a database programmer for an enterprise-level trading site. He has created many websites with Satya and worked with him on the GPS-enabled tracking software and encrypted e-mail projects.
    
      
Principles

Every stock transaction from all the major Stock Exchanges in North America is input for our stock-picking engine. A sequence of filtering operations is performed. Stocks whose prices exceed $100 or are less than $0.01 are eliminated from consideration, the former due to their relatively limited volatility, the latter due to a wish to avoid trading companies with an elevated risk of bankruptcy. The next screening mechanism involves various Technical Analysis indicators. For example, our engine is programmed not to recommend purchasing a stock whose chart or whose sector's chart exhibits a Double Top. Of the generally fewer than twenty issues which have survived the filtering process, our proprietary software, based on machine-learning and Genetic Algorithms, attempts to identify the most likely to rise. If the predicted rate of change is sufficient (and some days, no issue satisfies our criteria), a signal is sent by text message to all our clients at the same time. The entire process is automated.

Our examination of our results over years suggests that it is still impossible to consistently be correct more than 90% of the time in our predictions. There are unpredictable factors which are outside the realm of our analysis which can have a significant on a stock or, even the entire market. (earthquakes, terrorist attacks, death of a company principal etc.)
    
      
Dedications

This site is dedicated to our friends and mentors:

    STARVIN' MARVIN (Sniderman)
    Dr. Leslie Evans (Lev)
    Momma Miller
    Max (The Sport) Novack
    George Scott
    Kevin Higgenbottom
    Armand Vergil