SVM Capital, LLC, announced today is making substantial progress in using Health Discovery Corporation’s (OTCBB:HDVY) sophisticated pattern recognition technology (Support Vector Machines (“SVM”)) for predicting the movements of the financial markets.
In November 2008 SVM Capital began live trading with its first model utilizing SVM technology. Trading a composite of the four major stock indices (S&P 500, Dow Jones 30 Industrials, the Nasdaq 100 and the Russell 2000), the broad market experienced a four-month decline of 23% through the end of February 2009. Beginning in March 2009, the broad markets rallied through December 2009.
During the four-month period (November 2008-February 2009) when the markets and the four indices were down 23% in the value, SVM Capital declined only 7.6%, outperforming the market by 15.4%. The SVM algorithm correctly predicted market downturns, which prompted SVM Capital to either hedge its risk or increase its cash position.
During the subsequent 10-month market rally in 2009, the composite indices made an impressive 55% gain. Meanwhile, SVM Capital gained 26%. For the entire 14-month period (November 2008-December 2009, inclusive) the composite indices gained 19%, while SVM Capital gained 16%, with much less risk.
“It’s worth noting that while the four composite indices outperformed SVM Capital by 3% during the 14-month period, SVM Capital’s gain of 16% was achieved with far less risk,” said Mark A. Moore, Ph.D., Chairman, SVM Capital, LLC.
“We are pleased with these risk-adjusted results and lower risk is very important to today’s investors,” said Dr. Moore.
New Improvements to Trading Model Underway
SVM Capital continues to incorporate additional equity markets data in order to improve the initial trading model. A more robust model is expected to be trading live in the near future. This new model takes advantage of the addition of information to our data base and demonstrates again the capacity of SVM to deliver improved results while adding information. The new model has also demonstrated the capacity to perform well during periods of very strong market rallies such as we have seen in 2009. For the period of live trading from November 2008 to December 2009, the new model outperformed the composite four indices by a small margin. The composite of the four indices was up a total of 19% while our new SVM model was up over 20%, with lower risk. Again, the new SVM model outperformed the indices with lower risk.