Outline of our methodology (click thumbnails to see full sized charts)

• MathInvest examines all stocks in the Russell 3000 in a variety of ways. Current financial statement information is obtained and 16 factors are compared for each stock in the Russell 3000 universe. This first screening enables us to eliminate stocks which have high individual risk based on their fundamental factors. The remaining universe is then examined on a quantitative risk-adjusted basis. All stocks in the remaining universe are compared against each other, and forecasts are created for each stock. Lastly these forecasts are ranked according to several measures and a combined rank is created for the final list of stocks. Portfolios are then created for the top-N stocks. According to client requests, N can be as few as 20 or as large as 200. These rankings are performed every day, and the portfolios are changed to reflect the rankings.
  
  
• Stocks are never deliberately sold; they are merely replaced with other stocks that have risen to higher ranks. That means profits are never deliberately taken, and stop-loss orders are never used. This policy removes the most difficult decision in investing: when to sell. As the ranking process does its work on a risk-adjusted basis, no subsequent portfolio balancing needs to be done.
  
  
• Additionally MathInvest has created 5 “strategic overlays” which identify systemic risk in the equity markets, during which no equity investment is prudent. During those times the program signals a switch to alternatives such as Treasuries, or (in 2008, for example) to cash.

• Everything MathInvest recommends has been thoroughly researched. There has been no optimization of input variables. Operationally, all of our inputs are adaptive.

Our current Long Only program - Actual month end results since 200704. This is our more aggressive program	The pro forma results of testing our conservative long only strategy from 2003 to present.