"I just developed a new model of trading, and "tested" it using a real case. The stock on which it was tested is AA. AA released its earnings, so I thought to use the model to make a prediction for next day (which was today Thursday). A central point of the model is to predict a price level (called gravitational price GPL) . The gravitational price is a price such that the next day, we will observe a price above GPL, and a price below GPL A way to trade the model is for instance to look at the open price, and if it is above the gravitational price, you short because the model insures that there would be a price below GPL. There are other ways to exploit this further, but this is a basic application. Application for AA: On Wednesday evening the model said that the next day GPL for AA would be 14.65. AA traded at open at around 15.05, and a bit higher than 15.05 before open. If one is to follow the model one would have shorted AA at around 15.05, and waited until price is below 14.65 to cover. The model does not predict the drawdown (if any) and the ideal to enter and/or to cover. I have other models for those type of analyses. What is your first reaction to this model, and do you know if others have worked on a similar class of trading models? Any other useful/constructive comments are welcome." This article was first published on my financialtraders blog.