Saturday, March 24, 2012

2012.03.23 Weekly DJIA $ Gold $ Silver

Very Short Time Frame Predictions

 The Gold model has been working well, now  predicting significantly lower prices over the next month. $1450 +/- a lot looks possible.

The Silver model is also working well, suggesting some upward pressure on Silver - to the  upside of $35/oz in the first few weeks of April.

The DJIA model is probably broken - but I will continue to run it and post it so I can look like a genius if the DJIA drops 2000 points in a short time frame.  I don't have time to even think about constructing another DJIA model for at least  the next few months.

The DJIA model insists on treating the current  actual price curve as "noise" while the prediction curve bounds downwards. I will explain this briefly. If the model detects highly non-random time series behavior during a window of analysis, it weighs those points less than if the time series is exhibiting its usual and customary non-linear signatures. So - as long as the DJIA time series that "takes off" the predictive curve continues to exhibit an odd signature, the model will continue to discount them.  You don't have to be a mathematician to appreciate that the "texture" of the time series of the DJIA is much smoother than usual. That is exactly what is being translated into something the model can handle. This is not an apology for a bad model - it is what it is. There were very good reasons to include that module in the model.  As long as the model continues to run - I'll keep posting it.

Have a nice early Spring or Fall week!

Visit for free sample trading model downloads.