Friday, November 11, 2011

2011.11.11 Weekly Short Term DJIA $ Gold $ Silver

The quick convergence of all three models suggests that if they are wrong, they are wrong in the right way. On the other hand, if they are still predictive, they are all stretching to historical maxima. This suggests if there is a reversal, it will be quite steep.

Index / Actual / Predicted / Z-score
DJIA / 12153.68 / 11252.09 / -1.52
Gold / 1764.00 / 1582.93 / -4.67
Silver / 3377 / 2906 / -3.85

To save time, I will only post the long term predictions on a monthly basis, since week to week differences are trivial. If there is a major phase change in the model, as occurred with silver a while back, I will post it and note it. Have a great weekend!

Visit for free sample trading model downloads. 


Anonymous said...

Thank you for your time, hope your predictions will come true. Markets are manipulated big way hence the current non compliance with your model, I think. Anyhow I am regularly checking your site.

Regards, Lili

Paolo said...

Thanks Lili. Every time I run an update I keep my fingers crossed. While one can dispute the validity of a model based on its mismatches with reality - if it does not converge, then its game over for that model.

Rajeev Bharol said...

How many times in the past did you have to change the model for gold?

Paolo said...

The current model has been active for 13 months. Prior to that the model was rebuilt once in 2009 for non-convergence, and a child of the new model instantiated in 2010. So that would make three changes.

There were 23 versions of a beta model, some were new and some were child-of-parent models with additions and subtractions to the structure. The major improvements were concerned with determining the best way to link non-linear short-term pattern signatures to the longer cycles.

The current model is a compromise between weighing long-term factors and the ever-changing patterns of short-term model behavior To address the latter, the model adapts to the short-term price changes which in turn feeds back to the longer -term parameters.

You probably didn't need or want such a long answer, but there it is!

El Viejo said...

Is model any relation to this:

Paolo said...

I think the idea is informed by chaos theory, but I suspect there are some differences in methodology. I have wrestled with the multifractality of these time series - both price and time. Both the time and price scales can be mapped unto an N-space, and disassembled there, then brought back to the static scales of time and price. In N-space both scales are stretched and compressed to see if a pattern pops out. It sounds more complicated than it is because some of the methodology remains proprietary, preventing complete disclosure. I apologize for that - I hope you understand.

Rajeev Bharol said...

Thanks. Can you please enable "Mobile view" on your blog. It will be easier to access it from the phone.

You can do so in the blogger settings..

Paolo said...

Rajeev, I enabled the mobile option. Thanks for the nudge.

Anonymous said...

Thank you for your blog. Knowledge is one of the few assets that grows as it is shared. The Model is responding. RED