MarketTrak Question/Comment Message

Posted By: Tom D   Date: Thu Feb 14, 2008

Title: models of models of models?

  I find the recent management of your multiple model selection and juggling, and market recommendations based upon them, extremely confusing. Not intellectually confusing, but, from the perspective of someone trying to use your output for trading, there seems to be little logic presently expressed. You are now flat in the market while your best models are clearly long. And you have already conveyed that the model will switch to a currently bullish model this coming weekend. Do you need perhaps a "model of models of models" to decide when and which model is actually operative at a given time? I'm not convinced that a rules-based model regime-switching method really works for markets.

  It has taken more time than I expected to decide on whether version 7 or version 16 would be the one that we would use as the standard model. Based on the trading results over the last several months, forecast model version 16 has proven to be the best. On a time scale of several years, the trading return of versions 7 and 16 are nearly the same. Even version 10 did very well over the long term. The confusion should end this weekend when we make the switch to version 16 as we will no longer need to present the results of multiple models. We will base our trading logic solely on version 16 forecast results.

  I don't think we need to consider a model to switch between forecast models. It would be far too confusing and I don't see that we would gain much. The networks already switch internally between logic elements based on market conditions.


Return to message board

Bag Toss: high quality, made in America, and makes a great gift.

Copyright © 1996-2018 MarketTrak LLC.  All Rights Reserved.