Here's how I identified changes in status of the model:
The earliest date for which I was able to make this determination was January 1991, when he went 100% invested. If he was fully invested, then his model must have been favorable. Unfortunately, we are not able to make that determination prior to that, as he was only partially invested, and if he made any statements on whether the model was favorable or unfavorable, I don't have that information. On the other hand, in January of 2000, even though he only went 60% to cash, he explicitly stated "the model has turned unfavorable."
In October of 2000 he spoke of a "countertrend rally." By definition a countertrend rally cannot exist unless the trend is down, so it is evident that his model must still have been unfavorable. This is confirmed by the fact that in subsequent issues he made it clear by his explanations that the methodology he was using in attempting to identify a countertrend rally was different from the methodology of his model as he had previously explained it on many occasions.
In March of 2003 he again went fully invested, leaving no doubt about the status of the model.
