A secondary indicator that I watch is the S&P 500 Commitment of Traders [COT] report. I used to think that it was bullish if commercial traders (the so-called 'Smart Money') were net long S&P futures and bearish if they were net short. Having analyzed a subset of the data, I'm no longer so sure. In fact, it appears to me that the opposite is the case; I should be bullish when commercial traders are net short and bearish when they are net long. I have analyzed the COT data going back to the January 2004 and found that commercial traders were, with the exception of a handful of weeks (7 to be exact), net short the S&P futures from January 2004 through April of 2007. During this same period in time the S&P index itself advanced from 1,123 to 1,486, a rise of ~32.3%. By contrast, while commercial traders have been predominantly net long since May 2007 with the exception of 3 weeks, the S&P index has been flat, moving from 1,486 to 1,496 as of December 28th.
I used to think that the COT data represented commercial trader's speculative positions in the market. Now I'm beginning to come around to the point of view that the COT data represents a partial hedge of commercial trader's true positions, hence the inverse relationship between commercial trader's net position and the performance of the S&P. I have the COT data going back to 1994 and will have to make some time in the near future to see if the relationship I think I've discovered holds over a longer timeframe.