Some VXN thoughts:
Even though the index has only been around for a couple of years, after reading through the explanation of the VIX by the same guy who designed the VXN(thanks for the link, Augie), I went back and did the same analysis he did in his paper to quantify the VIX's ability to reflect the fear in the options markets for the VXN and the NDX.
I did two regressions, one for 2001 to October of 2002 and the other for October 2002 to the present. It does appear that the VXN is decreasing as a signal of price movement, as some have suggested. It is still significant, but the movements are shallower, both on the upside and on the downside. Here are the equations:
2001 to October 2002
.62 + (-14.10*Positive VXN) + (-9.41*Negative VXN)
October 2002 to Present
3.09 + (-8.79*Positive VXN) + (-6.24*Negative VXN)
Since May 20th, the downside predicted values have been routinely coming up too high relative to what has actually happened. Since there are only a few trading days since then, it's impossible to get a good read on whether the relationship between the VXN and the NDX has changed since then. Basically, we should be about 120 NDX points lower than we are now, given how the VXN has moved since May 20th. Of course, almost half of that came today, but it should still be a concern for bulls, because the VXN's movements still have very strong predictive power for price action in the NDX.
When the predicted value for a given day is as many points different from where the regression equation predicts it will be as it was today, the average point drop over the next five sessions is 47, which would take us back to retest the 1230 level over the next week.
Obviously, this is just one indicator, but given today's odd action in the option market relative to the NDX, I wanted to do some further analysis.
AI
Market Data 
Markets 
