Correlations (10 day) closed the week at .62, basically the lowest point of the move thus far. It is important to remember that long term correlations are still deep red, and that is going to take continued low short term volatility for an extended period of time to change that.
The VIX remains very elevated relative to correlations, nearly tapping the 95th percentile band just below 30. Our data suggests that in this band VIX is lower about 94 percent of the time averaging 17.1% against Friday’s reading of 27.35% with a T-test of about 37 (a definitively non random occurrence).
We look at this from two angles. The first angle is that VIX is too high. The second angle is VIX is too high, but should we be paying attention to it? This is demonstrated by our regime switch indicator. Basically, if realized volatility and correlations continue to decline, and at the same time VIX maintains its level, we conclude that VIX should not in fact be lower. We noticed our first uptick in the signal last week and we print a blue dot if it hits the 95th percentile (similar to bollinger bands, but dealing with the ratio of VIX/correlation, rather than price – and interestingly this signal works the same way for both tops and bottoms).
Last week we wrote about contango nearing extremes. That ratio normalized in the week with VXV remaining elevated to VIX.
Finally we saw liquidity conditions nearing extreme liquidity prints on Friday again, we would not be surprised to see another print this week – market participants are shuffling positions driving high volume but little price change.
To summarize this post, what we think will be most important in the coming weeks is whether or not VIX stays elevated relative to realized volatility and average correlations. If that happens, we believe the trend may be about to reverse. Until that happens it looks like continuation is probable.
Thanks and have a great week.