Last week, while some were sounding the death knell of the VIX, I posted an earlier version of the chart below that showed the SPX:VIX ratio had wandered well away from the context of historical norms and above the upper Bollinger band – something that usually happens only once or twice each decade.
Not surprisingly, yesterday’s action turned that relationship upside down, so that the SPX:VIX ratio now sits at the bottom Bollinger band and at a level relative to the SPX:VIX trend line (see that previous post for an explanation) that has not been seen since the latter stages of the dot com bust.
In the last two hours (which are not shown in the end of day chart below) the SPX:VIX ratio has bounced around in the 90-97 range, a move that brings it almost exactly halfway back to the 10% trend line and in close proximity to what I would consider a neutral range of 105-115. My guess is that it will be more of a falling VIX than a rising SPX that will bring these ratios back toward neutral over the next week or two.
I will periodically update this chart going forward and provide some commentary about what the ratio suggests about future market movements. Right now it suggests that the SPX still has a fair amount of firming to do relative to the VIX before it gets to an area where I would be concerned about the SPX taking another dive.
It will be interesting to see now if the SPX/VIX ratio can climb back above 110-115. The last time we saw something similar (in 2000), the SPX/VIX ratio was unable to climb back above the trend line, and eventually fell all the way down to 20. Do you see anything similar here as a possibility?
ReplyDeleteBrian,
ReplyDeletePart of the difficulty in plotting this trend line is the somewhat arbitrary point of origin, which compounds any 'errors' in placement over the course of the years.
In this case I just happened to begin the trend line where the Stockcharts monthly chart picked up back in late 1991. All things considered, it turned out about exactly what my intuition had predicted, which is why I have decided to keep it alive here.
In terms of how far the ratio might fall, I would look at Bollinger bands instead of absolute numbers and would be very surprised to see anything well below the lowest band.
Wild guess: we will probably not see this ratio get back down below 40 -- and a floor of 50 or more is fairly likely (largely because we have a trending number divided by an oscillating one.)