Thanks. This must be for serious statisticians. I did some reading on wikipedia and had to keep clicking on more unfamiliar terms. I may just stick with "r".
Interesting possibility for a pair. From Vix and More blog. Is anyone trading VIX/VXV? ------------------- Tuesday, October 12, 2010 VIX Sets Two New Records It is not every day that the VIX establishes some sort of new all-time record and it is rarer still that the volatility index sets two different records on consecutive days, but such has been the case at the beginning of this week. The first record, established on Monday, was in the VIX:VXV ratio â a subject that I covered on a regular basis in the first year or so following the launch of VXV, which is essentially a 93 day version of the VIX and whose formal name is the CBOE S&P 500 3-Month Volatility Index. Long-time readers will recall that for the first year after VXV was launched, the VIX:VXV ratio performed flawlessly (see VXV Is One Year Old.) In a post-Lehman world, however, the VIX:VXV ratio has been inconsistent as the persistent extreme contango has made the ratio more difficult to calibrate. I have been doing some work on this, however, and will share some of my thinking about how to tweak this ratio going forward. The latest record, established today, is in my proprietary VIX Futures Contango Index. I spelled out some of my thinking about the epic disconnect between the cash/spot VIX and the VIX futures in a post a month ago with the title of VIX Futures: What Are/Were They Thinking? In the end, the VIX:VXV ratio and the VIX Futures Contango Index both measure different aspects of the same phenomenon: how much current volatility readings vary from future market volatility estimates. I do believe that when estimates of near-term and long-term volatility show a record degree of divergence, some considerable opportunities are presented. As I have spelled out in a number of instance lately, my thinking has been that the back month volatility will likely collapse in order to bring the present and the future back into line. There has been some evidence of that happening during the past two days, but I anticipate that long-term volatility expectations will continue to decline. On a related note, VXX has made a new all-time low six days in a row and counting... http://vixandmore.blogspot.com/2010/10/vix-sets-two-new-records.html
saw comments about this trade on zero hedge yesterday. http://www.zerohedge.com/article/1m-3m-volatility-term-structure-plunges-steepest-years-vixvxv I thought I would collect this with spread on options
More thoughts on mean reversion with VIX/VXV (Note that this is copied from a 3 year old blog post) =============Tuesday, December 11, 2007 The VIX:VXV Ratio Yesterday I talked a little bit about what the CBOE has said about the VXV. Even though it is still early days, today I thought I would offer up a simple framework that might be useful for using the VXV as a timing tool. The chart below covers the first month of data from the VXV and calculates a ratio of the VIX to the VXV (the CBOE chart from yesterday chose to use the ratio of the VXV to the VIX, but I generally prefer to have the more volatile number in the numerator and the less volatile one in the denominator.) I expect that the VIX to VXV ratio will make it easy to determine the extent to which the implied volatility on SPX options suggests investors expect volatility to rise or fall in the 30 day (VIX) to 93 day (VXV) time period. In addition to the 10 day simple moving average and 10% and 20% moving average envelopes, I have included three horizontal lines in the chart below. The dotted black line is set to 1.00 and indicates no expectations for a change in volatility over the 30 to 93 day time frame. The two solid black horizontal lines are set to 0.90 and 1.10 and are intended to be easy visual references to indicate when volatility is anticipated to change by at least 10% in that 30 to 93 day window. The 10% level is somewhat arbitrary and largely dependent upon a desired signal to noise ratio, but it is supported by the CBOE data I highlighted yesterday and is consistent with much of my other research on the VIX. Just as is the case with VIX futures, I expect the VIX to have a tendency to fall and the markets to rise when the VIX:VXN ratio is above 1.10; and will look for the VIX to rise and the markets fall when the ratio is below 0.90, in classic mean reversion fashion. http://vixandmore.blogspot.com/2007/12/vixvxv-ratio.html
check that: http://seekingalpha.com/article/198661-a-vix-etf-pair-trade-vxx-vxz?source=feed it works until the markets goes down