Just five days later, the VIX seems to have peaked, yet the amount of backwardation in the VIX futuresterm structure has actually increased. Looking at the front two months of VIX futures (which is where investors in the likes of VXX and XIV should be focusing), I note that the front month (August) is now 7.25 points higher than the second month (September) VIX futures. This positive roll yield means that investors who are short VXX and/or long XIV are losing almost 1% per day due to daily rebalancing (rolling) that involves selling the front month VIX futures and buying the second month contract.
This also means that should the VIX spike higher from current levels, ETNs such as VXX of TVIX and others should see enhanced returns due to an increase in volatility plus favorable term structure and roll yield.
One problem with backwardation is that it tends to be fleeting. Of the 59 instances of backwardation in the front and second month portion of the VIX futures term structure going back to the inception of VIX futures in 2004, 37% lasted only one day and 56% lasted no more than two days, fully 83% of all instances of backwardation had ended within six days and only six backwardation events in seven years have lasted more than the current eight days. Not surprisingly, three of those six periods of extended backwardation were from 2008, two were from 2009 and the last one was from 2007.
To state what I hope is the obvious, detailed knowledge of the workings of the VIX futures term structure is mandatory for anyone who trades VIX ETPs. Not only does one need to know what the implications are of the current term structure, but also to have a sense of how that term structure is likely to evolve over time.