Thoughts on the VIX Futures Term Structure
Fri, Aug 19, 2011 | Jared Woodard
Here are the VIX futures at 3PM ET on Friday the 19th.
click to enlarge
A few things stand out to me:
- Backwardation forever: I don’t have all my data at hand, so I don’t know where recent turmoil ranks historically in terms of time that the VIX futures have spent in backwardation. My guess is that this isn’t the all-time record, but that it is a contender. That five point difference between September and November is incredibly rich if the world doesn’t end in the next few weeks; otherwise, it’s a great bargain. The trade is to sell, like, one September contract and buy one or two Novembers (where you might normally trade five or ten), and to set a hard stop.
- Christmas cheer: December is priced a full point below both Nov. and Jan. What gives? I know the holiday season is a bit quieter, but that seems like an unusual difference.
- Prediction: in March 2012, the implied volatility for April 2012 SPX options will not be 29%. It might still be 23% or 25% if we get more fireworks later in the year, but not 29%.
- Flatness: we’re all focused on the backwardation in the front part of the curve, but when was the last time the non-near term months were this flat? There’s no rational agent model or utility function that makes sense of this. What brings it all into focus is the model based on that time we all got drunk the night before our SAT exams and then ran out of time on the back half of the math section, so we just filled in the same answer for a bunch of questions before the clock ran out.
- I find myself caring less and less about the spot VIX. Maybe it’s because kids on Twitter keep confusing $VIX for some kind of technical analysis holy grail, or the occasional CNBC guest who acts like you can trade the spot level, or just because the actually tradable products provide more meaningful information in my opinion.
Tags: VIX, vix futures, vx
August 20th, 2011 at 2:36 am
Flat non-near term term structure is not consistent with any rational model, huh? Isn’t that the assumption in basic Black-Scholes that volatility is constant for all time? In the absence of other data, prices are a random walk. I find the usual contango term structure strange, because it says we think our random walker gets more and more drunk and agitated in the far future. Anyway, I guess the expensive and flat term structure is simply consistent with a long term pessimism where people have no clue how or when anything is going to change…
August 20th, 2011 at 6:16 pm
I trade these extensively. How about over 15 points during late 2008? 5 and 6 point spreads were the norm from late 2008 to early 2009. I have no idea when contango returns, but spot vix will have to run fast to highs 20′s imo before that can occur. The futures really took their time getting out of mid 20′s. This should last over a month more at a minimum.
August 22nd, 2011 at 1:52 pm
In 2008, the backwardation lasted for a couple of months. In the next months, we will probably make a lower low on the SPX and if VIX will be making a lower high, that will probably mark the returning to contango.