Monday, February 22, 2010

VIX Fair Value Model

I created a model which comes up with the fair value of the near-tem+6 VIX value.

The CBOE website has the historical VIX term structure.

As I mentioned before there is a strong relationship between high yield OAS and VIX.

Here is the model vs. the actual near-term+6 VIX:

I'm using the high yield OAS as the only input for the model. The R-squared is an impressive 94%.

So, what is this model telling us today?

The fair value for the Dec VIX is about 24.95%. The actual as of 2/19/10 was 23.95%, so about 1% lower than the model tells us.

As you can see this is the Dec VIX fair value, the VIX spot and the near term is a lot more volatile.

I'm working on a model to come up with a fair value for the shorter term VIX as well.

Let me know you thoughts.


  1. Hey,

    Great blog, congratulations!
    Where did you get your OAS data, is it freely available somewhere?
    Also, what exactly do you mean by "near-term+6"?
    Thanks, and apologies if these questions don't make any sense.

  2. By near term+6 I meant the 6th expiration. For example, today that would be 12/18/2010. You can find more info about the VIX term structure on CBOE website. Send me an e-mail to if you want the OAS data.

  3. So it's the VIX futures 6 months out? Great, thanks very much!

  4. Sorry, I meant 6 contracts out. Great blog by the way, very informative and thought provoking.

  5. Robert,

    did you decide to drop this blog? Why? I thought you were writing some good things. This VIX model is interesting, I would like to discuss, give me a ring.

  6. Have you heard about spurious regressions

  7. i trade vix futures exclusively. if you could share some fair val modeling i am certain i can add equal insights in return.