Maybe I should have read the CBOE whitepaper but I never did... Now I cannot find it.

I was looking at SPY/SPX just a moment ago and noticed that spot VIX doesn't match with SPY or SPX IV. Why not? I always thought that is what VIX was.

I can see current spot VIX at 10.71 at the moment while SPY/SPX IV is at 8.4%

EDIT!! this is just further info based on accepted answer.

SPY option chains look like someone thinking the world is gonna end.

enter image description here

  • $\begingroup$ I also checked VIX's own option chain that states IV of 76,5% which kinda makes sense. I mean volatility index is bound to be volatile but it does not help me to understand what VIX actually is. $\endgroup$ – Kimmo Hintikka May 4 '17 at 15:28
  • $\begingroup$ I don't see anything unusual in that graph above. $\endgroup$ – FinanceGuyThatCantCode May 4 '17 at 20:34
  • $\begingroup$ @kimmohintikka There is also a vvix that sits pretty steadily around 80. $\endgroup$ – will May 5 '17 at 7:37

What is your definition of SPX IV? ATM? ATM vol will be less than VIX. VIX is calculated by pricing a variance swap on SPX. A variance swap is priced theoretically using all strikes from 0 to infinity appropriately weighted - the final formula is an integral across all of the strikes. The VIX calculation uses the discrete strikes that are available in the market with weights that you would get from doing a numerical integration based on the strikes in the market. I believe the cutoff strikes for the calculation occur on both the put and call side at the strike where two consecutive strikes have no bid. Look up variance swaps on wikipedia - at the bottom there is a link to an Emannuel Derman paper that goes over the theory of variance swaps. Variance swaps levels are nice because it is not strike dependent for the options and gives extra weight to downside moves which is sensible given the terminal distribution of equities.

The CBOE paper: https://www.cboe.com/micro/vix/vixwhite.pdf

Wikipedia variance swaps: https://en.wikipedia.org/wiki/Variance_swap

Derman's paper: http://www.emanuelderman.com/writing/entry/more-than-you-ever-wanted-to-know-about-volatility-swaps-the-journal-of-der

  • $\begingroup$ AH ok my IV is automatically from ATM using IB TWS $\endgroup$ – Kimmo Hintikka May 4 '17 at 15:53
  • $\begingroup$ Got it - bear in mind that VIX used to be ATM vol as well - they used the puts and calls around the forwards level for each of the two expirations that wrap around the 30 calendar day theoretical expiration. That is a total of 8 options used to calculate the ATM vol. This was discontinued in 2003 I believe. This is all model dependent too for how vols are calculated which is a problem. The variance swap methodology uses mid prices that get plugged into a more complicated formula, but should be deterministic if two different people price it without errors! $\endgroup$ – FinanceGuyThatCantCode May 4 '17 at 16:00
  • $\begingroup$ It is true that since 2003 VIX is no longer equal to ATM vol, but a difference like 10.71 vs 8.4 seems far too large to be due to this. There must be other reasons (first of all, what is the maturity of the 8.4 figure?). $\endgroup$ – noob2 May 4 '17 at 16:53
  • 2
    $\begingroup$ @noob2 - his difference was due to the difference between ATM vol and var swap fair strike. I just added the extra comment about the old VIX method as an FYI if anyone ever looks up time series of VIX calcs. Overall - the VIX spot calculation is highly flawed for many reasons - calendar time calculations are bad as well as certain odd jumps that can occur when rolling to next pair of expiries used in the interpolation. The worst case is when T1 is 35 days and T2 is 63 days - in that case, they do an extrapolation to 30 days which causes big gap changes in VIX without a real change in vol. $\endgroup$ – FinanceGuyThatCantCode May 4 '17 at 17:46
  • $\begingroup$ There's also the issue of the theoretical varswap not matching the market varswap price. $\endgroup$ – will May 5 '17 at 7:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.