This forum is for traders to share information about OptionVue analytical software.

Combine put/call skews

6 replies, 4 voices Last updated by Andrew John 5 years ago
Viewing 5 reply threads
  • Author
    Posts
    • #9453

      Andrew John
      Participant
      @ajohnnd@gmail.com
      Points: 544
      Rank: Freshman

      So I’ve been tracking this in OV for some time, and I’ve come to believe that any differences in IV between the calls and the puts can be tied back to the Dividend Yield that is used to measure the IV.

      As far as I can tell, OV uses a single dividend yield to measure the IV’s and model the position regardless of expiration. In reality though, the dividend yield on indices is not static across different tenors, as the Index constituents go ex-dividend at different times throughout each quarter. This can lead to notable differences between the implied forward price that is calculated using the current spot, dividend yield, and risk free interest, and the implied forward price measured from the option chain. I’m inclined to believe the implied forward that is measured from the option chain is more accurate because there the differences between strikes is usually in the cents, while the difference between the implied forward based on a single dividend yield and the implied forward measured from the option chain can be in the dollars.

       

      I believe OV is using the IV’s that are calculated based on this single dividend yield to model out the position in the Analyze graph. This can lead to discrepencies in the analyze graph between Call/Put/Iron butterflies, especially for expiration months where this single dividend yield is not an accurate measure of the forward price of the index for that month. This is the reason I personally prefer to use Combined Put/Call skews in all circumstances.

    • #9430

      John Locke
      Keymaster
      @john-locke
      Going to The Trading Triangle LIVE 2016Locke In Your Success CoachAPM2Trade JournalsUltimate Income Trader Workshop

      David,

       

      Yes, you will get different Greeks for put/calls/irons.  Theoretically with the combined skews Greeks for put/calls/irons should be the same. With OV sometimes they are, sometimes not. I believe this is a program bug but even when they are not the same, I have found them to be close.

      • #9437

        David Fraser
        Participant
        @fradav68@gmail.com
        Points: 258
        Rank: Freshman

        Thanks John. Considering the importance of the issue and data avaiable it should be easy for OV to run an empirical assessment of the historical accuracy of combined call/put skew ticked vs. unticked on M3/iron fly positions.

        Based on a very small sample (SPX takes a long time to backtest, even if I run it as local data) I found that OV is actually very good at predicting price change in all-put BFs. However it´s not good with iron flys and if the combined call/put skew box is unticked it´s way off (negative deltas far too high). Again a small sample, 3 periods in 2016 (March-Aug) with multiple time points where I compared predicted and actual prices.

        It suggests that the call skew/pricing is difficult to model in low volatility, at in last 6 months, could be completely different in other periods. In the meantime I think I´ll stay away from iron flys.

    • #9429

      John Locke
      Keymaster
      @john-locke
      Going to The Trading Triangle LIVE 2016Locke In Your Success CoachAPM2Trade JournalsUltimate Income Trader Workshop

      Sorry for the delay in answering this..Yes “theoretically” put and call IV should be the same or at least within a tight range of 1/2% or so. And typically you’ll see that with near ATM options.

       

      Pressure, lack of pressure and bid/ask spreads will change the time premiums in individual options and thus calculate a higher or lower IV level on that option, especially as you get away from the money.

       

      In the not so distant past, when things were less efficient, big traders could take advantage of these differences and put on huge no risk or guaranteed profit trades (IE take advantage of the market makers) Today, the computers are so fast and efficient that if the traders try to take advantage of the skews, they are corrected so quickly they can only get a small position, if anything at all, and it’s not worth it. Therefore very large skews can develop. These large skews can produce drastically different Greeks readings and T+0 profiles.  As to which one is correct? Neither, usually.  You’d need to monitor the IV’s on each strike and make your best guess as to when one or both are out of whack. OR you can combine the skews and average it out.

      Len told me he put the combined put call skews in in for lightly traded assets where put call skews are common and he knew the projections of either the call or put were off and that they were not needed on an index because the skews didn’t happen.

       

      Well obviously it is happening on indexes now so I use it for that purpose.

       

       

       

       

    • #9418

      David Fraser
      Participant
      @fradav68@gmail.com
      Points: 258
      Rank: Freshman

      Would also like to understand this. With respect to postion deltas you can test for yourself on OV that ticking the combined call/put skew box makes little difference (at least on SPX) on a all put fly, but a big difference on an iron fly. You need to tick the combined box in the iron to match the put fly deltas.

      For example a Nov 20 lot 2025/2125/2225 is (1) minus 429 unticked and minus 321 ticked for the iron and (b) minus 339 and 308 for the put fly ticked/unticked. That´s a $210,000 notional difference on the iron.

      I always tick the box as I understood they should be synthetically identical (iron vs. put fly) and I thus trust the put position delta value. However, if persistent skews do exist (as you describe Andrew) then this might not be the correct way.

    • #9410

      Andrew John
      Participant
      @ajohnnd@gmail.com
      Points: 544
      Rank: Freshman

      “There are often differences between the call and put skews that persist over time in both futures and equities, especially the indexes and things like the VIX. By not using combined, the skews are a closer match to the calls and puts and result in calculated theoretical prices (fair values) that are closer to the actual prices being traded.”

       

      I’m wondering if we can dig deeper into this concept a little. I am of the opinion that skews should be identical for calls and puts, unless there are market frictions (such as the borrow on a short, inability to short, or discreet dividend). Can someone elaborate on how and why skews/IVs would be different for calls and puts. For simplicity, let’s limit to scope to Index and Futures options, where there tends not to be any frictions against shorting.

    • #9409

      Sherri Locke
      Keymaster
      @KDAKSj4ux
      Going to The Trading Triangle LIVE 2016Ultimate Income Trader Workshop

      Len says no on “combine put/call skews” and John says yes. When is it better and why?

       

      There is not a definitive right or wrong answer to this. The important thing is to be consistent, and if you are used to trading (or your system has been backtested and designed) using the combined skews you should use that. There are often differences between the call and put skews that persist over time in both futures and equities, especially the indexes and things like the VIX. By not using combined, the skews are a closer match to the calls and puts and result in calculated theoretical prices (fair values) that are closer to the actual prices being traded.

Viewing 5 reply threads
  • You must be logged in to reply to this topic.