Quantcast
  1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

OV accuracy in down moves

Discussion in 'OptionVue Forum' started by DavidF, Oct 1, 2016.

  1. DavidF

    DavidF Well-Known Member

    I´ve heard Jim Riggio use the broken handlebar on his motorbike analogy to describe his need to compensate for the fact that his platform (TOS I beleive) underestimates the vega hit on a down move. Thus his P/L would be about half what the model predicted.

    I´m finding the same thing with OV. However in contrast to the P/L line, if I look at the actual OV greeks of the position and make a mental calculation they seem accurate.

    For example on Wednesday my position was up about $3k, and OV predicted it would be up over $4k with a drop to 2150. Delta was about -50 at 2170, and -70 at 2175 so with delta only a profit could be expected, as per the t=0 line.

    However the vega is -1190 to -1900 between 2170 and 2150, which would predict that for every 1 point increase in VIX the position will lose about $1k to $2k.

    On Thursday my P/L was at $390 which is reasonably accurate if I take the vega into account, a 2 point jump in VIX taking about $3k off profits and the delta providing about $1k.

    It´s as if the T=0 line is only calcultating delta and ignoring vega.

    I´ve attached Weds and Thus positions below. So my question is this:- is OV also a broken handlebar motorbike model? I understand it´s hard to model as it´impossible to know what change in VIX will be associated with any change in SPX (e.g., Brexit where prior to the referendum VIX spiked with little change in SPX).

    Screen Shot 2016-10-01 at 09.00.21.png Screen Shot 2016-10-01 at 09.01.04.png
     
    Last edited: Oct 1, 2016
    Jack likes this.
  2. AKJ

    AKJ Well-Known Member

    be aware that VIX may not be the best, and in some cases, flat out misleading measure of volatility for your options trade. Your 50-DTE options are likely to experience different changes in volatility than the interpolated 30-DTE options that are used in calculating the VIX.
     
    Paul Demers likes this.
  3. Gabor Maly

    Gabor Maly Well-Known Member

    I have recently added SPX to my M3 arsenal and as opposed to RUT I also find that I would manually need to adjust volatility (in OV) to get a closer approximation of P&L even on a non-event down move. I have not seen this effect in my earlier backtests but would need to go back and analyze more. For sure vol has been very low lately and skew relatively high ....just like for RUT.
     
    Jack likes this.
  4. Steve S

    Steve S Well-Known Member

    No, because change in Vix is only partly change in vol and is mostly just "curve riding" for normal market moves, even if all your options are exactly 30 days. You need to at least consider only the "real vol change" portion of Vix, although the best solution is simply to always look at the real vols of your real strikes. I'm supposed to post a long boring thing next week that has a section explaining the Vix stuff in detail ... look for a post from me that appears to be ignored by everyone.
     
  5. DavidF

    DavidF Well-Known Member

    Thanks for your replies guys. I understand VIX isn´t correct term to use here, I should have written a 1 to 2 point increase in volatility. However, it still looks like OV´s t=0 line is based on deltas and not vega. Have never really grapsed the volty% change box and how to use it in relation to the vega values in the matrix, and whether or not OV are just providing a vega value that you can then manually input predictive values.

    However, that is also going to be off as it suddenly dawned on me after posting that Kevin Lee has already covered this for me in an earlier post

    https://forums.capitaldiscussions.com/threads/m3-vs-bwb-a-reductionist-approach.620/
     

Share This Page