volatility confusion between ib and ivolatility

Discussion in 'Options' started by newguy05, Aug 5, 2008.

  1. For ticker etfc, on ivolatility, it states the HV(10-30days) is between 122% - 130% and IV index is at 90%. On IB optiontrader, it states HV is at 7.69% and IV 5.36%.

    The HV/IV ratio seems to be correct between the two, but does anyone know how the actual numbers are so different? I know they are using different models to calculate it, but still the numbers are not even close.

  2. Hi Newguy,

    Correct me if I'm wrong (I'm tired), but I remember having the same difference once because one was daily and the second was expressed in annualized terms. Check it out, maybe the same case.

  3. Euler


    One explanation might be that one is listing daily volatility, and one is listing yearly volatility.

    The formula to convert between the two is:

    YearlyVolatility = DailyVolatility * sqrt(252)

    where 252 is an approximation of the number of trading days per year.

    This "seems" to fit the values you give. That being said, many other factors could be behind this discrepancy...
  4. dmo


    252 is indeed the approximate number of trading days in a year. But the standard is to use calendar days until expiration to calculate option prices and IV, not trading days. Assuming both ivolatility and IB also use calendar days to expiration in their option calculations, I think you should use the square root of 365 rather than of 252.

    It is possible that IB is doing something strange like using trading days - I seem to remember that was the explanation once before when somebody wondered why USO IV was so different from the IV on crude futures options (as calculated by IB).
  5. Hi newguy,

    I'm glad you received some answers about standard deviation. Feel free to say what you understand behind this word, and what consequences you could forecast. There is a great misconception on standard deviation, please keep this thread going. It could be useful.


  6. I think the large difference is due to the daily vs annual you guys said. In the IB video, the lady said they use the 100-step binary tree to calculate volatility for american style options. And i know ivolatility does it differently per their faq. I vaguely remember how to code all those pointers in comp sci class for a binary tree lol.

    I guess the followup question is if looking at the volatility in the same view, meaning all numbers are done using the same model. Does it really matter what the actual numbers are? as long as their ratio/comparison remains valid. After all isnt that how you use volatility, comparing IV vs HV vs historical highs/lows, to determine your trade.
  7. Hi Dmo,

    Right. It's an old debate to know whether there is volatility during week ends since earth keeps spinning or only during trading days. The question is just to keep same numbers to stay relevant.

    There are as many implied volatilities as different models. The most important point is what is daily volatility?
    Newguy, you wrote an HV at 7,69%. What does it precisely mean an HV at 7,69% ?

    ( Very interesting thread, keep it going)

  8. The point is the rule written by Dmo is right only for standard deviation.
    And a standard deviation and a mean absolut deviation are quite different. Hence, you can't compare "IV vs HV vs historical highs/lows, to determine your trade".

    So, if you talk about a yearly volatility (or yearly implied volatility as we usually do) as a yearly standard deviation, you have to be careful since it's just conventional. It's a number and just a number.
    10% daily standard deviation=16*10% yearly standard deviation (I took 256 trading days)=160% yearly standard deviation
    It doesn't mean neither a stock could go up by a factor 1,6 by the end of the year, nor it could drop down by a factor 1,6 (it would have a negative price).
    Implied volatility as an implied yearly standard deviation is just an annualized number.

    A simple rule: daily mean absolut deviation=0,8 daily standard deviation (if returns are lognormally distributed)


    implied volatility =16*daily implied standard deviation=16*implied daily mean absolut deviation/0,8=20*implied daily mean absolut deviation.

    Thus, a common 20% implied volatility leads to a 1% implied daily mean absolut deviation.

    "After all isnt that how you use volatility, comparing IV vs HV vs historical highs/lows, to determine your trade." the answer is no.



    For fun http://www.leeds.ac.uk/educol/documents/00003759.htm
  9. i got the sd part, but gonna need some starbucks before trying to understand the rest of what you are saying :)

    But fyi in IB: under Configure -> Volatility and Analytics -> Volatility Units -> you can change it from Daily (default) to Annual.

    Then the numbers will match the ivolatility model very closely.
  10. I have looked and looked for the configuration to change the IV display from daily to annualized. What version of TWS are you using? I'm running the latest (885.7) and can't find the menu selections you listed.


    I looked harder this time and found it under Configure->Misc->Volatility Units

    Never woulda thought to look under miscellaneous. :cool:

    Having IV displayed daily has been driving me crazy for months.
    #10     Aug 8, 2008