Software

Discussion in 'Options' started by spindr0, Mar 26, 2007.

  1. I bought the demo and was surprised at how crappy the valuation models are. At that cost, you should be able to overlay the results from models with stochastic volatility and price jumps, that calibrate to the market.
     
    #11     Apr 9, 2007
  2. The Pro version lets you manually specify volatilities for each strike and expiration.
     
    #12     Apr 9, 2007
  3. I would of add second PnL choice : separate for each month. This way you can look at the max loss and go back and adjust front/back month ratio.
     
    #13     Apr 9, 2007
  4. Just an example to better understand:

    say you model having both call and put calendars on:

    call cal's are Apr/May
    put cal's are Apr/Jul

    long or short, ratioed, never mind the position details.

    Then you'd like to see separate P/L development for

    - your combined Apr position (Apr calls + Apr puts)
    - your combined May ,, (May calls)
    - your combined Jul ,, (Jul puts) ?

    I'm no calendar expert, just trying to get your point.
     
    #14     Apr 9, 2007
  5. NP , I am definetly not advocating this enhancement...but here how I use it...
    I run a combine and separate (on the per-contract bases) PnL before I enter position. Lets say the worst case/max loss is at 0.34 cents on the custom End day. If 34 cents of combine loss breaking down to 1$ loss on front month and 66 cents gain on the back month , I will go back and readjust ratio to 2:3 for the entry.
    Just an idea ; this functionality could be useless to you/others.
     
    #15     Apr 9, 2007
  6. spindr0

    spindr0

    nonprophet:

    Thanks for your interest in this adventure. Here's what I have now.

    Not that I'd need so many but the program allows for entry of 28 options. 28 may be too many but sometimes, extras are needed because a 4 legged position might be legged into and there are fills at different prices for one or more of the legs.

    Once the spec's are entered, the data line calculates IV and delta. Other Greeks might be of interest to some here. I'm still working on English!

    The program calculates the average IV of all positions. This is pretty useless to me. I'd sooner see an IV calc of the average IV for positions held in each expiration month.

    I have the ability to enter a future look/see graph date. It could be tomorrow as with an earnings event or any date in the future, up to the expiration date of the nearest month in the position.

    I have the ability to enter a future IV guesstimate (AVERAGE IV for ALL poistions). This is the area that needs improvement. I'd like to be able to enter an expected future IV for each month involved in my position. I don't need it for each option, but hey, if you're going the distance, additional bells and whistles would be gravy.

    As I can do now, I'd like a risk/reward graph of the position (using the new guesstimated IV's) for the next day. If a longer time period is selected, multiple risk graphs (in different colors), evenly spaced over that time period.

    If you'd like a copy of this DOS shareware program that I'm using, in order to get an idea of what I'm trying to describe, ask away. Just remember, it's an antique :)


    "Are we nuts??"
    Ask the sane people that :)
     
    #16     Apr 9, 2007
  7. An apparently simple idea has me confused.

    I assume the 34 cents is based on 1:1 ratio.
    The model turns out then your front/back month P/L ratio is 3:2 ($1 : $0.66).

    So you adjust contract ratio to 2:3 so your front/back month P/L ratio becomes 1:1 .

    Is this the idea? I'm genuinely interested to understand this, if it'd be useful to you then it'll be useful to others (like me) as well. :)
     
    #17     Apr 9, 2007
  8. Running first what ifs at 1:1 , then ( based on my scenario from the above) entering position at 2:3 to have :

    Front month loss =1$*2
    Back month gains = 0.66*3
     
    #18     Apr 9, 2007
  9. Got that, but what is the specific advantage of (equally) spreading risk over different months?

    I mean, you don't spread risk over different strikes, so why over months? I just don't see that.
     
    #19     Apr 9, 2007
  10. The idea of a true (arithmetic) IV average ( (IV1 + IV2) / 2 ) seems beyond useless to me, honestly even per expiration month. Is that really how the DOS program does this?

    IMO the only slightly sensible way to "average" IV is a "combined implied".

    e.g. a 50/70 (IV) strangle. What's the "strangle IV"? 60? I'd say take teh strangle *market price* and search for the one IV that fits the bill.

    But the very question of avg'ing IV's of course collides with the BS assumptions so this is all dark territory. Or gruesome math with GARCH or local vols.

    For our practical use I'm having the same issue you describe with software like Hoadley. I think only realistic improvement would be to set IV for each separate leg, no avg'ing.

    I'm no pro-developer, but wanted to use

    http://elitetrader.com/vb/showthread.php?s=&postid=1292190#post1292190

    as a starting point. Instead of setting Vol in column H set a start and end-vol and either lineair/jump development. For each leg. That won't cover all possible market scenarios but will improve existing stuff.

    Are we on the same bus?
     
    #20     Apr 9, 2007