Option replication and exotics journal

Discussion in 'Journals' started by riskarb, Jul 14, 2005.

Thread Status:
Not open for further replies.
  1. In reference to this trade you did a while ago... the win/loss was .60 ($248600/$151400) . Which is a .6 : 1 reward to risk.. I thought u never strive for less than 1:1.

    Does it really matter? because it seems that the risk to reward is kinda irrelevant because the obviously the pricing of the exotic adjusts as the probability of being hit or missed on the touch increases or decreases...

    So lets assume that there is a 90% chance that the barrier will not be hit.. and you have 1:10 payoff.. so u have expectancy of 0.. which is equivalent to a 1:1 payoff with 50% chance of being hit. So basically the actual pricing of the trade balances out based off the probabilities.. how do u go about calculating the price level you want to set for the barrier?? Seems to me this has more to do with technicals and gut than actually being able to base it off some type of pricing model.


    --MIKE
     
    #531     Nov 19, 2005
  2. mahras2

    mahras2

    Hey B,

    I still get subscription to your thread on my email and just saw you surpass 1mio a while back. Congrats on hitting that milestone. Best of luck on your future trading.
     
    #532     Nov 19, 2005
  3. Right, 1:1 isn't cast in stone. I adjust barriers based upon a variety of factors. I solve for the 1:1 bets as a benchmark to price implied to stat-vols.

    Obviously I'd be less inclined to trade inside 1sigma[1:1] if I model increased stat-vols. All of these plays are essentially bets on forward statistical volatility priced in gamma.
     
    #533     Nov 19, 2005
  4. Riskarb,, it looks like u do a lot of synthetic straddles with otm not touch... do u ever trade the otm touches.. and how can u go about setting up a synthetic for a touch bet?

    Also what type of strategies are there for digital options.. the ones that bet either up or down over a certain price level at expiry.

    Sorry to be picking your brains with all these questions but I find this whole exotic fixed odds betting fascinating.

    --MIKE
     
    #534     Nov 20, 2005
  5. Yeah, I've traded otm touches as a replication strategy -- DAX touch//SPX no touch. I also have traded touch-outrights in FX when gamma was cheap. I don't like digitals... I'm far more comfotable taking gamma/vega bets. Digitals are a combo of a g/v/direction bet.
     
    #535     Nov 20, 2005
  6. Thanks Buddy -- my best to you and yours.
     
    #536     Nov 20, 2005

  7. Is there a way to hedge a touch trade from the onset.. how can it be made into a synthetic?
     
    #537     Nov 20, 2005
  8. It's the lay-off for the short synthetic. In terms of a put touch you'd simply buy spot and buy the touch. You can solve for the gain at the the strike and build the hedge accordingly. Obviously, as the mirror-image for the short syntheitc there can be significant convergence-losses.
     
    #538     Nov 20, 2005
  9. This does not make sense to me.. let me give u a real example i priced out at Betonmarkets again...

    Assuming gold is at 486. At 478 i can get a one touch expiring on Nov28.. ( 8 calender days)... for premium $1000 and payoff $2000 .. basically a 1:1. At the same time i buy spot gold as I buy the one touch.. with position size solving perfect hedge to barrier. That means I need gold to go up 16 points to get a 1:1.. ( i think what i did was crease a synthetic call?)

    Now if I price a seperate one touch for gold to hit 502 I get a reward to risk of 4:1. So its basically a 1:1 vs a 4:1. SO why would anyone want to do that type of trade?

    Looks to me the best types of trades are either synthetic straddles or just plain directional one touch bets... am i on the right track?
     
    #539     Nov 20, 2005
  10. It's difficult to avoid fractional-contracts due to the small sums, but you would solve for hedge loss to strike; in this case it's $800 per one contract to trade size in-line with your payout. So you've got a small loss at the strike($200) to maintain symmetry. Obviously we could scale the position to solve for a b/e at the barrier. The upside is windfall profits on a rally; the mirror-condition than in the short synthetic. Of course, the long synth in this example suffers from decay and convergence risk[stat-vol contraction/fat gamma risk] just as with a long vanilla straddle.

    It only looks promising if you believe the implied distro is underpriced with the barrier $8 on spot. If your 1sigma, 8-day forecast of stat vol is +/-$8.00 on spot, you're looking at a >65% prob. that your barrier will be hit.

    In any event, you're maintaining symmetry, or nearly so, with a single contract. It doesn't require a $16 move to be made whole; that's the requirement to earn the 1:1 payout.

    In answer to your last question... compare the position to its vanilla counterpart. Would you rather be long or short the atm straddle in Gold?
     
    #540     Nov 20, 2005
Thread Status:
Not open for further replies.