I traded a $20,000 miss into the long weekend earlier tonight. I model my exotics using multiple sources of real-time vols from volbroker and my ibank. Superderivatives is the standard for pricing in the industry, and I use the service. The model output payout on SD was $19,388//$34,610. I received $20,000//$34,610. Yes, the model expressed a discount to Oanda's market, but only a 3% edge loss. I don't expect to convert you. I have my issues as well. If you want to witness egregious markets, visit and price exotics on the following dealer-sites: http://www.fxcm.com/double-one-touch.jsp www.betonmarkets.com www.betsfortraders.com
How on earth do you find NT plays with R/R like that?? Awesome. May I ask what your typical duration is? I remember Natenberg going on an on about mispricings in the market. I had assumed that with todays technology, these didn't exist unless you were part machine (or all machine). Do you find that to be true? I'm still failing to truly grasp why this would be a better alternative to just trading in the underlying. Please go easy on me here, but here is why I say that. 1) Not only must one get direction right, but also must get time right 2) winners can't run 3) no such thing as partial profits, it either pays out fully or it lose fully 4) seems like you need a high IQ Best regards, MK
It was a one week bull trade on GBP, got lucky. Average duration is 4-days. There are few mispricings in a dealer market. It's always a bet; either on gamma[distribution] or vega. The AUD-trade was better than 1:1 at 30-deltas. I would've needed to sell 12.5mm AUD to replicate the return. Why sell a call over selling spot? I rarely trade touch markets w/o an underlying spot hedge or a replication trade involving a spot/touch position in a correlated pair. I am not debating suitability or methodology, only pricing.
Hi atticus, You certainly know your beans in this area! A few questions if I may, I'm finding this interesting to hear from a pro about these instruments. This may seem like a total newbie question, how important would you say it is know the gamma and vega of the priced box? I mean, if it is for directional spec, does it matter? how much does it matter? Wayyyy over my head there. Remember my point #4 above? Sounds complicated dude. Last night I was thinking about these instruments over vanilla options and then it hit me. The greeks are different. Maybe different is a bad choice of words -- they are fixed once the trade is put on. When Oanda first brought these box options out, I tried them. From memory, I thought they were all or nothing sort of trades (there's been a few wild parties since then so the memory may be faulty). I don't remember receiving any partial payouts, or any theta gain, or the delta changing as the underlying moved etc... It either played out as the box was setup and the maximum payout was received, or you lost the premium. If my memory is correct, then this makes the timing even more crucial with these instruments over vanilla options. I guess you can replicate theta decay at regular intervals by staggering several boxes and their expirations. I need to think about it a little more today and do some more reading. Thanks for the informative posts. Best regards, MK
Quote from MidKnight: Hi atticus, You certainly know your beans in this area! A few questions if I may, I'm finding this interesting to hear from a pro about these instruments. This may seem like a total newbie question, how important would you say it is know the gamma and vega of the priced box? I mean, if it is for directional spec, does it matter? how much does it matter? You're right, it's not terribly important as any 6th grader could solve for the barrier-hedge algebraically. Dynamic hedging requires some understanding of the curvature of gamma. Quote from atticus: I rarely trade touch markets w/o an underlying spot hedge or a replication trade involving a spot/touch position in a correlated pair. I am not debating suitability or methodology, only pricing. Wayyyy over my head there. Remember my point #4 above? Sounds complicated dude. Simply that I buy volatility box/swap via replication -- buying "n" vol on one pair and "n+1" vol on another, long correlation; i.e, selling vol on GBP and buying EUR vol. Last night I was thinking about these instruments over vanilla options and then it hit me. The greeks are different. Maybe different is a bad choice of words -- they are fixed once the trade is put on. When Oanda first brought these box options out, I tried them. From memory, I thought they were all or nothing sort of trades (there's been a few wild parties since then so the memory may be faulty). I don't remember receiving any partial payouts, or any theta gain, or the delta changing as the underlying moved etc... It either played out as the box was setup and the maximum payout was received, or you lost the premium. If my memory is correct, then this makes the timing even more crucial with these instruments over vanilla options. The formal term for a cash touch is an "cash or nothing" binary option. I guess you can replicate theta decay at regular intervals by staggering several boxes and their expirations. I need to think about it a little more today and do some more reading. Sure, you could choose linear decay, but it would increase your gamma risk. The gamma/theta ratio would increase[bad]. Thanks for the informative posts. YW Best regards, MK
I'll add that theta is a different animal when the option is a "money-line" bet. The theta follows the payout regime. IOW, it's a 'lumpy" theta and not discrete like analog options. It's modeled as discrete, but the gamma-sensitivity overwhelms.
I bought a large-exposure CHF/JPY no-touch today in the OTC market which expires on June 1, NY-cut. 99.10 strike; 70/100 payout; long CHFJPY spot as hedge. The barrier is outside of 2-sigmas for that duration and the spot hedge is weak. I chose to short d/g in CHF/JPY as my model predicts a contraction in stat vol on the pair. I foresee some spillover from today's selloff in equity markets, hence the long yen/short swiss price exposure. The closest analog in the vanilla World would be a short synthetic straddle or overwrite.
Thanks for the reply atticus - some of it makes sense! I'll play along and post some ideas I have next week with regards to NT plays on the 13 pairs in my world. These will be just straight NT plays for the duration of the week. At the very least, I might learn something from this exercise. Kind regards, MK