DP's exotics journal

[drownpruf]

I don't know the first thing about exotic options trading. But wouldn't you say these kind of equity swings(-12% on the first 2 trades, and now risking ~25% of nav in the next trade) are a wee bit extreme?

I took $10K to $300K in 18 months in one of these in 2003-5. I have BOM threads under "riskarb" as well. I took Trinitas Capital out of the business back in 2008 with SN exotics (DNTs on WMT, JNJ, etc). Yeah, DNTs are the meth of the vola-World. My "mentor" in exotics took less than $20K to $100MM organically in under 15y.

Oanda was offering pricing on touches as digital pricing. That was a pure-arb (netted elsewhere) where they would offer this touch at 17/30. They no longer offer them as a result. I am stating that the distribution is understated.

I know it's aggressive, but it's not the same as going long an ES overnight with $15K. The guy(s) holding the account(s) are worth 8-figs.

 

[drownpruf]

how do you measure volatility (mathematically) on FX as opposed to options?

any formulas?

ha, I use bog-standard vanna-volga on bloomberg pro fixed pages. GK pricing on vanillas (and VV) simply uses the fwd net of carry. IOW, watch the carry if you're modeling longer-term stuff on AUD, JPY, etc. You can stress (single touches) these statically by using touch prob or exp prob*2 under BSM/GK.

Most traders are better off trading the unimodal positions (touch vs, DNT). It's a lot less complex to buy/sell a touch and convert to a synthetic straddle (to use vanilla analogy). Again, I know guys making millions based on nothing more than buying the 70/100 no touch upside and risking half of the debit in long spot. Or simply treat the touch/NT as a short put/long call.

usd/xyz 1.10 NT bought at 70/100. Solve for spot hedge that results in a credit of 35 (half initial debit) at a barrier touch. I have mentioned this a number of times. There is no direct analogy to a vanilla, but an asymmetric (strike) short backspread/weak fly is a similar payoff in the vanilla World.

 

[drownpruf]

Disregard my last question. More specifically, if I am looking at it right they are listing a $100 payout for rise/fall of the SP500 tomorrow close. It costs approx. 57 if it closes above or 49 if it closes below. How did they come up with the costs to make them so different (assuming 50/50 chance)? Is it based on which side most of the people are betting on?

There are a ton of bucket shops trading atm digitals. It's analogous to opening a site betting on a coin-flip. An ATM digital fairval will be 50/100 tied to the forward rate. It's = spot if it's an intraday trade.

The edge is tremendous. Say you've got a couple thousand accounts betting on EURUSD. A market of 45/55 for the ATM digital. Sure, you're not netting directly client for client, but each trade produces a 10% edge to the house.

It's a coin flip in which you hit 55/100 when you win ($55 pays $100).

 

[voltrader1]

ha, I use bog-standard vanna-volga on bloomberg pro fixed pages. GK pricing on vanillas (and VV) simply uses the fwd net of carry. IOW, watch the carry if you're modeling longer-term stuff on AUD, JPY, etc. You can stress (single touches) these statically by using touch prob or exp prob*2 under BSM/GK.

Most traders are better off trading the unimodal positions (touch vs, DNT). It's a lot less complex to buy/sell a touch and convert to a synthetic straddle (to use vanilla analogy). Again, I know guys making millions based on nothing more than buying the 70/100 no touch upside and risking half of the debit in long spot. Or simply treat the touch/NT as a short put/long call.

usd/xyz 1.10 NT bought at 70/100. Solve for spot hedge that results in a credit of 35 (half initial debit) at a barrier touch. I have mentioned this a number of times. There is no direct analogy to a vanilla, but an asymmetric (strike) short backspread/weak fly is a similar payoff in the vanilla World.

interesting.. some of its past my head at this point but im gonna look into it.

whats your mathematical background? any higher education?

 

[drownpruf]

DP,

Thanks for the exotic options thread, never understood how it's structured.

How do you determine when to hedge the position, and why do you only hedge after a 70 pips drop, and only for 1/2 your original investment? Do you hedge on spot at initiation of position? Thanks.

Obviously you have to choose some point on the curve (gamma) to hedge the thing. I normally let these go (unhedged), but had a significant debit at risk. The $gamma/VV got too big to ignore and I had a "stop" at 3825. I run matrices in r/t which runs touch pricing. I wasn't willing to pay more than I did for that touch hedge. I don't solve for the moment-risk; I solve for the debit needed to effect the hedge. Specifically, it scrapes the pricing from BOM and they do offer a matrix, if I recall.

I was willing to risk half the DNT debit on the hedge. I had some room to recover (20 pips between touch and DNT barrier). It's moot as we're going to take out the DNT. There is a recurring theme of 1/2 hedges in spot and vola on this and my previous threads.

Yeah, the numbers look big in terms of %. The DNT hedge was well-timed and robust. I should've hedged the full DNT debit, but was ok with the risk. Sure, I should've gone with a 30K touch-hedge.

 

[drownpruf]

interesting.. some of its past my head at this point but im gonna look into it.

whats your mathematical background? any higher education?

Undergrad in physics and did some lin-algebra and topology. I am solid with maths but I always assume I know less than the guys writing the stuff. Options are a tool.

 

[drownpruf]

I don't know the first thing about exotic options trading. But wouldn't you say these kind of equity swings(-12% on the first 2 trades, and now risking ~25% of nav in the next trade) are a wee bit extreme?

I've got a 14-day touch 1.381. I assign the prob to exceed what I paid. I will begin to short spot on rallies and update the position when I do so.

 

[drownpruf]

Touch options (singles) are a pure play on touch probability -- hence the reduction of the figure to x/100.

 

[voltrader1]

Undergrad in physics and did some lin-algebra and topology. I am solid with maths but I always assume I know less than the guys writing the stuff. Options are a tool.

not surprising... seems like a lot of physicists get into derivative markets and do very well..

to a young person coming up would you recommend a math or physics degree if they wanted to get into derivatives and understand them better? and also for job prospects in the industry.

 

[drownpruf]

not surprising... seems like a lot of physicists get into derivative markets and do very well..

Cosmology. Chandrasekhar limit, dark matter (quark stars, hadronic, IMO); bubble-models. I did/do have an interest in astronomy, less so the maths. I know virtually nothing about QM. I have had Dobs and Schmidt-Cassegrain scopes since my teens.