Same question would apply to IR products too.

Long-duration structuring has nothing special, in a sense. What's special is the availability of certain market parameters. So in the case of FX, you'll need the existence of very long dated interest rates. Are they tradeable, or just a number on a screen? How should you model volatility in a Black sense?Equity stuff: dividends? In the long dated world, they are not what they seem - the existence of credit issues / index members survivorship bias...

Thank you, Papercut. I was wondering because of the job adds.

Long dated FX (PRDC products) are a bit special.You get an unhedgeable one way expoure to FX smile, which you need experience to manage.example structure (issuer)Pay: Notional *Max(USD/JPY -110),0)*1%Recieve: Whatever?The floor on the coupon paid represents a vanilla FX option, the client has bought from issuer. All the FX options have a common strike, but the FX curve is downward sloping, so near dated options are far out of the money, while far dated options are at or through the money. Out-strikes are priced at a different implied vol to ATM options => FX smile.This can cause quite nasty vol squeezes that you don't get in other markets, and the modelling is also pretty poor, so intuitive understanding is important.Which IR products are you interested in?

Thank you. I guess one of the difficulties is to find those long dated options to hedge the smile risk, maturity bucket by maturity bucket, right?What do people sometimes refer to as cross risk? Is it the dVega/dS - so, again, smile risk?As far IR as products are concerned I was thinking about range accruals, TARN's, snowballs... Thanks again.

Thank you. I guess one of the difficulties is to find those long dated options to hedge the smile risk, maturity bucket by maturity bucket, right?-YesWhat do people sometimes refer to as cross risk? Is it the dVega/dS - so, again, smile risk?-This can mean lots of different things. Cross gamma is change in e.g. IR Delta for change in Spot FX (very big issue in LDFX)-Also can be what you mentionAs far IR as products are concerned I was thinking about range accruals, TARN's, snowballs... Thanks again. -Nothing that special about these products -General liquidity in the long end-Increased 2nd order aggressiveness of greeks in Tarns and Callable structures (long dated products are more leveraged/front loaded)

QuoteOriginally posted by: bsycheng...You get an unhedgeable one way expoure to FX smile, which you need experience to manage...*sigh*Why, exactly, is the FX smile unhedgeable?

Show me the price for a 30yr risk reversal on USDJPY...It's not unhedgable per se, but I would bet that all the dealers on the street have massive positions warehoused.

QuoteOriginally posted by: bsycheng Show me the price for a 30yr risk reversal on USDJPY...I'm not a market maker. However, I'm pretty sure I can price one using Bloomberg, MATLAB, a stapler, band-aid and some coffee. Ouch! I just got a paper cut.QuoteIt's not unhedgable per se...You don't say.Quote...but I would bet that all the dealers on the street have massive positions warehoused.They do? So why would this be hard to hedge? Or price?

It's not a complicated product, that's not the point.It's a thin market. There are no natural sellers of long-dated vega, especially for low strikes. Vol curve is very steep.30yr vol is 18%+ vs 1y Vol at 8%.It's not possible to efficiently hedge that far out on the vol surface and make money. This is what I mean by unhedgeable.... Theoretically it isn't, but practically it is.Could you explain what fits your definition of unhedgebale, if anything?

bsycheng,Re: cross risk, it also exists in single options, so say in PRDC's isn't possible to approach it in a similar way?Thanks again...