one-offs

[destriero]

Tails maintain convexity from 0-inf. Small haircut ($3K debit). Crazy rho on long-dated but that is simply opportunity cost (moot) and this is a 2W combo. Allows you to trade short-gamma locally with no bear-risk.

Beat tail is conservative due to vol-corr and upside due to stickiness, but I typically model a touch of the the bear trough to x as vol-corr (- to index) will shift the curve higher.

It's an ideal portfolio prot. I arb VX against it (long switch say Dec/Jan for convergence gains as SPX/cash convexity exceeds leverage to VX). IOW, you can arb the long switch in VX against VIX SOQ.


View attachment 327339

Regarding VX... You can see the leverage to a 3% move downside which would result in perhaps a touch of 20 VX in the Dec. Arbing VX against isn't cheap but you can trade outright short VX against in lieu of the switch -> say, long Dec/Jan; Dec/Jan/Feb fly, etc. I can sell 10x outright in VX against this structure w/o risk.

 

[destriero]

Tails maintain convexity from 0-inf. Small haircut ($3K debit). Crazy rho on long-dated but that is simply opportunity cost (moot) and this is a 2W combo. Allows you to trade short-gamma locally with no bear-risk.

Bear tail is conservative due to vol-corr and upside due to stickiness, but I typically model a touch of the the bear trough to x as vol-corr (- to index) will shift the curve higher.

It's an ideal portfolio prot. I arb VX against it (long switch say Dec/Jan for convergence gains as SPX/cash convexity exceeds leverage to VX). IOW, you can arb the long switch in VX against VIX SOQ.


View attachment 327339

Further, the more straightforward trade is to lever-up long Delta1 in SN or SN vol as a sort of dirty dispersion. I always carry the bear-convexity hence my returns are massively -corr to mkt while mkt-performing or better, upside. Hence I only show PNL when outperforming.

 

[traider]

vol-arb. long vol-corr, Derman and tails. Zero vol-shift. Pure arb.

View attachment 327225

Why is the PNL for wings so much greater than body?

 

[tooriginal]

Still trying to understand this. But would you say this position is trading against an overly sticky strike market? And part of the modelling is assuming sticky delta will take over for a large enough move?

 

[destriero]

Still trying to understand this. But would you say this position is trading against an overly sticky strike market? And part of the modelling is assuming sticky delta will take over for a large enough move?

Sure, but at some point gamma (upside) goes to zero and it becomes Delta1. You have to model conservatively to assure that the stress on SD exceeds the loss to vol-corr upside.

 

[destriero]

 

[Adam777]

Wow this is gold ... and just when I had my pitchfork skew and flys all sorted with TWS data in excel. You've clearly explained how to find the max vertical skew between the 2 ET threads, but I didn't exactly understand how you're finding max horizontal skew

 

[destriero]

Wow this is gold ... and just when I had my pitchfork skew and flys all sorted with TWS data in excel. You've clearly explained how to find the max vertical skew between the 2 ET threads, but I didn't exactly understand how you're finding max horizontal skew

Through the complex fly/skew lock.

 

[destriero]

Yeah, The PF is obsolete.

 

[ironchef]

@destriero, what do I have to do, to read up in order to understand what you and the others are talking about here?

It is all Greek to me.