Delta hedging on teenies

[newwurldmn]

Quote from Brighton:

I'm pretty sure 'teenie' is derived from the smallest bid/ask increment that was in place before stock market decimalization, i.e. one-sixteenth of a dollar or 6.25 cents.

In current options jargon, it sounds like a teenie is defined as a far OTM option, but how far? A Delta of 20 (.20) or less? A Delta of 10 (.10) or less? Whatever you want it to be as long as it's small?

Teenie means small premium. Generally 5 cents or 10 cents... way out of the money, probably worthless option.

 

[Brighton]

Thanks. Now I know the type of trader Atticus is disparaging when he refers to "the idiot selling teenies."

 

[sle]

Quote from newwurldmn:

Teenie means small premium. Generally 5 cents or 10 cents... way out of the money, probably worthless option.

Sell a tiny, drive a Lamborghini!(c)

In reality, you most probably do not want to buy these (waste of money) and most probably do not want to sell them (unless it's an airport trade). Especially the ones in index, hedgies have a tendency of selling really short-dated teenies and if you win 'em you are almost always best of leaving them untied or delta-hedging them at much lower vol.

 

[cdcaveman]

Quote from sle:

Sell a tiny, drive a Lamborghini!(c)

In reality, you most probably do not want to buy these (waste of money) and most probably do not want to sell them (unless it's an airport trade). Especially the ones in index, hedgies have a tendency of selling really short-dated teenies and if you win 'em you are almost always best of leaving them untied or delta-hedging them at much lower vol.

whats an "airport trade" and your saying.. if your gonna expose yourself.. expose yourself.. hard to hedge them.. if you do slow play them/under hedge.. assuming a lower vol number on them and hedging to that number would be slow playing them.. meaning under hedging

 

[TigerBalm]

Just to make the discussion more interesting, I'll play devil's advocate and take the opposing point of view :

Maybe it's the case that in the long run, the explosion in IV on down moves, which happens 5% of the time, will compensate for the loss of premium and commission the other 95% of the time?

That's because it's not just IV, there's also the sensitivity of the IV to a change in price and to the level of IV itself, aka vanna and vomma. vomma is to vega what gamma is to delta, a 2nd order effect.

Instead of theory, see this article for a concrete example: http://www.optionworkshop.com/read/vanna-and-put-explosion
Here, an increase in IV of 17% resulted in an increase in the Put price by 163% ! As the author writes, "the point I want to make is that when tension enters the market, the OTM puts literally explode in value. Very simply put, when IV rises, the effects of Vanna and Vomma blow up the OTM put’s premiums, deltas and vegas in a way that looking at a simple risk graph cannot predict."

Also, the way you put it, "a 50% change in IV results only in a smaller change of the Put value" makes it sound as if a 50% change in IV is a big deal. But here again, we are talking about a 2nd order term, the volatility of volatility, for which 50% is not a big deal. See a chart of VVIX against the VIX, for example.

http://www.barchart.com/chart.php?sym=$VVIX&t=BAR&size=M&v=0&g=1&p=D&d=X&qb=1&style=technical&template=

 

[cdcaveman]

Quote from TigerBalm:

Just to make the discussion more interesting, I'll play devil's advocate and take the opposing point of view :

Maybe it's the case that in the long run, the explosion in IV on down moves, which happens 5% of the time, will compensate for the loss of premium and commission the other 95% of the time?

That's because it's not just IV, there's also the sensitivity of the IV to a change in price and to the level of IV itself, aka vanna and vomma. vomma is to vega what gamma is to delta, a 2nd order effect.

Instead of theory, see this article for a concrete example: http://www.optionworkshop.com/read/vanna-and-put-explosion
Here, an increase in IV of 17% resulted in an increase in the Put price by 163% ! As the author writes, "the point I want to make is that when tension enters the market, the OTM puts literally explode in value. Very simply put, when IV rises, the effects of Vanna and Vomma blow up the OTM put’s premiums, deltas and vegas in a way that looking at a simple risk graph cannot predict."

Also, the way you put it, "a 50% change in IV results only in a smaller change of the Put value" makes it sound as if a 50% change in IV is a big deal. But here again, we are talking about a 2nd order term, the volatility of volatility, for which 50% is not a big deal. See a chart of VVIX against the VIX, for example.

http://www.barchart.com/chart.php?sym=$VVIX&t=BAR&size=M&v=0&g=1&p=D&d=X&qb=1&style=technical&template=

your barchart link is bad....

 

[cdcaveman]

Quote from TigerBalm:

Just to make the discussion more interesting, I'll play devil's advocate and take the opposing point of view :

Maybe it's the case that in the long run, the explosion in IV on down moves, which happens 5% of the time, will compensate for the loss of premium and commission the other 95% of the time?

That's because it's not just IV, there's also the sensitivity of the IV to a change in price and to the level of IV itself, aka vanna and vomma. vomma is to vega what gamma is to delta, a 2nd order effect.

Instead of theory, see this article for a concrete example: http://www.optionworkshop.com/read/vanna-and-put-explosion
Here, an increase in IV of 17% resulted in an increase in the Put price by 163% ! As the author writes, "the point I want to make is that when tension enters the market, the OTM puts literally explode in value. Very simply put, when IV rises, the effects of Vanna and Vomma blow up the OTM put’s premiums, deltas and vegas in a way that looking at a simple risk graph cannot predict."

Also, the way you put it, "a 50% change in IV results only in a smaller change of the Put value" makes it sound as if a 50% change in IV is a big deal. But here again, we are talking about a 2nd order term, the volatility of volatility, for which 50% is not a big deal. See a chart of VVIX against the VIX, for example.

http://www.barchart.com/chart.php?sym=$VVIX&t=BAR&size=M&v=0&g=1&p=D&d=X&qb=1&style=technical&template=

how do you like live vol pro...??? you should add something there about the relatiive pricing in term structure.. would make for a good article.. term structure risk good subject if your into higher order subjects.. nice site..

 

[TigerBalm]

Quote from cdcaveman:

how do you like live vol pro...??? you should add something there about the relatiive pricing in term structure.. would make for a good article.. term structure risk good subject if your into higher order subjects.. nice site..

I'm not the author of that article (and I'm not shilling for them)...I just found it while trying to learn about vanna, vomma.

I DO have an idea regarding teenies (that may just help us buy our future Lamorghinis! ) that I'll post here later (outside right now).

 

[kapw7]

Quote from TigerBalm:

I'm not the author of that article (and I'm not shilling for them)...I just found it while trying to learn about vanna, vomma.

I DO have an idea regarding teenies (that may just help us buy our future Lamorghinis! ) that I'll post here later (outside right now).

Looking forward to that!
I have also been looking at what can be generally put under the category skewness-kurtosis trades and I have come up with a lot more questions than answers

How do you define if an OTM is cheap or expensive? How do you estimate the historical vol of vol and vol spot correlation (eg a simple regression or very advanced filtering?) How do you estimate the implied vol of vol and vol correlation? Do you use some advanced model (Heston etc) or do you use some sort of model free estimate of implied skewness/kurtosis?

How do you even measure your delta?

In general the risk premium lies in selling OTM positions which can easily lead to huge losses especially when I cannot even calculate my delta etc.

One interesting idea I found is in this paper:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1868605

Seems too good to be true (from the prespective that no quant knowledge needed) and as sle said before, you cannot trust academic papers on trading and this one has not even been published (says working paper - no other details)

 

[newwurldmn]

Quote from kapw7:

Looking forward to that!
I have also been looking at what can be generally put under the category skewness-kurtosis trades and I have come up with a lot more questions than answers

How do you define if an OTM is cheap or expensive? How do you estimate the historical vol of vol and vol spot correlation (eg a simple regression or very advanced filtering?) How do you estimate the implied vol of vol and vol correlation? Do you use some advanced model (Heston etc) or do you use some sort of model free estimate of implied skewness/kurtosis?

How do you even measure your delta?

In general the risk premium lies in selling OTM positions which can easily lead to huge losses especially when I cannot even calculate my delta etc.

One interesting idea I found is in this paper:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1868605

Seems too good to be true (from the prespective that no quant knowledge needed) and as sle said before, you cannot trust academic papers on trading and this one has not even been published (says working paper - no other details)

Yes. Skew is overpriced. But it's hard to monetize. You have to run levered books to make any real money and then you have dgamma/dspot and dvol/dspot issues.