Lets talk variance

[.sigma]

I'm surprised ET has a "Fixed Income" forum yet no Volatility forum.

Anyways lets get a discussion going on variance.

I am a complete neophyte when it comes to variance, I just know the basic sq. of vol stuff, but I'm trying to understand this stuff better.

I've read @sle and I know TBS knows quite a bit about variance as well and others here.

Has understanding what variance is, the functionality of variance swaps, how to price them etc.. helped you in understanding the concept of vol better? I'm assuming yes. This is why I'm learning more about var as it compares to vol.

I'm curious what you variance guys think of the BSM, because BSM uses vol as a measure of financial deviation, not var... yet I'm reading Bennett and seeing that the industry created the modern $VIX formula to move away from ATM vol (VXO), towards variance-based calculation.

Do vol forecasting or value models like GARCH use variance as the st.dev?

 

[TheBigShort]

I know TBS knows quite a bit about variance as well

.simga I am glad you find my posts insightful. However, I am still very new in the vol space. There are a few people on this board who have 20 years of experience. I have 1/10th of that. So please read my stuff with a brick of salt.

In regards to your post, I am not quite sure where you are trying to go with this. The math can get pretty heavy when you get into pricing OTC derivatives. In a previous thread, you were hesitant to learn coding or math. As a retail trader, you probably won't find an edge in being able to price a variance swap (especially since we cant trade them). There is money to be made in modeling OTM options but I will leave that for someone else to talk about.

It is easier to work with variance which is why when you talk to a PM/Trader in the vol space they use the word variance quite a bit. A vol PM might be pricing many OTC derivates and will therefore be working with variance rather than converting to vol all the time. But variance is just sigma^2, while vol is sigma*sqrt(T). Vol is usually annualized which is why we multiply by sqrt(T).

 

[TommyR]

realised vol must be sampled daily for true volatility, with exponential weightings obviously-expert. market moves come in discrete pulses, they are almost always not moving, then a pulse comes, exerts some force and it moves. at what frequency do pulses of this power arrive?

 

[tommcginnis]

Recognizing BSM' volatility as an agglomeration of individual strikes' IV, I have always been more interested in the individual strikes, or at least the behavior of the local neighborhood. This means that I care less about skew, and allows (perhaps) more attention to unexpected differences between expirations. Those differences point out a lot of event-driven pricing, which -- with a little mapping -- was pretty easy to take advantage. ("Yay!")

 

[Same Lazy Element]

There are two primary reasons to look at variance. First is to understand the flows that variance swaps create. Second is to trade related products.

 

[.sigma]

In regards to your post, I am not quite sure where you are trying to go with this.

1. I don't think experience has much to do with it. Your posts are just as insightful if not more than some of the vets of this site, I'm glad we have you in our era.

2. I'm trying to grasp and understand more advanced concepts of derivatives because I believe knowledge is power. Of course one must also "know the ledge" when it comes to knowledge. Yes I am hesitant to learn coding but determined to get the basics down at least. Regardless of coding I believe understanding variance, and things like variance swaps can help me assess vanilla optionality better.

For example I've seen dest compare bfly's as some type of knock-out barrier.. not sure if that's the exact terminology but I definitely think once one has a good grasp of vanilla options it wouldn't hurt to study up on exotics (and variance).

I also read in Bennetts book that modeling variance can help you determine if vol is cheap/expensive. Not sure how, but that's what figuring out shit does lol.

I'm not trying to find an edge using variance per se.. like ive said its more of an overall better understanding of the derivative world. And yes variance-swaps and exotics are OTC but liquidity has poured into these products and I believe eventually exotics and the like will be available for us retailers, but thats another thread for another day.

 

[.sigma]

realised vol must be sampled daily for true volatility, with exponential weightings obviously-expert. market moves come in discrete pulses, they are almost always not moving, then a pulse comes, exerts some force and it moves. at what frequency do pulses of this power arrive?

What do you mean by "realized vol must be sampled daily for true vol"? Do you mean converting volatility into variance?

I agree with your point on the markets overall non-movement. And that's interesting about the incoming pulse, this power arrives from some source and I believe its more than one. The key is to spot it before hindsight bias creeps up on you.

 

[.sigma]

Recognizing BSM' volatility as an agglomeration of individual strikes' IV, I have always been more interested in the individual strikes, or at least the behavior of the local neighborhood.

Not sure what you mean by this?

You are saying you pay more attention to individual strike IV% instead of the overall IV% of the underlying?

 

[.sigma]

There are two primary reasons to look at variance. First is to understand the flows that variance swaps create. Second is to trade related products.

Can you elaborate more on what you meant by "flows that var-swaps create"?
What flow? Money flow? Debt-instrumentation?

What are related products? I'm thinking it has to do with the VIX products and associated vol etf/etn's?

 

[tommcginnis]

Not sure what you mean by this?

You are saying you pay more attention to individual strike IV% instead of the overall IV% of the underlying?

Absolutely. I don't buy gasoline by the average price in my state -- I buy it from the station in front of me.