The problem with short gamma

[nitro]

Quote from CPTrader:

Variance swaps are a great vehicle it seems for pure volatility trades.

How can one get access to these instruments? Hwo liquid are they?

What houses are the major players for these OTC instruments.

Why doesn' the CME or EUREX create exchnage traded versions of variance swaps?

Thanks

VSs are an institutional product. Even when we were trading our measly $10M, we didn't have access to these things. I think these are strictly interbank agreements, AFAIK.

In another thread,

http://www.elitetrader.com/vb/showthread.php?s=&threadid=191669&highlight=suggestion+to

I proposed to the CBOE (and later to the CME because you will see in that thread that a futures on that suggestion is probably the only way) that they implement a product close to the spirit of a VS.

If you are managing >$100M, I suspect you can call GS or Citadel etc, and get a quote on a VS on a given asset in denominations of $250,000 or more (probably way more).

The point is that there is no product in existance, with the exception of possibly VIX, that allows you to trade volatility purely without having to worry about continuous delta hedging or skew risk. What we want is a sort of VIX for each individual equity, but VIX is too messy because it mixes months (there should be seperate "VIX" per expiry month withoug mixing options from different months), although I would even be willing to live with that.

VSs are great because they in one price make an entire continous option chain liquid, and they are far more immune to the underlying than things like strangles and straddles

 

[CPTrader]

Thanks nitro.

Quote from nitro:

VSs are an institutional product. Even when we were trading our measly $10M, we didn't have access to these things. I think these are strictly interbank agreements, AFAIK.

In another thread,

http://www.elitetrader.com/vb/showthread.php?s=&threadid=191669&highlight=suggestion+to

I proposed to the CBOE (and later to the CME because you will see in that thread that a futures on that suggestion is probably the only way) that they implement a product close to the spirit of a VS.

If you are managing >$100M, I suspect you can call GS or Citadel etc, and get a quote on a VS on a given asset in denominations of $250,000 or more (probably way more).

The point is that there is no product in existance, with the exception of possibly VIX, that allows you to trade volatility purely without having to worry about continuous delta hedging or skew risk. What we want is a sort of VIX for each individual equity, but VIX is too messy because it mixes months (there should be seperate "VIX" per expiry month withoug mixing options from different months), although I would even be willing to live with that.

VSs are great because they in one price make an entire option chain liquid, and they are far more immune to the underlying than things like strangles and straddles

 

[MasterAtWork]

Hi nitro,

As a trader who used to price variance swaps, I don't get it.

Quote from nitro:

"...A perfect hedge with a constant aggregate gamma for all underlying levels would take infinitely many options struck along a continuum between 0 and infinity and weighted inversely proportional to the squared strike. This is etablished rigorously in Section 3.2. Note that this is a strong result, as the static hedge is both space (underlying level) and time independent."

http://math.uchicago.edu/~sbossu/VarSwaps.pdf

"It took some serious withdrawal symptoms and final acceptance through a painful process, but once flushed of the demon, it opened for some interesting ideas. I am now in the hunt for a high frequency underlying (stocks) system. I am pretty sure that I have a decent options system if I didn't have to worry about delta neutrality, but that is not my reality since I don't have access to variance or vola swaps. Starting from a simple premise as to how Rentec trades, almost all of the repercussions of an underlying HFTS must be highly constrained, so at least the search space is not infinite, or at least countable, aleph-null."

Since you quote a variance swap, you need a lot of vanilla strikes to construct a portofolio that would replicate the variance payoff. You need to delta hedge your vanilla portofolio to keep a pure volatility exposure. As a matter of fact, your delta hedging strategy costs need to be included in the variance price you'd quote. That's called a fair variance price that determines the strike of your variance swap. If you don't you would lose money for sure.

Hence my point is the following one : if you 'd get a var swap, you 'd pay for delta hedging strategy in it. So delta hedging cost is not an issue here.

 

[Maverick74]

Quote from nitro:

VSs are an institutional product. Even when we were trading our measly $10M, we didn't have access to these things. I think these are strictly interbank agreements, AFAIK.

Nitro, there are variance swap futures that are exchange traded. There are 3 and 12 month swaps on the S&P 500.

 

[nitro]

Quote from Maverick74:

Nitro, there are variance swap futures that are exchange traded. There are 3 and 12 month swaps on the S&P 500.

I need single equity vola swaps, not swaps on the entire SP 500.

 

[nitro]

I don't understand what you don't get. So the price of delta hedging is already included in the price of a VS. Great, I don't have to worry about staying delta neutral then, right? (other than that I paid for it.) And if I sell another month VS against one I am long, these two months delta hedging costs embedded in the VSs prices should more or less even out, and therefore I am trading just vola.

Perfect!

Just so that we are clear, I don't want to replicate a VS with vanilla options. I just want a variance swap.

Quote from MasterAtWork:

Hi nitro,

As a trader who used to price variance swaps, I don't get it.



Since you quote a variance swap, you need a lot of vanilla strikes to construct a portofolio that would replicate the variance payoff. You need to delta hedge your vanilla portofolio to keep a pure volatility exposure. As a matter of fact, your delta hedging strategy costs need to be included in the variance price you'd quote. That's called a fair variance price that determines the strike of your variance swap. If you don't you would lose money for sure.

Hence my point is the following one : if you 'd get a var swap, you 'd pay for delta hedging strategy in it. So delta hedging cost is not an issue here.

 

[Maverick74]

Quote from nitro:

I need single equity vola swaps, not swaps on the entire SP 500.

I know, I was making the comment in reference to you saying they are only an institutional product. Also, making single equity vola swaps available is not your problem. They could all be listed tomorrow on the CBOE and that would not help you. The spreads would be 100 yards wide.

Look, these are order flow products, not market taking products. If Goldman wasn't earning the spread on these over the counter, I doubt they would make any money on them.

 

[nitro]

Quote from Maverick74:

I know, I was making the comment in reference to you saying they are only an institutional product.

Right, I was aware of the VS on the SP500 being exchange traded. I may be able to make use of them in some future project, unless they have the problem you state below, i.e., that they have enormous B/A spreads.

Also, making single equity vola swaps available is not your problem. They could all be listed tomorrow on the CBOE and that would not help you. The spreads would be 100 yards wide.

Well, it is my problem, but your point is that the solution that I claim as my savior would not help because it would introduce another problem, they would be essentially not trade-able do to enormous spreads.

Look, these are order flow products, not market taking products. If Goldman wasn't earning the spread on these over the counter, I doubt they would make any money on them.

Ya well, that would not surprise me that there is no interest in making markets on them, for a variety of reasons which I don't want to get in to (I don't want to get into a political war with the powers that be). But I don't know if I agree with your premise anyway. People made the same arguments when options first came out at the CBOE ("who would sell insurance on equities? You have to be crazy!"), and now look, we have penny wide spreads in a large portion of the equity-option universe and the penny-pilot program is growing every couple of months.

 

[heech]

Quote from nitro:

Ok, let's simmulate it. Come monday, I will give you a calendar straddle making you gamma short (short front month long next back month). You hedge the position delta once a day at a predetermined time, for a week and see how you do. You don't get to adjust the straddles strikes.

You up for the challenge?

And what would this prove, exactly? What if I was profitable a week, or what if I was negative after a week? The former indicates I was right, and the latter indicates you were?

That's not the case, of course. It will be profitable if realized volatility was less than IV, and a loser if realized volatility was higher than IV.

FYI: I'm trading $2.5mm, and this property of volatility is a key aspect of my strategy.

 

[nitro]

Quote from heech:

And what would this prove, exactly? What if I was profitable a week, or what if I was negative after a week? The former indicates I was right, and the latter indicates you were?

That's not the case, of course. It will be profitable if realized volatility was less than IV, and a loser if realized volatility was higher than IV.

Well, I am not saying it would prove something or nothing. You claim you would be "...profitable if realized volatility was less than IV, and a loser if realized volatility was higher than IV." Ok, let's see if that is all it requires, by your own hand. We will have a measure of IV and realized at each hedge. We can wave our hands in the air all day long and mentally masturbate back and forth on ET all day long, but nothing beats a real-world test, even if that test is humble. I have done this thousands of times on real positions and I am telling you that is not all that it requires, but we keep going back and forth in circles like insane people.

My claim is that the PnL will be dominated by how well you delta hedge, not just by IV vs realized vola, but because of every argument I have made in this thread and in particular, Peter Carr's insight. As soon as you introduce a non-trivial position that includes both puts and calls, long and short, multi-month options, you are extremely sensitized to the non-linear path of the underlying.