How liquid are VIX options?

[shooter]

Hate to add to the confusion but, if I'm not mistaken thinkorswim does not model VIX options correctly....the theoretical pricing of all VIX option maturities is based off front month VIX on thinkorswim, whereas in reality each maturity should be priced off the corresponding month.

i.e. June VIX options should be priced off a blend of the corresponding June/July VIX (weighted?) calculation, but TOS's theoretical pricing is based off of current March VIX.

Sorry if I'm mistaken.


Also, as far as vol calculations on TOS for calendar spreads and whatnot, you can click on the little screwdriver and adjust volatility for each month. i.e. for April VIX options you can adjust volatility up 8%, and May volatility up 6%. But if you click on 'plot lines' ---> +1 Vol Step, it will increase volatility the same across the board, which obviously doesn't happen in real pricing.

 

[TskTsk]

I think you're misunderstanding what sle is arguing; hedge ratios are referring to the number of futures of one maturity you'd have to be long to offset a position in a different maturity. VXX will lose its value in the long-term because of negative roll yield a typical contango environment (hence the constant reverse splits as the price falls), but that doesn't drive large price swings in the VXX; VXX's holdings consist of a futures position weighted between the front two contracts; if one goes up, the other one is almost certainly going to go up as well along with it.

And yes, if it's the paper I'm thinking of (put out about 3 years ago), it's a very helpful one.

I know what hedge ratios are. I'm simply arguing that calculating hedge-ratios is pointless, because they are essentially exposed to contango / backwardation. So if you start investing based on those hedge-ratios, you're simply speculating in contango / backwardation. And that phenomenom is essentially what drives the swings in the VXX, or to a large extent anyways. And indeed, in heavy backwardation, VXX tends to increase, whereast in heavy contango, VXX tends to decrease...in fact it's one of my primary signals I use to reduce my own short VXX exposure....

Also, if VXX reduces it's holdings of one contract, then it must neccessarily increase it's holdings of the other contract. The two are always inverse.

Are you saying I won't be able to rely on TOS's profit/loss graphs when I'm trading live with them? Here are the graphs of the 2 theoretical VIX positions:]

Indeed what I'm saying. You will notice on your TOS graph that the two options are modeled against the same underlyer. Which is not correct...

 

[darkshogun]

Hate to add to the confusion but, if I'm not mistaken thinkorswim does not model VIX options correctly....the theoretical pricing of all VIX option maturities is based off front month VIX on thinkorswim, whereas in reality each maturity should be priced off the corresponding month.

i.e. June VIX options should be priced off a blend of the corresponding May/June VIX 'blend' calculation, but TOS's pricing is based off of current March VIX.

Sorry if I'm mistaken.

It is confusing. I don't know how the different pricing models work. I like visuals. If my graph says I'm making a profit of about $420 if VIX goes from 14 to 20, and the actual profit/loss of the position in the TOS account statement confirms it, what could really be wrong about it? I am new to VIX so I apologize for my obtuseness.

 

[darkshogun]

I know what hedge ratios are. I'm simply arguing that calculating hedge-ratios is pointless, because they are essentially exposed to contango / backwardation. So if you start investing based on those hedge-ratios, you're simply speculating in contango / backwardation. And that phenomenom is essentially what drives the swings in the VXX, or to a large extent anyways. And indeed, in heavy backwardation, VXX tends to increase, whereast in heavy contango, VXX tends to decrease...in fact it's one of my primary signals I use to reduce my own short VXX exposure....



Indeed what I'm saying. You will notice on your TOS graph that the two options are modeled against the same underlyer. Which is not correct...

How would I correct this?

 

[TskTsk]

How would I correct this?

Hell if I know, I dont use TOS for modeling VIX. I dont even trust it for equities...before new years 2013 I remember talking to their support chat, as their greeks were way off on some equity I cant remember, when compared to IB. After some back and forth we concluded that TOS indeed calculated it wrong and IB was correct in their pricing...whats scary is the delta figure was off as well. After that I dont trust TOS. IB isnt perfect either but its much more professional.

However my tip if you want to model the VIX : Trust nobody. Instead get a good understanding of it, so you can model it in your brain. Beats any risk graphs..

 

[darkshogun]

Hell if I know, I dont use TOS for modeling VIX. I dont even trust it for equities...before new years 2013 I remember talking to their support chat, as their greeks were way off on some equity I cant remember, when compared to IB. After some back and forth we concluded that TOS indeed calculated it wrong and IB was correct in their pricing...whats scary is the delta figure was off as well. After that I dont trust TOS. IB isnt perfect either but its much more professional.

However my tip if you want to model the VIX : Trust nobody. Instead get a good understanding of it, so you can model it in your brain. Beats any risk graphs..

I'm sure they're all cagey in some way or another and we never get the entire truth. But I'm going to ask a lame semi-newb question here - If I'm using TOS as my broker and their graphs to plot my strategies, as long as I'm making a profit when I should be making a profit (when their version of VIX increases) and losing money when I should be losing money (when their version of VIX goes down), does it matter in the long run who has the 'correct' model, if there really isn't one?

 

[monkeyjoe]

I'm sure they're all cagey in some way or another and we never get the entire truth. But I'm going to ask a lame semi-newb question here - If I'm using TOS as my broker and their graphs to plot my strategies, as long as I'm making a profit when I should be making a profit (when their version of VIX increases) and losing money when I should be losing money (when their version of VIX goes down), does it matter in the long run who has the 'correct' model, if there really isn't one?

People are trying to tell you that it's highly unlikely that their incorrect model happens to perfectly predict profit and loss scenarios. In other words, "as long as I'm making profit when I should be making profit" is not likely to happen if you rely only on the models in TOS.

 

[TskTsk]

I'm sure they're all cagey in some way or another and we never get the entire truth. But I'm going to ask a lame semi-newb question here - If I'm using TOS as my broker and their graphs to plot my strategies, as long as I'm making a profit when I should be making a profit (when their version of VIX increases) and losing money when I should be losing money (when their version of VIX goes down), does it matter in the long run who has the 'correct' model, if there really isn't one?

It doesnt matter as long as you're aware of the limitations of the model. But as moneyjoe said above, it's highly unlikely that the model will transfer to real life in the way you think it will (i.e. "as long as I'm making profit")

Shortly speaking, TOS models the options directly against the underlying VIX index, which is what they will eventually settle against. If you plan on taking the options to expiry, and dont care about how they will act before expiry, then the TOS model can be used. Do not expect to use the "Day Step" feature or the "Vol Step" feature and get any accurate results.

 

[rallymode]

Backwardation based on inversion (that is, a rally in the front accompanied with a decline in the back) is very rare and usually happends in smaller moves.

Also, Aug options will gain on both delta (change in VIX) and vega (change in vol of vol), so I would not be so sure if they are bad hedges.

sle, i realize you are speaking in general terms but actually it's been quite common over the past few months(admittedly out of norm) for the curve to flatten due to near term vol being bid relative to the mid-back end. And that's on up and down days. Hedging 30-day vol with 6 month vol regardless of ratio has certainly been a crappy trade from what i am looking at. In fact i'd venture a guess that the inverse pair has been a solid winner

 

[sle]

sle, i realize you are speaking in general terms but actually it's been quite common over the past few months(admittedly out of norm) for the curve to flatten due to near term vol being bid relative to the mid-back end. And that's on up and down days. Hedging 30-day vol with 6 month vol regardless of ratio has certainly been a crappy trade from what i am looking at. In fact i'd venture a guess that the inverse pair has been a solid winner

I am talking general terms here, just a generic idea that you can play relative value within the curve fairly well and hedge out sources of variance. As for the 1 vs 6 (or about), actually both directions have worked for me past year. It's like gamma vs vega - in some cases you want to be long calendars, in some cases short calendars. Once you get into the options world, it gets even more interesting, as you get the whole Samuelson effect and the skew involved, which makes for a lot of opportunities.