Which IV to use when modeling spreads?

[schreibdave]

Happy Holidays everybody

I just stumbled across this forum so this is my first post here.

I am working on mastering credit spreads and I have a question about how I should be modeling implied volatility.

I use a program called 'Edutrader' which will create a risk graph of the proposed position and the variables are price, time and volatility. I am not sure what input to assign to IV. Say I model a call spread, and each leg has a different IV associated with it. Should I average the 2 IVs in order to establish the initial "position IV?" More importantly, If I want to model what my P&L will be at various points in the future, what IVs shoud I input? A decline in the stock is usually associated with higher IVs, and an increase in the stock usually associated with a decline in IV. Should that be my basis for making assumptions? Or should I look at what the historic volatility has been and use that as my basis for making projections? If so, historic volatility of what duration? - 30 days? 60 days?

Maybe I am overthinking this, but I would appreciate any input you folks can provide. Thanks

 

[slik]

unless its a very tight spread, the implied vol of the strikes can be different due to the variation in the vol surface. Ideally, you should be able to enter different vols for each strike or create a vol surface that accuarately maps the market using parameters like slope, call/put curves, cut offs, vcr, scr and slide rates. VCR is the change in vol on a spot moving e.g. IV at x is 10%, IV at .75x is 10%+whatever you want to put in- that kinda answers part of your question.

I would be very careful in trying to predict vol though- there are vol predictors and what not that uses historical data, but this is no ordinary market at the moment. If you must use historical vol, tie in the data observation periods with your expected holding time of the option or expiration time with the observation period (eg 30 day option with 30 day moving averages of vol and associated weightings for macd etc etc)

 

[schreibdave]

Thank you Slik. Unfortunately I am not familiar with a lot of the terminology/methodology that you describe, but maybe I can simplify my question... Lets say that if I have a bull put spread with an average IV of the two legs of 50. How should I model future IV? There are no earnings forecast and I have no knowledge of any market moving announcements coming up. As best I can tell, my options are 1) Assume it moves to some measure of historic volatility - so for a 30 day option I might use the 30 day HV. 2) I could look at what the current IV is a few strikes closer to ATM and a few strikes further from ATM and use those measures to estimate what the IV might be if the underlying moves in either direction. Given the software platform that I have, I think those are the options I have given the data that I have access to. What do you think?

 

[slik]

hey, it really depends on what kind of play you are interested in- if you are delta hedging (delta neutral) the implied vol and the other greeks would be managed in a different way than if you just for example putting on a directional play. ( where you just want to pay a premium for a spread say)

Always remember that implied vol is the premium that people are willing to buy/sell options at and these will differ significantly for at the money options to say puts in equity indices; puts trade at much higher vol due to supply and demand- we are in a bear market and vol explodes when say the dow/ftse/dax touches year/multi year lows. That is all a vol surface is, different implied vols for different strikes (calls, atm, puts)- quadratic and tertiery equations can be used to describe this surface and the parameters I mentioned are such parameters.

This implied vol is different to realised vol which is the average (standard deviation) of the underlying movement on a tick/hourly or minute basis. In a broad sense, they are related in that if the underlying starts moving violantly, up or down or both, then the realised /historical vol is higher and this will start pushing implied vol up through market makers wanting to own options to profit from the gamma of more violant underlying movement.

To answer your question directly about modelling future vol I would say there are 3 approaches you can take, but I would always suggest using the implied market vol as your starting point. The 2 alternatives are:

1) that if you are looking to profit from gamma movement, look at the historical (underlying movement) vol for as you say, a period of observation matching the days to expiry. To extend this, you can look at the expected p&l's from gamma and comapre this to the implied vol and expected theta of the options

2)If you want to look at vol movement as the underlyier moves, map more expected p&l's from scenario analysis (e.g. if you expect vol to double on a move down within 15 days of a 30 day option, plug these parameters into your analysis and look at your Greeks and p&l, then look at worst case scenario where say the underlying goes to your long strike and vol is smashed through the floor).

In short though, there is no accuarte way of predicting future vol. Ironically, the best estimate is implied vol (implied vol is a forward looking parameter that the market assesses)

 

[schreibdave]

Hi Slik

Right now my focus is on selling bear call spreads and bull put spreads.

This conversation has been helpful to me, so let me ask a few follow up questions:

1) You say - "I would always suggest using the implied market vol as your starting point."

What measure of IV would you consider the "implied market vol?' The IV of the ATM calls or puts? An average of all IVs for that month? An average of the IVs for the two legs in my spread?

2) Since my trades are based mainly on having an opinion about direction and time - and not on having a real opinion about future volatility, I am inclined to just look at a chart of IV, see which ways it is trending and assume that that trend will continue. For setting stops I would also model the worst case scenario where I am wrong on direction AND on IV and set my stops accordingly. Would you suggest a different method on dealing with the IV issue?

Thanks

 

[slik]

1) Use the IV's of the strikes you are using, if they are significantly different, use an average- it will give you a reasonable estimate of the Greeks position.

2) I worked within fixed income options and in my experience I would say that hisorical implied vol is very volatile in itself and any meaningful trends can be blown out of the water in the blink of an eye.. or the rumour of an insolvency.

Futhermore, as you quite rightly say, the vol is more dependant on the underlier and I would guess vol getting bid on the rally as credit spreads blow out, so I think what you suggest is all you can really project really- mapping best/worst case scenarios as accurately as you can with the vol/underlyier movement.

Maybe look at the highs and lows of historical vol and the current implied vol (make sure you are comparing like for like though e.g. 30 day options on the at the money strikes) and that will give you boundaries to stress your position and setting stops.

If you look in Hull, I do remember a section about volitility predictors, but that is an area 1) I have no experience in 2) very few market makers would look at as a tool (being a mm myself)- As I have said, the recent markets make the vol of vol very high in itself!