Implied Volatility vs. Real Volatility

Quote from justrading:

The Ferrari driver does not need to know the physics of how the gas in his tires helps him. If it affects the way he has to manage tire wear, or specifically how the tires performance varies through the race, then all he needs to know is what he has to do as a driver given those factors. R&D chaps figure out the technical details.

As for jargon being used for precision, that is perhaps an oxymoron. Try 'large negative gamma position' for precision - everyone and his dog can understand that. Is there a lexicon somewhere that specifies short gamma implies large gamma risk? All my searches for 'define short gamma' have not turned one up, so at precisely what point does a negative gamma position become short gamma? Where is the boundary?
More...

No. short gamma and large gamma aren't the same thing. All negative gamma is short gamma. All negative vega is short vega. All negative delta is short delta. The only place where I found it weird was in swaps. CDS terminology is : long CDS means long protection. I never traded CDS, but my old convert guys did and this is what they explained to me when I was looking at some potential relative value trades.

The car analogy: unless you have people who are completely dedicated to working with/for you (90% of their income comes from you) then you are the entire Ferrari team.
 
Quote from newwurldmn:

Given that a trader is using options to compete against others it is a good idea to view them as a race car where understanding the internals on your own will likely fond you opportunities.
More...

Good comment. Plus I imagine top professional traders have teams of whiz kids (lol I'm becoming too American reading this forum) who will for example do all the hard work to adapt a complex math model to the trader's needs and then write code to execute very fast etc.

We, on the other hand, have to do everything on owr own. I am not suggesting you/me have to learn everything out there but we need to be realistic and not oversimplify or underestimate some things.
 
Quote from newwurldmn:

No. short gamma and large gamma aren't the same thing. All negative gamma is short gamma. All negative vega is short vega. All negative delta is short delta. The only place where I found it weird was in swaps. CDS terminology is : long CDS means long protection. I never traded CDS, but my old convert guys did and this is what they explained to me when I was looking at some potential relative value trades.

The car analogy: unless you have people who are completely dedicated to working with/for you (90% of their income comes from you) then you are the entire Ferrari team.
More...

You said "You are short options that have a lot of gamma risk.....", which I expressed as large gamma. Isn't all large gamma risk something to be wary about, especially if short gamma?

Re the car, my original point was you do not have to understand the math backwards and forwards to profit from it, or protect yourself. You want to know how to use the idea or concept to profit, that's all. You post here, and trade, using either a mobile device or PC. You do not need to know the design specs of every chip and board in the device to profit from using it. Nothing wrong if you want to, but is not necessary.

I have no idea how IB programs/calculates the Greeks on the options trading page. I just use them. Ditto with Hoadley.
 
Quote from justrading:

You said "You are short options that have a lot of gamma risk.....", which I expressed as large gamma. Isn't all large gamma risk something to be wary about, especially if short gamma?

Re the car, my original point was you do not have to understand the math backwards and forwards to profit from it, or protect yourself. You want to know how to use the idea or concept to profit, that's all. You post here, and trade, using either a mobile device or PC. You do not need to know the design specs of every chip and board in the device to profit from using it. Nothing wrong if you want to, but is not necessary.

I have no idea how IB programs/calculates the Greeks on the options trading page. I just use them. Ditto with Hoadley.
More...

"Short gamma" means that gamma is your principle risk. "Short vega" means that vega is your principle risk.

Understanding the design specs of a PC might matter if you derive your pnl from complex technical infrastructure. Stat Arb guys care about this stuff. I don't do any stat arb so it doesn't matter to me. I do do a lot of options so I spend a lot of time understanding how options move.

Greek models may matter. I just use european black scholes. I understand where my models fail and when they are reasonable. I also am aware that this is a significant limitation in what I can trade. There are firms that spend a lot of time on their greek models and they have some very clever strategies around them. The more you know the more you can do and your competitors are doing a lot of it. You can play the same games of them or another game. In the end, you have to come up with your own framework and if you want to trade options it helps to understand what you should be looking out for on your own.

EDIT: If you are using the greeks as a form of risk management, then you should understand how the calculations are done.
 
Quote from justrading:

You mean risk management?
More...

Kind of. It's all risk managment.

Don't look at an option as max(S-K,0).

Look at it as delta + gamma + vega + theta + second and third order effects if they are relevant to the situation.

When I was trading professionally, I never knew what the dollar price of the options in my book were. I only knew the implied vol, and the delta, gamma, vega, theta, rt vega, etc. that i was running.
 
Quote from quatron:

I've done that. My multi-leg execution strategy does not pay the spread, it trades on theoretical (fair) values on average. It quotes the legs and trades a spread maintaining vega-neutral exposure while executing.
My backtester simulates execution of quotes/limit orders very accurately, according to the price-time priority in the book. Live results are almost the same as backtested.
More...

Thanks for the reply. Sounds like you're using TAQ data for your backtests. I'm not quite ready to take on all that data management. My question was more focused on interval or EOD data. For example, if I get a backtest signal to sell the ATM straddle based on HV/IV analysis and the EOD NBBO quote on the straddle is 3.25 / 4.00...what is a reasonable assumption for a fill price? The answer to this becomes more critical the more illiquid the options are; anything short of full NBBO on those and I feel like I'm fooling myself with noise inside the bid/ask.
 
Quote from newwurldmn:

"Short gamma" means that gamma is your principle risk. "Short vega" means that vega is your principle risk.

Greek models may matter. I just use european black scholes. I understand where my models fail and when they are reasonable. I also am aware that this is a significant limitation in what I can trade. There are firms that spend a lot of time on their greek models and they have some very clever strategies around them. The more you know the more you can do and your competitors are doing a lot of it. You can play the same games of them or another game. In the end, you have to come up with your own framework and if you want to trade options it helps to understand what you should be looking out for on your own.

EDIT: If you are using the greeks as a form of risk management, then you should understand how the calculations are done.
More...

OK, I was going on Natenberg's "A large gamma number , whether positive or negative, indicates a high degree of risk; a low gamma number indicates a low degree of risk." He did add further on that large negative gamma is particularly risky. Knowing that gamma spikes close to expiry, I get flat in good time.

Are you trading European style options? I thought the defining difference was no early exercise permitted?

Re knowing the games, the mathematical models are transparent, and there are hundreds of traders (thousands?) who know them well enough to profit from nuance there. There are probably loads of computer programs trading and milking every last cent possible from minor variances from the norm. By the time I look at all the IVs of the various strikes and expiries for the option chains for a single stock and compare with HV, the programs could have milked options for a hundred stocks.

What I find more informative is the sort of thing I read recently. A trader had profited from theta decay, and had several hundred dollars profit on the Friday. Thinking that the weekend would increase the theta decay, he held. On the Monday there was a gap opening and not only did he lose all the unrealized profit, he went into negative territory. Apparently market makers start discounting the weekend theta decay from Thursday afternoon or Friday morning. If true, then this is practical stuff that no model tells you.

Still, to each his own and I meant no criticism of you or anyone else for using whatever approach works for you. There is a particular type of beast on ET that believes there is only one way to do something. I belong to the 'there are many ways to skin a cat' group.
 
Quote from justrading:

OK, I was going on Natenberg's "A large gamma number , whether positive or negative, indicates a high degree of risk; a low gamma number indicates a low degree of risk." He did add further on that large negative gamma is particularly risky. Knowing that gamma spikes close to expiry, I get flat in good time.

Are you trading European style options? I thought the defining difference was no early exercise permitted?

Re knowing the games, the mathematical models are transparent, and there are hundreds of traders (thousands?) who know them well enough to profit from nuance there. There are probably loads of computer programs trading and milking every last cent possible from minor variances from the norm. By the time I look at all the IVs of the various strikes and expiries for the option chains for a single stock and compare with HV, the programs could have milked options for a hundred stocks.

What I find more informative is the sort of thing I read recently. A trader had profited from theta decay, and had several hundred dollars profit on the Friday. Thinking that the weekend would increase the theta decay, he held. On the Monday there was a gap opening and not only did he lose all the unrealized profit, he went into negative territory. Apparently market makers start discounting the weekend theta decay from Thursday afternoon or Friday morning. If true, then this is practical stuff that no model tells you.

Still, to each his own and I meant no criticism of you or anyone else for using whatever approach works for you. There is a particular type of beast on ET that believes there is only one way to do something. I belong to the 'there are many ways to skin a cat' group.
More...

Yeah there are lots of people who are arbing things like that theta on thursday for the weekend, etc. Personally, I think it's a dangerous strategy unless you run a diversified book that captures bid/offer rather than pays it.

I trade American listed single stock options. I use European black scholes because it was real easy to implement. I know when it doesn't work and when it's reasonable enough for what I do. It's important to understand what's under the hood - especially for options.

I was responding to the comment about Graethal (whatever his name is) offering a more simple explanation. I am not really sure what the point of this discussion is anyway.

Good luck with studying of the options. My advice is to understand the pricing mechanics. It will help you understand what the greeks mean and how they will change. As a retail trader it's hard to make money any other way with options except for buy writing, etc. In that case, focusing a lot of the greeks is less important.
 
Quote from filthy:

i like gatheral's book but i don't see any reason at all that a vanilla option trader would need to read it. the models he discusses are really only necessary when spreading and hedging exotics. a possible exception would be someone who is trying to trade vix options against sp options, but even then the model driven approach is not what i would go with.

BS (and binomial) models are good not because there assumptions are realistic. they are good for reasons unrelated to their assumptions. they turn fast moving things (option prices) into slow moving things (implied vols) that are easier to keep track of and compare. BS isn't a model that (anyone other than taleb's strawmen) thinks represents reality. it is just a translation device.
More...

Hey filthy, big fan. When's the new book coming out?
 
You must log in or register to reply here.
Back
Top