General Discussion on Option Modelling

Bruno

Member
Decided to cover a bit more pricing issues to be in a position to be able to plot all sorts of combos for instance on Excel.
The B&S spreadsheet is available on request: it contains Raphson-Newtown and a few other goodies like interest rates extraction from the internet.
Video available here https://my.aeromir.com/go/u.ffe2/fid/10940/hash/8698454FEAE86858BE35.html
With the available tools now, it is calculate and plot projections using Individual Volatilities (called EIOIO on OptionVue).

For those also interested in Nelder-Mead, here is a link to the spreadsheet with fairly simple VBA code: https://www.dropbox.com/s/wfw121vobn3ijgn/neldermead_downhillsimplex.xlsm?dl=0
 

Attachments

status1

Active member
Hmmm... I can't seem to get the same price for the put and call as it's shown in TOS

I changed the date to Nov 15 since there is no SPX expiration for Nov 14
I also changed the underlying to 3006.72 for today's closing and left the 3000 strike the same
Than I also changed the volatility to 14% Everything else I left the same
So I got 46.68 for the call and 40.32 for the put while on TOS I have 39.50 as the mark for the call and 32.85 for the put

Is there anything else I should change to make it closer to the TOS numbers ?
 

status1

Active member
There seems to be a problem when I try to close it
It says excel has stopped working and after I click on debug I get this error message
 

Attachments

Bruno

Member
Hi Status1,
Apologies for this quick & dirty update to the B&S model spreadsheet.
I added a query to retrieve an options chain from Yahoo from which I could derive a simple vol skew.
There could be a way to be more specific but at this time the query takes no parameter hence it downloads the default one and that is the DEC chain that is loading on the spreadsheet (at this time).
I believe you would be able to get what you want by using simple RTD calls from either IB or TOS in the right-hand table (1% OTM etc.)
Excel then extrapolates ATM IV and the slope of the vol skew.

I have tested it on Office365 32bit and I have never had any such error at this end.
Please download the updated version (same link) and try again.

Hope this helps
Bruno
 

status1

Active member
Thanks for the explanation
I thought this was a calculation using BS to get the price of the options but I guess not
 

Bruno

Member
Assuming B&S is a valid model, there may sometimes be minor errors in the treatment of dividends etc, but 99.999999999% of calculation errors arise from putting in the wrong IV estimate in the equation.
 

status1

Active member
I am using Office 2010 and it works at work on Windows 7 64 bit but it's not working at home with windows 7 32 bit
I get the same error crash when I am closing the excel spreadsheet
 

Bruno

Member
Correct IV ? It is the million dollar question for which even simpler pointers (i.e. theory) would take a long time to answer. I can recommend a few books (for instance Options Pricing Formulas).

About the crash, I don't think that is due to a 32 vs 64 bit environment. There may be something else, probably the way web queries are handled. There is nothing in the VBA code that could cause such error.
 

AndrewS

Member
The only "correct IV" is the one obtained from the option price this moment. A minute from now everything can change and buying or selling pressure driven by human emotion will change the equation, sometimes dramatically. Based on recent history, computer models will usually give you a fairly accurate prediction of IV in the near future but the further into the future you want to go the more fraught with error that prediction can become. Evaluating future IV is one area, for now, where computer models need to be matched with street smarts and maybe a bit of good old fashioned luck.
 

Bruno

Member
Volatility is essentially non directional and it is indeed quite difficult to predict it. One major flaw of B&S and consorts is that they cannot handle price jumps (discontinuities). There are "jump-diffusion" models but they only do marginally better.
It is a tricky topic and I'll expand on it when I get to volatility surface analysis.
 

Marcas

Member
Andrew put "correct IV" in quotation marks. Rightly so! There is no such a thing as correct IV value in general. There can be correct iv for given model and applied parameters. This "correct IV" returns price that is equal to market price (chicken and egg dilemma). This correctness can be verified only for single time-point that equals "now".
Modeling t-lines is most of the time 'good enough' to get an idea what will happen to our position if nothing extraordinary happen to markets.
Predicting IV is pretty much impossible. It is the same as predicting where market will go. Some modeling techniques can be 'better' in 'predicting' Vol changes in some market conditions in selected periods of time. Sure, any statistical methods can, and should, be applied but you can gain sense of what will happen to Vol simply by watching it while trading.
Question trader should ask himself is this: Do I need 'correct IV' for my trading? Will it give me an edge? If the answer is 'yes' then go ahead, if 'not necessary' then ask yourself 'what knowledge/skill set will give me an edge' and focus on that.
Beside that, for option traders knowledge and understanding of modeling is very important although I'm not quite sure if it is essential.
 

status1

Active member
What I was trying to find out is if this BS formula can calculate the put and call price for SPX and I understand that the iv is very important for that calculation So my question was where you get the iv from to make the calculation ?
Is it just some obscure number that only market makers use that no one else knows ?
Or is it just not possible to calculate the put and call prices since the iv is not available ?
 

Marcas

Member
You can be lucky and get it right by using fancy model or by guesstimating. What you should know is that modeled t-line is most likely highly inaccurate and be ready for worst case scenario by putting appropriate hedges or by sizing your positions. You never know for sure how it will be :) That's trading.
 

Marcas

Member
Or is it just not possible to calculate the put and call prices since the iv is not available ?
It is absolutely possible to calculate iv on your own. No problem. Problem starts when you pursue 'correct IV'.
(That holds if we are talking liquid assets like SPX or RUT. It is different matter if you look at rarely traded underlying like DDGPUXX :) )
 

garyw

Active member
@status1 : Are you seeking a prediction of IV at some time in the future, or just want to know what is is now (or a real time spot value), or at some time in the past? -- It is unclear which you mean! The first requires the sacrifice of a chicken and expertise of a Jamaican lady (Mambo voodoo priest), but the second may be computed.
 

Bruno

Member
There are mathematical / statistical ways to model IV specifically. Local volatility has its own dynamics (auto-regression, clustering, mean reversion etc.) and its own drivers, mostly fear translated into hedging activity. Volatility also follows like price (considering both are tightly intertwined through a model like B&S) the necessity to follow a pattern to cancel out any arbitrage opportunity. This is how a volatility model can expand to neighbouring data points to attempt to turn a local volatility model into a global one.

Now, to answer @status1 : Market Makers are on the sell side. They have to provide liquidity and they have their own book to look after. Let's not forget also that even on the buy side, there is a huge amount of activity to just hedge assets. All things considered, a market is a market, and price is meant to reach an equilibrium at any point in time between buyers and sellers. Market participants don't trade volatility as such: they trade products that are closely linked to volatility.

Comment on @Marcas: Absolutely correct. As a matter of fact, assuming a model like B&S does a fair job, one can technically only then trust the T+0 using individual volatilities. Everything (e.g. vol surface or as mentioned above generalisation endeavours) else is a conjecture. That being said, that is no reason to throw the baby out with the bathwater. If constructed properly, those conjectures are mostly valid within the realms of the model's hypothesis and its inherent weaknesses.
 

status1

Active member
I just wanted to see if the current price of options can be calculated I am not planing to use it in any modeling of t+0 line or sometime in the future or the past
I have seen some calculators online where you just enter the all the known data and it gives you the price of the option One of the unknown is the volatility
Sure I can guesstimate the iv but that is just adjusting the IV based on the known option price so basically you are calculating the iv not the option price
 
Top