How to "compute" option prices from daily values

[Derrenoption]

If anyone is able to specifically identify a single instance where an SPX PUT option IV moved in an opposite direction to VIX

I think that would be possible if you wrote a script. If you put the VIX in memory and then read in each options IV in memory and then compare each option against the vix for the same period/date(s). I think you are looking for divergence? That is very interesting.

 

[Derrenoption]

The IV input is easy to extract by 'reverse engineering'. Give me the Expiry date and I will give you the correct one... Wait, it's tomorrow right?
So IV will be around 25.5?

I am not sure what you did here. For example if the daily bars LAST price of AAPL is at 117 and we look at strike 114 which has a MIDPOINT value of 0.62.
The our goal is to find out what the MIDPOINT of strike 114 was when AAPL was at 115 during the day.

Then we would need to know what the IV was for strike 114 when AAPL was at 115. But to find out what the IV was using the BSM, we need to know the option price in that equation which is the original parameter we are looking for so that seem to be a dead end?

Did you do anything in your formula that I missed there? You found out IV around: 25.5?

 

[stepandfetchit]

I think that would be possible if you wrote a script. If you put the VIX in memory and then read in each options IV in memory and then compare each option against the vix for the same period/date(s). I think you are looking for divergence? That is very interesting.

My suspicion is that Jack is looking at a more macro time frame and extending his knowledge of slower moving events to intra-day movement without re-thinking the micro dynamics of what we are addressing. <-- not stated well; Without hard evidence; and coupling with other loss of precision in statements from him, my confidence level is in the high 90's that this is a wild goose chase for something that does not exist. I believe he is thinking about something else! Even if he is correct, your data sources do not have the capability to reflect it. The SKEW value is EOD only so is merely a poor representation of the EOD price data you already have, so should be ignored as it will add error in your specific case. (SKEW being the CBOE SKEW INDEX).

 

[JackRab]

My suspicion is that Jack is looking at a more macro time frame and extending his knowledge of slower moving events to intra-day movement without re-thinking the micro dynamics of what we are addressing..

Uhm, what do you mean by macro time frame? And what are we chasing exactly? It thought OP was looking for a way to compute IV's of specific options?

Look, in the end it's quite simple. You need an options pricing model with the correct inputs. Most inputs are easy; time to maturity, interest rate, dividend, underlying stock price. The IV what we're after is a bit more difficult... but in general is also not too difficult to get. If I know the price of the ATM straddle I can compute every options series. The skew isn't fixed, but doesn't change by that much in normal circumstances.

The model in the excel file I have uses B&S as the main model. B&S has it's faults, but you can use it with some workarounds. The IV I get might be off somewhat, but that also depends on whether you're looking at vols in business days or calendar days. And somewhat depending on the model you use.... but in general, it works.

Also, you can't use CBOE skew index for single stocks... there's no way that the index skew will match those of single stocks. I assume that skew index is either the skew of say S&P500... or an average of a basket of single stocks? The skew in banks is different from skew in techs and pharms are different again. Even within a sector stocks have different skews... depending on the risks characteristics of the individual stock. S&P500 skew isn't even anything close to an average of the underlying stocks' skew. Same as that the S&P500 ATM IV's are not an average....

 

[JackRab]

I am not sure what you did here. For example if the daily bars LAST price of AAPL is at 117 and we look at strike 114 which has a MIDPOINT value of 0.62.
The our goal is to find out what the MIDPOINT of strike 114 was when AAPL was at 115 during the day.

Then we would need to know what the IV was for strike 114 when AAPL was at 115. But to find out what the IV was using the BSM, we need to know the option price in that equation which is the original parameter we are looking for so that seem to be a dead end?

Did you do anything in your formula that I missed there? You found out IV around: 25.5?

Okay, so... assuming it's the AAPL Oct21 114 put we're talking about...

If I look at my screen now, with the last prices of friday close being
AAPL last traded prices were 117.66 and 117.67 in the last second before close (of the options).
114 put 0.14@0.16

If I put all variables in my model, where I use ATM vol of 17.2 to match the last quotes of the 117 and 118 calls and puts... I get this:

Skew seems very cheap here actually, I wonder why. Maybe the Samsung problems are putting a floor under the stocks of AAPL.

Anyway, this is IB's screenshot:

Using these variables, I can just change the underlying stock level to see what the 114 put's price was when AAPL was at 115. But likely the vols were a lot higher etc. But... if you know what price the option traded at that level, I can give the vol easily. And, very importantly... all the options are related to one and other. So, the ATM price basically sets the main IV... and then your add skew. So if skew is fairly constant (which it usually is) then on the basis of an active traded ATM option, you can compute the whole series. And even other months, with some assumptions about the vol distribution over time. That's how market makers do it. You have to start somewhere.

I used to quote options prices in very illiquid series, where we had no info because nobody was giving quotes. So you start by making assumptions about the ATM front month IV, based on historical vols... future expectation with events in mind. Toss a certain base-level of skew in there, also based on past movements. Add dividend, which generally is easy when the company has paid dividends in the past. Uhm, than do the same for other maturities...
And you wait and see what happens. If you're offer gets lifted, you raise the IV...

As with anything, there are assumtions in everything. Even professionals//market makers have slightly different vols/models. But in the end, the market is what the market is. And it changes constantly. They all keep databases of the inputs they've used historically... so they can assess whether they and the market prices 'correctly'.