IV can be computed exactly, if you know the "correct" option premium. Usually it's computed from (Bid + Ask) / 2, ie. the MidPrice, regarded Bid and Ask are non-zero... Of course also using the other option parameters like Spot, DTE, Strike, Risk-free-rate etc.
Well get to it earth! Determine what you think the "correct" price is, calculate your IV and start printing some extra pocket change. Personally, I prefer to do it the other way around... I make an estimate of what the vol will be and then look at the price. I figure it's easier and since it's mathematically just six of one, half-dozen of another it ends up working out better for me.
I think my approach to compute the MidPrice from Bid and Ask, is IMO much easier. Just tell us how you make your estimate, which data do you need for it?
I've been running tests on thinly traded options, and it would appear you are correct, the live IV is based on the midpoint. However, on thinly traded options, on stocks that don't trade that much volume, the bid and the ask can change dramatically without any change in the underlying. That's what I was referring to when I said "they do that all the time." For example look at the June $77.50 calls on PZZA. They jump around. One minute the bid will be $3, look at it a minute later and it'll be $1.50 Same with the ask, it'll go from $4 down to $3.50 and then pop back up to $4. Why is this even important? I don't think you can get a reliable input into your system when they jump around like this all day. Even with the underlying not moving. https://www.option-price.com/implied-volatility.php
I don't trust closed-source software. Those pricing engines are all over the web, they could have been written by a cock-eyed half-wit, they're too slow to actually implement in a trading system. Sometimes I'll play with them for fun or just to double check something. They're good for messing around though... I don't think you could turn a big profit with them though.
Forget it, it's buggy! Example: for the Call option below it gives a wrong IV of 500. My method gives this result: Spot=100 Premium=90 Strike=100 DTE=60 RiskfreeRate=0% DividendYield=0% DaysInYear=365 : IV=811.387040
Lol. I was just using a 90 day EWMA (a=0.94) like everyone else apparently, but switched to garch a while back... tough for me to say if it made a huge difference. I don't trade anything too close or too far from expiration though.
No, I was just using it to verify what my trading software was showing. I took multiple screen-shots as the bid/ask changed because I wanted to see if my trading software's IV was in fact the mid. It was. That's the only reason I used that calculator, just to see what it came up with the same IV using the mid. And it did in fact match.
LOL! IMO an academic trash method So, you have to keep a database of timeseries to be able to make your calculation. Not necessary in my case. Which one is easier?