Don't know about the ATM straddle, but 84% of not breaching 1 StDev down and 84% chance of not breaching 1 StDev up should lead to . . . 68% success in the prediction. Right?
Oh yes it does help ...... I kick started your thread after 20 hours of zero replies, now you have an active thread. BUT all those replies (and future replies) will not answer your OP, you will end up with more questions than answers. That is a fact.
It's as if you can't read. I received 2 very helpful tips: 1. average ATM call IV with ATM put IV 2. calculate the IV the way the VIX is calculated I'm sorry the trading world doesn't wholly conform to your style, but you really need to open your mind to the idea that you might have some things to learn.
The volatility parameter, sigma, is the actual or realized volatility of price. It's definition is one standard deviation of the logarithm of the daily percent price change annualized. For example if you are looking at 45 days until expiration, you would calculated the historic realized 45-day volatility of the underlying. You could use that number in the formula, but volatility likely will be different 45 days from now. So what number do you input for sigma? Volatility forecasting is both art and science. No need to be a volatility guru, just a volatility Forrest Gump. That is know when volatility is relatively low or relatively high. If your sigma calculation is at the low end of the historic range, you may want to bump up your sigma number, or maybe go out a little further in standard deviations to account for the anticipated rise in volatility in 45 days. Also, this is for what happens at expiry. The probability of touch any time before expiry is approximately twice the probability of finishing OTM. Attached is a simple spreadsheet that calculates the 45 day volatility for ExxonMobil.
You might want to go to the DOE-EIA site and look up the 'Market Prices and Uncertainty Reports' they create each month for crude oil and natural gas. If you poke around the html page you'll see pdf and xls downloads in the upper right. The Excel download is nice, it has relatively recent data in it and you can copy and paste current data or link it to a data source. All you need are futures prices, ATM IV and DTE (or create your own DTE counter). The file provides results in table format and also draws a chart with 1 or 2 SDs up and down. You can insert 1.5 or 3 SDs. Overall, it's a good learning tool and the formulas are right there. Customize the existing file or start fresh without having to worry if you've got everything nested just right (the formulas are long). And when you're finished, you can brag about/bore your friends & family with big-league talk about "trading volatility cones." ;-)
Actually your answers are helpful to me. Short term (less than a couple of months?), statistics and equations do help getting the expected and probability of outcome. The problem is that there are too many smart folks in this business so everything is priced in and there is really no advantage. Longer term however I have not found statistics and equations to be very helpful. You are right, it is better to use other means to assess what the likely outcome will be. As to the original poster's question of what IV to use, in my not so smart opinion, the market gives you a price, so the market has a number for IV therefore you can reverse the formula to calculate the market IV. If you want to forecast what the spot price will be different from the market, you in effect have to get a sense of where future IV will be and use that in the equation be it historical, call/put average, extrapolation....and not anything that is currently available? I don't know if I am making any sense?
Not exactly. I am attempting to bound the future price of the underlying as +/- X StDev. StDev is a function of volatility, so therein lies the question. I like the idea of comparing HV to IV to get a feel for where IV might go. It's all (seemingly) a bit of an art, and of course the model itself has some built-in assumptions that may only loosely conform to reality. I get it, it's a model, and it is not always accurate. Sometimes we do the best with what we can.