Coffee has jumped recently and I wanted to see how two common methods of ATM IV measurement compared on a day with a large move in the underlying (today). A quick and dirty analysis is attached. If one is just using the data to keep an eye on things via a time series chart, the IV Index provided by ivolatility.com (using interpolation to get a constant maturity) seems to be fine. It should be spot-on compared with the 'traditional' measurement if a 30-day IV Index increment matches the term of a particular option month (see shaded comparisons). One big honkin' caution is in order, however, when there's a significant skew in the IV term structure , e.g. coffee today or natural gas the last couple of weeks. Note: I'm just comparing the two measures on February 3. The IV Term Structure/Forward Vol Curve has additional dates, but they're not the same for each data set. Nothing earth shattering, I know. I'll have to come up with another project if I ever expect to win the Nobel Prize.
(a) is it ATM forward on both displays? (b) do both models calculate vols from the right forward? (c) is it business days or calendar days? (d) is it interpolated or closest strike? etc...
What vol do you think IVol is using for the interpolation? Do you see that the durations don't match? ATM skew? The interpolation seems pretty well fitted. The fwds don't match and all of the interpolated vols are higher. What exactly do you think you've found?
You know all this stuff, so to you it's pointless. For others, especially those who are somewhat new to options and rely on the "IV" that is charted on common platforms from retail brokerage firms - IB, TOS, LiveVol Securities, TradeMonster, OptionsHouse and Scottrade - my point is that the constant maturity/interpolated volatility they see on the default IV/HV chart might be fine as a way to take the temperature of the market. However, if the product is a physical commodity being affected by the weather, a stock before earnings, or pretty much any product/situation where sharp sentiment is front-loaded, the constant maturity IV, can be quite misleading. I like the constant maturity IV for an easy comparison across products and to find extremes, but when it's time to trade, it's time to dig into the current and historical data for an individual contract month. Edit because Atticus removed the 'pointless' comment. I wasn't looking for anything special. I just wanted to see how the two sets of data looked on Feb 3 only, which was a really big day in coffee. It's just an observation - and an obvious one to those with your experience - that the constant maturities can be deceiving in volatile times.
Interpolated vols that are 2bp off of the traded IV and six days off the listed maturity. My point is that the constant maturity set is derived from those same traded vols and look well-fitted. I don't see anything here that you wouldn't see on *any* day. You used a large range day to make your point and the interpolated vols are all higher. Shouldn't they be lower to make your point?