What are the differences between -- and how should we compare -- historical volatility and implied volatility?
Time for a little Twitter mail! Full disclosure: It's not actually my Twitter mail, I just got included at the end of the conversation.
Let me unpack this a little, since it's a tough question to answer in 140 characters. HV is historical volatility, same as realized volatility. It's the actual volatility of the underlying. The "10" is the time frame, so HV10 looks at the last 10 days of trading.
I do tend to use HV10 for all comps, but there's nothing particularly wrong with using more days. I use the "10" because, to me, it's more representative of the here and now. And options tend to price on the volatility in the here and now. In fact, that's a much bigger driver than expectations of future volatility. The best way to say it is that if you want to predict future volatility, simply look at the recent past.
Implied volatility (IV) under HV is atypical, but far from unprecedented. Over time, we generally see IV30 with a 3-to-4 point premium to HV. But that certainly varies. HV lags, and right now it still includes some relatively volatile sessions from one and two weeks ago. IV looks forward and sees a long weekend, and not all that much in the way of expected news into January earnings. In fact, an individual stock that has gapped post-earnings is the perfect example of HV deviating wildly from IV. Specifically, HV will shoot up and IV will decline with the news out. Options may look "cheap," but they're quite likely not. It's an apples-to-oranges question.
And that leads me to the latter part of the comparison. Why compare HV10 to IV30? Why not HV10 to IV10 -- basically, CBOE Short-Term Volatility Index (VXST), if you're talking about the S&P 500 Index (SPX) -- and HV30 to IV30, etc.
The answer is again apples to oranges. HV lags and is measured in trading days. HV10 is basically two weeks' worth of realized vol. IV looks forward and is measured in calendar days. Even if you adjust the syntax (call is IV23, e.g.), there's no particular reason why you'd need to look at the same time frame going two different directions.
Disclaimer: Mr. Warner's opinions expressed above do not necessarily represent the views of Schaeffer's Investment Research.