Tuesday, September 29, 2009

Mail Time- Implied Vol vs. Historical Vol

Hi Tyler,

I got a couple of questions about the spread between historical volatility and implied vol.

When you study this spread you’re actually seeing different volatility periods right? The implied is for the next 30 days and the historical is from the last 30 days. Do you make any kind of adjustment to the series or just leave them like that, with different dates. If the implied vol. you're studying is for an option that expires in say 20 days, do you still compare with 30 day HV or do you use the same number of days for the historical volatility as the number of days to expiration?

Thanks a lot!



When I’ve talked about comparing historical volatility to implied volatility in the past, I’ve typically been comparing 30 day HV to IVolatility.com’s IV Index Mean. As the name implies, 30 day HV does look back over the past 30 days so you’re correct in that assumption. As far as the IV Index Mean, I’m almost 100% sure the graph I use looks forward 30 days. Here’s how IVolatility describes their IV Index Mean:

“The Implied Volatility Index is calculated by using a proprietary weighting technique factoring the delta and vega of each option participating in IV Index calculations. In total, we use 4 ATM options within each expiration to solve for the Implied Volatility Index of each stock. This IV Index is normalized to fixed tenors (30, 60, 90, 120, 150, 180 days) using a linear interpolation by the squared root of time."

I believe the IV index they’re using in the graphs I’ve shown is the 30d IV, which shows what volatility is expected to be (based off current option prices) in the ensuing 30 days. Keep in mind you don’t have to use the IV index mean. Remember each option has its own implied volatility, so you could simply use the IV of whatever individual option you’re trading or the average IV of the options trading in that expiration month. Most brokers display that information. Here’s a graphic from Think or Swim showing the IV levels of RIMM options for each expiration month. This picture was taken on Sept. 24th, the day before earnings (click image to enlarge).

[Source: ThinkorSwim]

Notice how the front month (OCT) options were trading at higher vol levels than the later months. This helps show the expectation of higher volatility in the short term (due to the imminent earnings announcement), but lower volatility in the back months. At the time the IV index mean was around 56%- which seems close to an average of the IV of every expiration month.

In tackling the question as to which HV I use, this is where personal preference comes in. The conventional approach (at least the way I was taught) is to use the historical volatility that looks back the same amount of days you’re going to be in the trade. So if I’m entering an option trade for approximately 30 days, then I’ll compare implied volatility to 30d HV. If it’s a 60 day option trade, then I may use 60 day HV instead. It really comes down to which HV reading you think gives the best indication of future volatility. Sometimes I use 10 day HV because it’s more sensitive to recent price action and thus gives a more accurate picture as to how volatile the market is right now. Let’s say you have the following volatility readings on XYZ:

10d HV 50%, 30dHV 30%, 60d HV 35%, IV Index Mean 47%

Which HV reading should we use in comparison to current IV? My answer would be the one that is closest to how volatile the stock will be throughout the duration of your trade. Is there any way to really know that before hand? Well, not really. It’s an imperfect science. Currently 10d HV is much higher than 30 and 60 day HV showing that the stock’s volatility has picked up recently. If you think that increase in volatility will persist going forward, then I’d probably use 10d HV in comparison to IV. If, on the other hand you think the increase in stock volatility is an aberration and it’s likely to revert back closer to 30 or 60 day HV, then I’d use those readings in comparison to current IV levels. In the case of 10d HV vs. IV (50% vs. 47%), it seems like options are right in line. If we compare 30d HV to IV (30% vs. 47%), then options seem a little rich and I’d probably lean toward being a seller.

If you missed my four part series on Historical Volatility you can view them here: Part I, II, III, IV

IVolatility.com also has some great resources within the Knowledge Base or Education part of there website. I'd recommend checking that out if you seek more information on volatility.


It's probably important to add that most data providers use trading days when calculating HV and calendar days when calculating IV. So while 30d IV consists of 30 calendar days, it comes out to approximately 21 trading days. Thanks to Bill Luby for pointing that out. His post The Gap between the VIX and Realized Volatility sheds more insight.


Decio said...

Thanks for the answer and congrats on your blog!

Bill Luby said...

Tyler, I think most data providers use trading days for HV and calendar days for IV.



Tyler Craig said...
This comment has been removed by the author.
Tyler Craig said...

Thanks for pointing that out Bill. I know there are still a few details I'm not aware of concerning volatility. Your site has certainly helped a lot and I welcome any feedback or corrections on anything I post.


One inspiring post.