Wednesday, September 9, 2009

Historical Volatility Demystified Part II

Historical volatility can be measured over any time frame you desire. Commonly used time periods include 10, 30, and 90 days. If you want to view different lengths simultaneously, some software programs allow you the ability to overlay them on top of each other which aids in making comparisons. Take a look at a chart of the S&P 500 index which compares historical vol over three different time frames (click to enlarge).

[Source: EduTrader]

Here are a few important nuances with the different time periods:

1. Much like moving averages, short measurements of HV such as 10 days move quicker and are more erratic or noisy. The advantage to using a shorter term level of HV is it can provide a quicker indication as to whether or not volatility is picking up or dropping, but it gives more false signals due to its erratic nature.
2. Longer measurements of HV such as 30 or 90 days are slower moving and less erratic. As such they provide slower indications as to whether or not volatility is picking up or dropping.
3. When you get an extended expansion or compression in volatility (i.e. the stock continues to increase or decrease in volatility), shorter term levels of HV tend to lead the longer term levels of HV. This should come as no surprise as we’ve already asserted that shorter term HV moves quicker than longer term HV.

Assessing a longer term view of any length of historical volatility (10,30,90, etc...) causes one to see that it is mean reverting. Whereas a stock could trend up indefinitely, HV typically oscillates in a range around its average. For example, between 2004 and 2007, 30 day HV on the S&P 500 Index oscillated around 11, moving as high as 16 and as low as 6 (click to enlarge).

[Source: EduTrader]
We can reasonably expect that when HV gets too high or too low it will at some point return to its average. Unfortunately the “average” level perpetually changes, which means that just because 11 was the average between 2004 and 2007 doesn’t mean it will always be the average. As market conditions change, the norm for historical volatility will undoubtedly change also. For example, in the first chart, you can see current 30 day HV on the S&P is at 16. Compared to 2004-2007 that seems high, but compared to the last 6 months it actually seems quite low. Due to the bear market we’ve experienced volatility levels have been quite elevated through 2008 and 2009.