What Is 'Normal' Volatility in Options?

Includes: SPY, VXX, VXZ
by: Surly Trader

What is really a normal volatility level for the S&P 500? This is a very valuable question to address when thinking about buying or selling options. If implied volatility is low compared to historical volatility levels, then you might think about purchasing options. If implied volatility is high, then you might think about selling options. There is not a perfectly right answer in this realm, but one place to start is to look at recent historical volatilities versus current implied volatilities:

[Click all to enlarge]

Recent Historical Vol Bouncing from 3.8% to 20.9%

The recent rolling historical volatility graph shows a 10-day annualized low of 3.85% to a high of 20.94%. It has been my feeling that these volatility levels have been fairly low considering everything that is going on with global economics and global politics, but looking at this chart does not give a lot of support to that hypothesis.

One thing that you will note is that the implied volatility level, cyan line, has been significantly higher than the realized levels shown. This indicates that you would earn a positive risk premium in selling one month options over this time period. If you sold calls and puts at the money, and delta-hedged the options to a delta-neutral position, then you should have captured the implied volatility to realized volatility gap.

Getting back to the original topic: How can we make a judgment on whether current volatility levels are high or low? I like to look at volatility cones. In the volatility cone below, I plotted annualized volatility distributions going back to January 2001.

The way this works is simple: I start off calculating 10-day volatility for every 10-day period between January 2001 and May 2011, then look at the distribution of those annualized results. I move on to 20-day volatility and do the same until I finish at 360-day volatility. What results is a cone shape that shows that annualized volatility ranges can be significantly low or high over 10-day time periods, but the ranges get smaller as you move out to one-year time periods.

Volatility cones get their name from their shape.

The numbers that are plotted below the green line represent the 50th percentiles of the distributions. For 10-day historical volatility, we see a 50th percentile of 15.78%. For a 200-day volatility, we see a 50th percentile of 18.69%. Notice that these are not averages, but medians. The average would be brought higher by the skewed distribution.

So how would I interpret these specific results? Looking at the last 10 years of S&P data and specifically at the 100-day volatility cone level, I would say that the current 100-day realized of 11.69% is on the low side of historical volatility levels.

Disclosure: I am long SPY.

Additional disclosure: I am short SPY via short positions in SPY put options