Great blog! I have been trading butterflies lately on individual stock names and trying to learn about them. I’m fairly new to options and love the added dimension of volatility. You mention in your blog that skew is important for flies since you’re short the body and long the wings. So a flat skew would benefit a fly. I used ATM vol to price the wings of a fly using various strikes out to see what flat vol price would be and how far off it is from current price. Then I just keep updating them in excel. Thing is it seems to be fairly consistent and don’t vary too much.
My question to you is, do you think that skew flatness offers an edge? I mean, how much does that translate into the price of the fly when it reverts back to the norm? Is it enough? I check on high price stuff like AAPL. Seems like at most 10-15 cents… which isn’t much on its own.
Also, how do you know what is flat and what is steep? I noticed you look at the vol/50 delta and vol/25 delta for example and those figures seem to be consistent across various asset classes. Is that how?
Thanks for reading!
Skew flatness does offer a few points edge. I read the following in Volatility Trading by Euan Sinclair (Chapter 3):
One way of determining the relative importance of different types of movement is principal component analysis (PCA). This is a mathematical technique used to reduce the dimensionality of data sets. For an introduction refer to Alexander (2001a). Derman and Kamal (1997) applied PCA to S&P 500 and Nikkei options. They examined the daily change in the volatility surface where it was parameterized by delta and time to maturity. Skiadopoulos, Hodges, and Clelow (2000) applied PCA to changes of S&P 500 implied volatilities for given maturity buckets, using both strike and moneyness metrics. Alexander (2001b) applied PCA to daily changes in the deviations of strike volatilities from the at-the-money (ATM) volatilities.
Alexander’s work is most directly useful to us. She showed that a parallel shift of the implied volatility smile accounted for between 65 and 80 percent of the total variation of volatility. A tilting of the curve explained a further 5 to 15 percent of the variation, and the curvature component explained about another 5 percent of the changes. So the most important thing to understand is the dynamics of the overall level of volatility, followed by the slope of the curve.
So from this excerpt, movements in overall implied volatility account for 65 – 80% of the variation in IV while skew accounts for 10% – 20% (slope & curvature). I have found that most of the time, skews are normal to flat but it is still important to be aware of the environment.
In order to see what is flat and what is steep, I download historical data from a Bloomberg terminal of 10 delta puts/calls and ATM puts/calls. I divide 10 delta put/50 delta put & 10 delta call/50 delta call. I check those numbers against historical averages to see if it is historically steep, average, or flat. Check this post out for a detailed explanation: SPX Put and Call Implied Volatility Skew In Perspective.
I hope I answered your questions, if not, let me know!