Volatility selling strategies have performed better than one would expect through the 2008 tsunami. These strategies relate well to my previous article “Picking up Nickels in front of a Steam Roller“. The first point that they make is that up until 2008, the markets since 1990 have been relatively calm and there has been a consistent option premium that has rewarded option sellers via a positive gap between option implied volatility and actual realized volatility.

The second part of the piece focused on the relative performance of different volatility strategies. They explored a 95-105 strangle (sell 5% out of the money put and 5% out of the money call), ATM Straddle (sell a put and a call at the money), Delta Hedged Straddle (sell ATM Straddle and delta hedge the straddle daily), and a recently popular Variance swap which basically pays the difference between realized and implied volatility based upon the swaps terms.

The first thing that stands out is that a delta hedged strategy has the highest information ratio, just think of this as the return per unit risk (standard deviation). The second most notable issues is the face that the variance swap wiped out over 5 years of gains due to 2008. The main reason for this was the incorrect pricing of the convexity premium in the variance swap, which is a fancy way of saying that the distribution of volatility was much more fat tailed than expected. On the opposite side I was very surprised by the outcome from a Naked ATM Straddle. Its rebound from the hits it took during the end of 2008 was very fast due to the mean reversion type returns of the S&P 500. In the past I had been leaning towards a 95-105 strategy, but this certainly adds some validity to the ATM strategy for those who are willing to take the added volatility of returns.

Can you provide a link to the source document from Barclays? Or tell me the title and I’ll try to find it on Bloomberg.

Thanks.

Randy –

I am replying via e-mail.

ST

Sorry I’m late to the party. You have a great blog.

Do you have more details of these strategies?

On the face of it a simple straddle writing system seems like a winner,

Surely, it’s too good to be true?

Steve

Hello,

interesting article!

What expiration was used when testing the strangle/straddle? Always the front month?

And how many contracts were traded? Always a one contract strangle/straddle?

Thanks an regards

May I ask where is the best information on how to best delta hedge the straddles?

Thanks so much,

Joseph

I don’t know if you still read comments here, given it’s an old post, but I just saw it for the first time (linked from a newer post of yours). So, my question is, how would selling a delta-flat strangle compare? That is, selling for example the 25 delta put and the 25 delta call, whatever the strikes are at any given time. This probably means something like 93-94% puts and 103-104% calls or thereabouts.

I tried to replicate the ATM straddle, and OTM strangle back-test using both weekly and monthly options. My data is from Bloomberg going back to 1994. I had to make a few assumptions about the weeklies and the OTMs because the Bloomberg data only has ATM implied on monthlies (it has additionals from 2005 onward). But I am not getting such rosy results. The above chart implies a 11% annualized return, but calibrated to market data, I am showing a 3-4% annualized with an IR of .4. Also, when I compared the implied to VIX, I am showing a premium of around 3%, and if comparing the implied to 1 month forward historical, the premium drops to 1%. Could you also send me the source document so I can find out if I am missing something here? Thanks.