What's the Best Way to Sell Premium?

Matt Moran of the Chicago Board Options Exchange (CBOE) highlights an interesting recent study on their weekly and monthly PutWrite indexes:

"Written by Oleg Bondarenko, professor of finance at the University of Illinois at Chicago, and sponsored by CBOE, the study -- 'An Analysis of Index Option Writing with Monthly and Weekly Rollover' -- analyzes the performance of the two PutWrite Indexes through the end of 2015."

Just to review, the CBOE S&P 500 PutWrite Index (PUT) measures the return of a fixed strategy of selling one-month at-the-money (ATM) puts on the S&P 500 Index (SPX) and then rolling each month. The CBOE S&P 500 One-Week PutWrite Index (WPUT) does the same, but for weekly options. The quantity of puts "sold" varies, but is set up such that there's enough theoretical cash on hand to cover max losses.

Intuitively, such a strategy should do better (or really, less bad) in a market decline, and underperform in a rally. It should also outperform in small rallies and general churns. And it should have less volatility than simply owning the index.

That always sounds odd. Less volatility via a strategy that literally sells volatility? From a portfolio perspective, that's true if you think about it. You're muting the gains and you're limiting the losses. Anyway, what did the paper find? Here are their highlights.

"Long-term performance. Over an almost 30 year period, the PUT Index outperformed the traditional indices on a risk-adjusted basis. The annual compound return of the PUT Index is 10.13%, compared to 9.85% for the S&P 500 index. However, the standard deviation of the PUT Index is substantially lower, 10.16% versus 15.26%. As a result, the annualized Sharpe ratio is 0.67 for the PUT Index and 0.47 for the S&P 500."

OK, makes sense. Pretty similar returns, but with less volatility. That's what you get with buy-writes also, and put sales are essentially the same thing.

"WPUT Index and PUT Index over recent history. The data history for the WPUT Index begins in January 2006. Over the last 10 years, the PUT and WPUT indices delivered similar risk adjusted performance and both outperformed the S&P 500 index and other benchmarks. The annual compound return is 6.59% (PUT), 5.61% (WPUT), and 7.09% (S&P 500). The annualized Sharpe ratio is 0.52 (PUT), 0.50 (WPUT), and 0.46 (S&P 500)."

I'm actually surprised PUT and WPUT underperformed over the past decade. We've really seen a little bit of everything. PUT and WPUT should have done relatively well through the 2008 implosion, underperformed at times from 2009-2014, then outperformed again in the churn and decline since then. Perhaps the devil is in the details.

"Lower risk. Relative to the PUT and S&P 500 indices, over the last 10 years, the WPUT Index has lower standard deviation, beta with respect to the market, and maximum drawdown. In particular, the standard deviation is 11.51% (PUT), 9.85% (WPUT), and 15.11% (S&P 500). The maximum drawdown is -32.7% (PUT), -24.2% (WPUT), and -50.9% (S&P 500). The longest drawdown is 29, 19, and 52 months, respectively."

Well, this part does make sense. Again, rolling premium sales will mute the extremes and outperform in the middle.

"Annual premium income. From 2006 to 2015, the average annual gross premium collected is 24.1% for PUT and 39.3% for WPUT. Premiums for WPUT are smaller, but collected weekly instead of monthly, which results in higher aggregate premiums."

It's interesting that you generate that much more premium if you roll weekly vs. rolling monthly. It sounds great, except the actual returns from the weekly strategy trail the monthly strategy. That's mainly because the higher dollars come with higher gamma and vega that you have to theoretically defend.

Think of a really extreme case: rolling one-day ATM puts. You would clearly generate the most premium possible, as whatever you sell is pure premium that will decay to zero by the end of the session. But it comes with huge gamma. If the index declines a bit, you might lose 100%, 200%, etc. that day. More premium does not necessarily equal greater returns, as it probably comes with more risk.

"Liquidity. Trading volume in Weekly S&P 500 options has increased dramatically over the last 5 years. In 2015, on average it was about 340 thousand contracts per day, representing 36% of all CBOE S&P 500 options."

On the margins, this adds to overall market volatility. It implies the average duration of open options has declined over time. The shorter the duration, the more exposure to sudden moves and the more exposure, the more forced covering of said moves, et al.

All in all, I don't find all that much shocking info here. I like the concept of all these indexes, but I did always wonder whether a premium selling strategy did better rolling longer- or shorter-term options. The answer is that they each have their pluses and minuses, and personal preference should swing the decision more than anything else.

Disclaimer: Mr. Warner's opinions expressed above do not necessarily represent the views of Schaeffer's Investment Research