We ran a large backtest to identify the best maturity, delta, call value as a percent of stock price, earnings strategy, and implied volatility profile for call selling on S&P 500 components. We found some interesting results. Symbols used in the test were MSFT, AAPL, AMZN, PFE, JNJ, T, GOOG, BRK_B, JPM, V, PG, UNH, INTC, VZ, HD, MA making up 30% of the S&P 500 weighting. Days to expiration (DTE) used were 10, 15, 20, 30, 45, 60, 90, 120, 180, 240 days to expiration. Deltas tested were 5 to 50 by 5 deltas. IV Percentile and Skew/slope Percentile tested were under 33%, between 33% and 66%, under 66%, over 33% and over 66%. Call market price as a percent of stock price tested were 0.5%, 1%, 1.5%, 2.0% and 2.5% and no filter. We call this Spread Yield%. Exiting after 75% profit is made and no early exit were tested. Avoiding earnings by trading out before the announcement was tested along with trading through earnings. In all, there were 3.5 million combinations of the above parameters run back through 2007 and here are the results. Results The best maturity was 30 days to expiration and best delta was 0.10. Avoiding earnings was better than trading through earnings. Holding to expiration was better than exiting after profit. A call value of a percent of stock price that was best was 2.0%. The best use of IV percentile and slope was selling a call when the IV percentile was above 66%. Slope percentile was not important for identifying the best calls to sell. Implementation Scanning for the above parameters is easy using the ORATS web tool. Trading through our brokerage APIs such as Tradier or Thinkorswim is made possible. Monitoring fills can be done to see if the market goes against the trade to cancel is available. Paper trading is available through Tradier brokerage. Position tracking and exit alerts along with delta and profit can be found in our Risk web tool. ORATS has backtested many parameters for identifying covered calls. The best are described here. Using web tools, ORATS makes it easy to implement the trading of these calls. Please contact us at 312.986.1060 or support@orats.com to start. More information and blogs here
Hi Steve Ophir does a nice job. We have a more intricate backtester- with data going back to 2007, scanner and live market greeks and data. We concentrate on proprietary options calculations which help us with trading triggers and accurate vols. We are past market makers, been in business since 2001, and have clients with over a trillion of AUM. Try it out 14-days free: https://info.orats.com/elitetrader Matt
^^ thanks for posting this. does the DTE take into account the shorts calendar? since there are several weekly option expiration dates that can be between the shorts calendar, did you see any irregular changes in IV or premium percentage during a mid month or end of month option expiry dates? for example most all of the companies you mentioned have weekly options, so could the DTE be skewed?
The days to expiration were tested with min max ranges. For example, the 30 days to ex had a min of 20 and a max of 50. Each time the call was rolled, the new expiration would be within that range. The IV percentile was observed at each trade.
Hi Matt,nice simultion.Like that you can run the backtest over a portfolio as opposed to a single stock.Looking at the results,it certainly makes intuitive sense.If you can get 2% of spot for a 10 Delta call,thats really high vol. And in a runaway Bull market( other than 2008 and the 2020 V correction),it makes alot more sense to sell the 90 delta put as opposed to the 50 delta put.. My question to you is it a fair backtest to compare the profitability of a 90 delta put vs a 45 delta put?? I would like to see "equal deltas" compared,i.e. 2 45 delta puts vs 1 90 delta put.Is that not closer to apples to apples? Thoughts??
Hi Tao, good question. First, I should mention that the 2% option price divided by stock price has a range around it quite wide. As you point out, a 2% is quite high. Most of the yields are lower in the range. You are asking if it is fair to sell a lower delta in a bull environment. You are converting a sell call + stock to a sell put and asking if it is preferable to compare the two -- 2x 45 delta puts vs 1x 90 delta put-- on an equivalent delta. 90 delta short put is equivalent to a short 10 delta call and long stock. A 45 delta short put is equivalent to a short 55 delta call and long stock. However, since there is discrepancy in the delta, to be fair you should sell 2 45 delta puts, not knowing we'd be in a bull market and of course the longer delta does better in that environment. I see your point. The issue many investors have is they own the shares and consider selling calls against them. What I like about the backtest is that there is a strategy that works even in a poor environment for call selling. This strategy is still additive whereas higher delta calls were a drain. You suggest that to have a fair comparison, you would give the investor selling the 55 delta calls double the long stock and sell double the amount of calls. These investors would have outperformed but since they are short more units they had more risk so I don't think that would be a fair comparison.
Thank you for sharing. Posts like yours are the reasons why I spent so much time on ET going through and reading all those posts. I ran similar backtests using SPY historical data from 1993 to 2019. I wrote my own BSM code after I learned VBA excel. I calculated HV from STDEV of SPY to simulate IV and calculate call options from them because I don't have historical option data. I also ran some fixed IV backtests. It pretty much supported your results: Best profitable covered calls were around 30 days +/-, with slightly OTM calls. I appreciate your post because you validated that the codes I wrote are probably correct. I also know it is a lot of work (at least for me). Best regards,
If you look at his graph,the covered call strategy performed nearly as well as MSFT stock..Didnt give up much in terms of performance..In a down market,vol picks up and you could probably get close to 2 percent for the calls..