Quitting day job to collect weekly premiums - realistic?

[ironchef]

Selling premium on individual stocks requires stock picking skills and market timing skills. Doing it relentlessly and mindlessly will not work in the long run. Larry McMillan has said that selling covered calls this way works out to a T-bill-like rate of return. My few backtests, although not extensive, indicate this view is correct. He suggests using some simple technical analysis to avoid bad times and bad stocks (stock above the 20-day MA is the example he used, although it may not be the best approach).

I can support your statements based on my own experience.

Mechanically selling covered calls in 2013 resulted in very poor returns (compared to buy and hold). For six months I sold hundreds of covered calls, lost my stock holdings and then had to buy them back. To add insult to injury, I had to pay capital gain tax. I now do covered only on occasions and it is profitable that way.

 

[sle]

When you say "works out to T-bill returns" do you mean it against the notional? Also, I have question the results of that back test - it would imply that there is no structural risk premium in SS vol which sounds very unusual.

 

[ironchef]

When you say "works out to T-bill returns" do you mean it against the notional? Also, I have question the results of that back test - it would imply that there is no structural risk premium in SS vol which sounds very unusual.

I think the best answer came from CBOE (BXM=BuyWrite index):

There seemed to be no "outsize" returns either by selling puts or selling calls compared to buy and hold. And the results were before commissions and slippages so as a mom and pop retail trader, I did a lot worst net commissions and slippages.

 

[sle]

Outsized returns - of course no, but there is obvious excess return over SPTR due to risk premium. Selling straddles shows it the clearest, IMHO (just average results from call and put selling). Also obviously you pay for that excess return via higher 3-4 moments of your profile - scary drawdowns. But this is for index vol which is a different beast - I was asking specifically about SS vol and T-bill statement.

PS. You have to pay a lot of scrape to lose that excess edge BTW unless you are really getting raped.

PPS. Not persuading anyone to go and do any of these strategies, just saying that presenting them as "no edge and then you die" is just as silly as saying "it's guaranteed income"

 

[drcha]

We need to look over a lot of years.

2013 was mentioned as a case in point-a very bullish time. And backtests up to 5/31/07 are posted (geez, could the CBOE cherry-pick any better?!). While helpful, neither of these time periods should be considered a complete view of what may happen to the premium seller.

Finally, beating the S&P may not be the ideal goalpost. If you can match the average return with less volatility, you are essentially beating the index.

 

[sle]

A true vol selling back test - sell 1 month variance (aka VIX) since 1991 when VIX data starts. It shows that capturing risk premium is profitable but scary business. I've done similar tests on monthly straddles (delta hedged since 1996) and it has a similar profile, with slightly lower sharpe and smaller draw downs.

Now, what most people here are talking about is not the excess return but their personal utility function wrt to that return and risk. That's a different conversation all together.

 

[drcha]

SLE, so true. Tests are one thing, and the gut check is another.

 

[sle]

SLE, so true. Tests are one thing, and the gut check is another.

Yup. Best example is speeding while commuting - while the time improvement is measurable but small, the results of failure can be catastrophic yet it's a relatively rare event. Each person picks his own spot on that utility curve and acts accordingly.

 

[ironchef]

Now, what most people here are talking about is not the excess return but their personal utility function wrt to that return and risk. That's a different conversation all together.

Your posts are very insightful. Thank you.

Here is a one person example:

My investment account is 100% in individual stocks and I judge my success/failure against SPY.

Second, I only trade options on those stocks. I judge my trading success/failure against the returns of the individual stocks.

Frankly I don't know how to deal with risk profile, sharpe ratios... etc. so I simply ignore them figuring that over a long period total return is what counts? I am probably wrong so do you have any suggestions for me?

Thanks.

 

[sle]

I think intuitively you already doing some risk-adjustment so it's just a question of formalizing that process.