21 October 2009

Selecting Condor Strikes - A Study

As I have outlined in another thread, people who trade iron condors are usually taught to select short strikes at one standard deviation from the current price action.

As I am developing a conversation elsewhere, I think extrapolating option volatility to "monthized" levels is not optimal.

What I want to do in this thread is select hypothetical strike levels, assuming that an iron condor is put on on the expiry Friday of the previous cycle. Some people like to initiate ICs with 2 months to expiry, so we'll have a look at that too.

I'll be using historic volatility, just so it can be automated.

I'll report on how these strikes are threatened or not as time goes by, to test how well this selection process works.

Note that this does not represent any trades I may be doing, it is just an exercise.

On the chart the red lines are 1 standard deviation (68% theoretical probability of staying inside).

The Blue lines are 1.28 SDs (68% theoretical probability of staying inside), which you would probably never get enough premium to trade.

The Green lines are at 0.6 SDs and represent what a low probabilty condor trader might select. It is presumed this trader is prepared to adjust more aggressively.

The top chart with heavier dotted lines is to represent a trade initiated 16 Oct with Nov expiry. (1 Month)

The 2nd chart with lighter dashed lines is to represent a trade initiated 17 Sept with Nov expiry. (2 months)

The 3rd chart lighter dashed lines is to represent a trade initiated 15 Oct with Dec expiry. (2 months)

I have also posted this at http://www.internationalstockforums.com/index.php/topic,29.msg64.html#msg64 That will be an easier place to follow allong as time goes by. Comments most welcome there.

Click to expand image.

2 comments:

Mark Wolfinger said...

"people who trade iron condors are usually taught to select short strikes at one standard deviation from the current price "

I don't accept that premise. i believe it's false.

Wayne said...

Yep, agree.

That's what I'm trying to prove as time goes by; hence my comment - "I think extrapolating option volatility to 'monthized' levels is not optimal."

Cheers