PEBBLEWRITER.COM IS LIVE!



*************** PEBBLEWRITER.COM IS LIVE! TO SIGN UP, CLICK HERE ***************


Sunday, May 22, 2011

Does OPEX Matter?

This past Friday morning, with SPX down about 9 1/2, I made a crazy call based on some indicators I've been watching [more later tonight or tomorrow.]  They said the market would not only stop falling, but would completely reverse itself and open up, leaving a nice bullish hammer going into the weekend and thus, portending a big up day Monday.

Indeed, SPX fell a little more, then made a spectacular comeback with well-formed 5-3 patterns all the way... until an hour and 15 minutes before the close.  When THEY screwed with my plan. You know who I mean.  The evil OPEX manipulators -- thrashing my portfolio (and worse,  my pride) for the umpteenth time.

Is it me, or are they really out to get us?  I did a quick little bit of research, and found out... "maybe."  I studied the last 42 OPEX cycles since 11/16/07, dividing them into positive (21), negative (13) and neutral (8) cycles based solely on net price movement during the cycle.  I then compared performance on the day of option expiration with the day after expiration for each cycle.  The results are intriguing.

  • When the OPEX cycle was up, OPEX day was positive 62% of the time.  When the cycle was negative, OPEX day was also positive 69% of the time.  Neutral cycles were even.
  • In positive OPEX cycles, OPEX was usually followed by a reversal -- regardless of whether OPEX day was positive or negative.  Negative OPEX days produced positive reversals 100% of the time; positive OPEX days produced negative reversals 62% of the time.
  • In negative OPEX cycles, negative OPEX days were always followed by another negative day (100%), while positive OPEX days usually produced negative reversals (67%).  
  • When the cycle was neutral, the day after OPEX reversed 75% of the time, regardless of whether OPEX was up or down.
  • In general, negative OPEX days were followed by positive days (69%), unless in the midst of a negative OPEX cycle.  And, positive OPEX days were followed by negative days (69%).
I also looked at the size of the day to day moves.   In general, most negative OPEX days produced similarly-sized positive reversals -- regardless of the cycle trend.  But, positive days in positive OPEX cycles produced outsized (2.14X) reversals, as did positive days in neutral OPEX cycles 1.63X).


There are limitations to a study like this.  For one thing, I generalized when characterizing the OPEX cycle trend.  Many would say the recent trend has been down, although SPX rose between 4/15 and 5/20 -- thus qualifying it as an up cycle.  Also, there wasn't a single instance of a positive cycle with a negative OPEX day until after the market bottomed in Mar '09 (and, yes, 5 of the 16 cycles during the bear decline were positive.)

I also didn't measure the gain or loss on the days studied, but focused instead on the total range and whether SPX was positive or negative on the day.  So, a big decline that reversed to close slightly positive was an up day; if it closed slightly negative it was a down day.

There's no particular justification for such methodology; it simply fit my desire to identify what trading opportunities, if any, OPEX might produce.  Has it?

This past cycle was positive (although, again, open to interpretation) and Friday was down 10.33.  Despite what the futures currently indicate (down 5.50), the model would suggest an up day with a range of 11 1/2 or so.   On the actual 8 similar days (down OPEX in up cycle), the 5 bigger OPEX declines averaged 10.6 and were followed by 12.1 point rallies (smallest was 5 points.)  The smaller 3 declines averaged .8 points and were followed by average 10.3 point gains (smallest was 7 points), so size of the OPEX decline didn't seem to matter much.

Also, the 3 days following the reversals were mostly down (5 of 8), although the trend for the remainder of the next OPEX cycle was up (5 of 8 times.)

Do I know any more than before I sat down to study all this?  Perhaps.  One of my favorite analysts put a huge red candle on one of his SPX charts for Monday that, I'll admit, scared the crap out of me because it wasn't at all what I expected (he put a bullish alternative up, too.)  Although medium and long-term bearish, I'm one of those looking for one last push up.

Regardless of what Monday brings, it's pretty clear to me that OPEX does matter.  The preponderance of reversals indicates to me that OPEX moves are "unnatural" in the sense that the subsequent market action "undoes" what the market makers did.  As someone always on the lookout for contrarian opportunities, this is useful information.

In the future, I'll look for trading opportunities on reversals (especially in up markets) and have a better idea what to expect.  And, when I'm itching to capitalize on the action unfolding in that last hour on an OPEX Friday, I'll take a quick look at the historical data.  It could be that THEY really are out to get us.

*******************************************

NOTE:  Monday, May 23  1:15am

The futures are down 6.50 to 1321.25.  If we were to close down tomorrow, it would be only the 6th time in over 5 years when a negative OPEX Friday was followed by a negative Monday.  Three of the previous five were in the midst of the 2008-2009 bear markets.

1 comment:

  1. Hello Pebble Writer,

    Hope this cheers you up..I still believe we are not done with the bull market yet and it is quite easy to get off track with the results of a day's close..Monday Open will be weak but how weak will depend on how strong 1321 - 1325 levels holds up. If that level holds up, I believe there will be a rally to close the gap in the mid 1350's. I would suggest keeping a stop below these levels or break even until clarity emerges. If 1318 breaks, there are targets all over the place until 1274 so I would exercise caution holding longs..I did some studies on fractals and harmonic trading and did find some good results. I will share them with you once I do some back testing...Have a good weekend

    Vikram

    ReplyDelete