It’s here a little early, but this morning should see the end of the rally from 1560 on Jun 24.  Look for the market to top out at the .886 Fib retracement of the 1687 to 1560 decline to complete a Bat Pattern.  This is the higher of the two targets we’ve had in our sights since May 21 [see: If It’s Tuesday] — which was the last time we saw a sharp rally 17-point ramp job on Fed comments that were decidedly not bullish.

The original forecast, shown below, called for a much quicker return to this level from a low of 1600, followed by a decline to 1560.  Instead, we are reaching the interim high after the 1560 decline.  Time will tell how this translates in terms of waves.  But, we should get a sharp pullback regardless.

For anyone short from yesterday, expecting a pullback before this final thrust like I was, the options are to hold and wait it out or play the upside and re-short.  I’ll do the latter.

UPDATE:  9:44 AM

SPX has come to a screeching halt at 1670-1671.  We could still see another point or two, but no doubt there are others who are wondering the same thing and won’t wait to short.

I’m pulling the plug, will revert to a full short position here at 1670.57 w/ tight stops.

Note the dashed line SPX tagged just now at Point D.  This is one of the three trend lines from 1994/2002 we’ve been tracking (445.45 on 12/8/94 and 776.76 on 10/9/02.)  Its yellow cousin is just above at 1685, and the red version is down at 1619.

Note the consolidation that has occurred around each — validating their importance.  A sustained move above the yellow one or below the red one will tell us much about future prices.

Also, an update to yesterday’s ramp job tally: this morning marks the 11th ramp job for 10 or more points since the 1687 high on May 22.  The S&P 500 might be down 16 points since then, but without those ramp jobs it would have been far, far worse.

My back of the envelope calculations:

UPDATE: 2:36 PM

The eminis just completed its own Bat Pattern, so SPX should be free to reverse now.

Obviously, Bat Patterns don’t always produce meaningful reversals.  That’s why we use stops.

But, the nice thing about them is that they occur at the .886 Fib.  So, if they bust, a double top — another shot at a reversal — is only 11.4% (of the decline) away.  In this case, if 1672.72 doesn’t produce a reversal, the previous top of 1687.18 is only 15 points away.

I’ve spent the last several hours looking at the 12 significant Bat Pattern over the past six years. I limited the study to those of 95 points or more which involved a reversal off a decline rather than a rally.  Both are legit; I just wanted to see what recent history said about situations like the current one.

To review, a Bat Pattern involves a drop from one significant point to another.  They should be at least meaningful interim highs/lows, though this is obviously subject to interpretation. They are shaped like a big “W,” with the height of the right side only 88.6% of the height of the left (from the very bottom.)

The trick with Bat Patterns is the Point B.  It must come in at less than 61.8% of the initial drop from X to A.  If it’s greater than or equal to .618, you’re probably looking at a Crab Pattern which finishes at the 1.618 extension (or more.)

One common problem is deciding which reversal to label as Point B.  Here, there are no absolutes; but, my practice is to choose the greatest reversal to date.  In the current case, that came at 1626 on Jul 1 at roughly the .500 Fib.  It was only 22 points, but there are no other candidates as large since the 1560 bottom.

BTW, we could have chosen the Jun 18 high of 1654 as the origin.  The 1626 reversal would be at its .707 Fib, which means it could not be a Bat Pattern.  The pattern from 1687 to 1560 qualifies perfectly.

I’ll publish the actual data after the close. But, the bottom line is that of the 4 patterns that actually qualified (B<.618), every one of them posted a reversal within a few points of the .886.  The extent of the reversals:  44 points (2/11/11), 61 points (5/2/11), 106 points (8/31/11), and 130 points (10/11/07.)

The last one is the biggest pattern on the chart that represents the 88.6% retracement of the drop from 1576 to 666 that began on 10/11/07 and finished on 9/14/12.

The Bat Patterns that “busted” weren’t really Bat Patterns.  They featured either a Point B that exceeded (even slightly) the .618 Fib or a much bigger reversal at, say, the .786 Fib (indicating a probable Butterfly Pattern.)

But, of those that did “bust,” a reversal at the double top occurred in almost every case — typically down to the .886 or more.  The result of this, of course, was another chance to break even.  So, even someone completely asleep at the switch, who shorted at the .886 but did nothing as it climbed up to the 1.000 typically saw a reversal back to their entry point.

As I wrote back in December 2012, as SPX was leaking up past its .886 retracement of 1474-1343, the gap between the .886 and a double top is the most challenging area for harmonics traders.

SPX tagged 1676.60 a moment ago.  Is it a slight overshoot (not uncommon) of the .886?  Or, is it a double top, instead?  If it tops 1687, will it stop at a double top, or is there more upside to come?

Given the acceptable form of the price movement up from 1560 and the channels at play, I’ll stick by the Bat Pattern call.  If it closes at 1677 and they ramp it up again tomorrow, my downside is probably another 10 points, which I should be able to recoup in the likely event it reverses at 1687.

I’ll take a quick break after the close, and be back within the hour to post the study data as well as a revised forecast for the next few months.

continued for members…

The study was a little more involved than I expected!  Eight hours later, I have results — along with some great data on Bat Patterns.  I looked at patterns with a move from top to bottom of 95+ points.

The Summary:

• 14 previous large Bat Patterns — or at least what looked like one — since Oct 2007
• 4 met the precise definition: most significant reversal at B < .618
• of the 14, avg .886 overshoot was 0.13% (range:  -0.6 to 0.5%)
• of the 4, avg .886 overshoot was .38% (range: 0 to 0.9%)
• of the 14, avg reaction at .886 was -2.1% (range: -0.4 to -5.1%)
• of the 4, avg reaction was -2.5% (range: -0.8 to -3.6%)
• 13 of 14 provided a 2nd chance, reacting at the previous high to below the .886
•  3 of the 4 did the same, average retrace at double top = -4.5%

The best performing Bat Patterns (and imitators) were well-formed, with a clear Point B that was less than the .618 and no other significant reversals at other points.  For instance, a large .786 or .886 reversal typically resulted in a Butterfly or Crab Pattern.

As a group, their reaction at the .886 relative to the double tops was greater than the imitators.  And, their reactions at both the .886 and double top was significant relative to the imposters.

Given that there are no other Points B in the current pattern and SPX’s top today was only 0.14% greater than the .886, the odds still favor a reaction here that, statistically speaking, should be about 40 points over the course of 8 sessions, followed by a rally to 1687 and a larger reaction of about 50 points.

Just for grins, let’s put a pin in a reversal from current prices to 1634 in the next week or so.  This would be followed by a return to 1687 around Aug 5 for a reversal to 1638 in mid August and rally up to the 1.272 of 1823 by December.

Obviously, this scenario would be undone by any move down through 1560 or up through 1687 without first reacting.  But, the odds favor a return to test these levels if SPX should blow through the .886 here.

This more or less jibes with my expectations — though I’d not be surprised if the path to 1823 were mostly without deviations given TPTB’s recent intolerance for even the slightest dips.  Nor would I be surprised if the entire rally came unraveled in one fell swoop for no good reason (other than investors coming to their senses.)

GLTA.