Retail sales dropped 0.4% versus last month’s +1.0% (revised down from 1.0%.)  In a market where bad economic news provides the Fed cover to extend QE, we have to wonder whether this is “good bad” news or “bad bad” news.

Surprisingly, the “adjusted” data on which most everyone focuses came in worse than the unadjusted data.   I guess even the bureaucrats wouldn’t mind letting a little air out of this market.

In an environment of flat to negative real income growth, retail sales obviously can’t continue to grow in real terms unless savings fall.  And, that scenario rarely ends well.   In a perfect world, falling retail sales would mean sentiment has probably backed off as well.  We’ll find out at 9:55am when Michigan data comes out.

SPX shed about 7 points in the opening minutes and bounced at the former neckline (red, dashed) of the IH&S that didn‘t play out on Mar 25.

With all the charting I do, I sometimes stumble across an important old pattern I’d forgotten about.  The way this neckline kept popping up in critical situations, I began to suspect it was somehow connected to something more important.

*  *  *

BTW, consumer sentiment just came in at 72.3 vs last months 78.6 and expectations of 76.  This is the lowest since last July, when slower job growth and higher food prices were taking their toll on expectations.

*  *  *

Getting back to that neckline…

It’s actually part of a channel system we’ve examined from time to time (in red, below.)

Seen here in a little more detail…

So, a drop back through it in a few minutes won’t be a minor event.

UPDATE:  11:00 AM

SPX is continuing to sell off nicely.  I’ve added the channels corresponding to the neckline (in red, below) back into to our charts.  It makes things a little busier, but I have a feeling we’ll be seeing more of this system.

I remain short from 1597 [yesterday’s 11:30 update.]   As usual, we’ll take a shot at divining where this correction might take us.

The technical elephant in the room is the previous 1576.09 high — now just 5 points below.  I wouldn’t be at all surprised if the market closed right there — forcing the decision as to whether or not to hold short into the weekend.

Unless 1576 is taken out, any correction will be viewed as a backtest of an important, previously exceeded level of resistance.

This is the exact same situation SPX faced in October 2007 when it exceeded the previous all-time high by 23 points (versus yesterday’s 21 point beat of the 2007 high.)

Of course, it had already had the good sense to back off the first time it approached the Mar 24, 2000 1552.87 high.  But, when it finally pushed through, it spent a total of 7 sessions above 1552.87, closing above it only 5 times before reversing for a 170 point decline over the next 7 weeks.

Are we in for the same kind of shellacking this go ’round?  Before reading any further, I strongly suggest you gather some important supplies:  a comfortable chair, a cold beverage of your choice, and a liberal amount of pixie dust.

continued for members…

There are many similarities in the patterns, but many differences, too.  Consider the RSI charts, for instance, which clearly show the impetus behind the current downturn.

The 60-min chart shows we’ve tagged the top of a three month old channel (purple) and the more recent red channel.  But, obviously there’s no negative divergence here.

The daily chart shows plenty of negative divergence: four successively higher price peaks with successively lower RSI peaks.  And, the top of the falling purple channel looks pretty solid.  It even features the same slope as the previous purple channel that led from the May 2011 high to the Oct 2011 low.

The weekly chart isn’t quite so clear.  The white RSI channel slopes downward. And, there has been considerable negative divergence relative to the other channel top tags (525, 621, 1163, 1344.)

But, zooming in a bit, we can see some stark differences with the 2007 peak.  In 2007, weekly RSI made three lower peaks in a row as prices surges pushed higher: 1431, 1522 and 1576.

While here in 2013, RSI has increased with prices (1292, 1422, 1474 and 1597.)

Zooming in further, we can see that the latest rise has more in common with the Feb 2011 1344 peak than the May 2011 1370 peak.  Note the succession of higher prices and higher RSI values until the white channel top tag at 1344.

This was the Feb 18 high, just prior to a 7% correction to 1249 (Mar 16.)   SPX bounced there and reached 1370 by May 2.

So, the lack of negative divergence correctly warned that, while a correction was coming, it would soon be followed by a higher high.  In other words, it wasn’t the “big one.”

A similar pattern occurred in the months leading up to the 1219.80 high in Apr 2010.  SPX, which had just tagged the .618 Fib of the 1576-666 crash, subsequently fell 209 points (17%.)  It took seven months and QE2 to get back to 1219.  Prices kept climbing all the way back to the aforementioned 1344 in Feb 2011.

So, why was the decline from 1219 so big compared to the decline from 1244?  First, it was the .618 Fib of the 1576-666 crash.  As such, it might have represented the full extent of the bounce prior to SPX moving lower than 666.  If QE2 hadn’t come along in the nick of time, who knows?

The decline from 1244, on the other hand, didn’t come at a major Fib level at all.  The next major Fib level was the .786 at 1381.50.  And many folks, including yours truly, fully expected SPX to reach that and complete a Gartley Pattern before the real fun began.

Also, unlike the 1344 top, the 1219 high didn’t quite reach the top of the white channel.  We see this quite often, when prices or RSI fail to reach a clear chart pattern objective (such as a channel top.)  The disappointment can yield more dramatic effects.

What’s it all mean?  At 1597, we’re much closer to the channel top. So, I don’t think we have to worry about the disappointment factor — especially after making a new all-time high.

And, the lack of negative divergence on the weekly RSI argues for still higher prices after this correction — whatever it ends up being.  The extent of the downside — from an RSI perspective — could be anywhere.

Some of the more suspect targets are shown on the chart below.  They could equate to anywhere from a 45 to a 200-pt decline.

The daily RSI, remember, is the one that shows plenty of negative divergence.  So, I’m inclined to believe we’ll see a sell-off beyond this morning’s lows.  But, with so much support below —  those Crab Pattern 1.618’s  at 1553 and 1555 for instance — SPX might fall only another 30 points before the weekly pattern takes over and it pushes higher.

If that should happen — and, here’s where you’ll want to have that pixie dust handy — we could see the 1972 analog play out.  Note the highlighted area, wherein prices whipsawed gradually higher after having exceeded the Nov 68 all-time high.  They finally topped out at 1.272 of the 36% drop between 1968 and 1970.

The equivalent point for 2013’s SPX would be 1823 — which could take as long as early 2015.  I’ll play more with the exact price levels and timing over the weekend.  But, unless more downside momentum builds over the coming week and some of the long-term channels start to break down, this is looking like a real possibility.

The key, of course, is those two little letters: Q & E.  If the Fed keeps injecting hundreds of billions of dollars into the banking system, banks might eventually start lending instead of hoarding those dollars gambling them away in the derivatives markets.

Eventually, there should come a tipping point where more/cheaper money only drives up prices — producing an inflationary environment like that of the 70’s.  They had an oil shock to contend with back then (we don’t….yet) but otherwise there are some interesting parallels such as the end of a war, cutbacks of government spending, etc.

Stay tuned.

The first step is to see how serious a pull back the market currently has in mind.  The bottom of the purple channel is currently around 1563 — only slightly higher than the .618 of the 1539 to 1597 rally at 1561.60 and the previous 1474-1343 Crab Pattern 1.618 of 1555.57 — which would represent a whopping 2.6% correction.

If those levels don’t hold, 1527 is on the table.  And, if that fails, 1497-1509.  If, on the other hand, the bounce of these past 30 minutes continues into Monday, there are two likely scenarios on the table:  (1) a backtest of the broken TL from the 2000/2007 highs currently at 1593.30 (the .786 of the last leg down is 1593.63); or, (2) new highs — starting with 1600 and seeking out resistance at higher channel lines (the top of the purple channel is currently 1630.)

Why no upside harmonic target?  There is no remaining Harmonic Pattern resistance other than the 1.272 at 1583.  Yep.  A strong move up and it could be a doozy.

More later.