Sell in May? UPDATED

The old adage is buy in Nov and sell in May, and maybe that is what we should now expect. The seasonal pattern is quite dependable, and Sy Harding reports that it is seen not only in the US but globally, and goes as far back as 1694. Still, it doesn't always happen. He also notes that we would be just as well off mathematically to sell 5% from the top as selling 5% after the peak, but often we are psychologically much worse off selling in a downside panic. Better to be wary near a top, and prepare to short. With this in mind, here are some weekend speculations. 

Thursday's Special STU noted an exhaustion gap, and as expected today was weak. Next week we should see a deeper pullback. Friday's STU stated that "there is no satisfying way to label the move up from the February 5 low, or the even shorter-term swings over the past several days, without compromising some of Elliott's guidelines." As I have been saying, the market has been at a moment of entropy (in a Chaos Theory sense) where it could go either way - hence hard to label. Still, we can always play around with what we do know, or think we know:

1. We seem due for a correction, as indicated by the exhaustion gap and a bunch of other oversold indications 
2. We don't seem done yet, and given we passed the 50% retrace, the next most likely level to hit is the 62% retrace, or Sp1229 
3. We have been in a 33 trading day pattern since Nov2, as noted by the STU Friday (see chart)

 

The 33 day turn pattern suggests a turn now, and another one in mid-May. The prior ones have been low to low; maybe this is high to high? We might see a pullback that lasts several weeks, then a final run up into May. 

Our challenge is to guesstimate the wave structure. Since Feb 5 the market has sallied forth up a tight channel, and so far has broken in three waves, suggesting either a zigzag or an impulse. A pullback now and a final surge later would fit a wave 4 correction and then a final fifth wave up into May. That makes this a five-waver up, which either kicks off something bigger (eg a much larger zigzag), or finishes wave C of a correction. One model is as follows:

• return to a prior wave count which made Nov2 as ending an X wave
• count the 3 waves up to the Jan high as A of a flat
• count the drop to Feb5 as B of the flat
• then we would be in a five-wave C to a slightly higher high than Jan

This "ending flat" pattern has been shown by several ewave sites, including this Wave Principle chart. updated from when first suggested, a week ago:

 

My prior weekend speculative count had been a top in the 1158-1170 range, which we have busted a bit above, but still close enough, so it could have topped a few days ago - and perhaps enough negative events are swirling around to drop the market, including passing ObamaCare, the jitters in bonds, the possible conflict in Korea, and the seeming bubble bursting in China, but too early to tell. 

The old one is still on, but tenuously. If we blow by it, the next stop is the 62% retrace of the whole Hope Rally at Sp1229, which turns out to be close to the 1240 level where wave C of this flat = 1.6x A, a common enough relationship. Let me propose a new one based on this speculative wave count of an "ending flat" to end a triple correction from Mar9: 

Sp1225-50 by the end of May.

Comments

[trendlines]

Hi Yelnick, good analysis. Coincidentally, i had a post by the same title last week, with a target around 1220, based on a long-term chart:

http://trendlines618.blogspot.com/2010/03/s-long-term-chart-sell-in-may.html

The depth and nature of this correction is crucial to determine what happens next.

[betterdays]

Anyone follow Mark Lebovit of VRtrader? He was rather coy on The Nightly Business Report Friday .. covering stocks
I got the impression he didn't want to divulge the fact that stocks are HEADING MUCH HIGHER , Even though his forecast model is CALLING FOR STOCKS TO END THE YEAR ON THE LOWS of the year .

[OracleLurker]

since all the fractal guys seem to have gone quiet (broke?) i have to ask - anyone else see the fractal similarity between where we are now and Oct '07? Normally if something like this was a hot topic of discussion i would say it only guarantees we're trucking higher but i don't think i've seen it mentioned anywhere - and that could indicate i'm seeing something that isn't there. Also Sp1210 is right at the July '08 lows and the middle of the Paulson induced madness of Sept '08 - i think that is the maximum possible extent of this move before a return to 1150.

[JT]

Full Moon on Monday!

[Wave Rust]

I have rarely seen an X wave I couldn't name as something else as time passed.

So, I just call them a 2 or a b ,,, but never an X.

wave rust

[Hank Wernicki]

the market has been at a moment of entropy ... yep

got a buy for XAU / GLD

got a buy for ORCL

Monday's tough, but I think the DOW should rally

[Wave Rust]

"perhaps enough negative events are swirling around to drop the market, including passing ObamaCare, the jitters in bonds, the possible conflict in Korea, and the seeming bubble bursting in China"

any drop now is a buy for all of those 'bad news' reasons. this correction however brief, is minuette for spx (or lower on some other indices). Since March 12th, this will be the 3rd 4th (or 'b', if you prefer), as the overlaps begin to unwind.

If the 5's don't overlap early on, they might complete the minute 3 of minor 3 of intermediate (3) of P1 a few days before 'big effing surprise' tax day. Then correct in the minute 4th, since all the big teenaged traders will be hung over from spring break. By Cinco de Mayo the run to new highs should be crunked up.

From my 5 legged stool, it's "a buy this low, and then, look away until the end of May".

long at easter, short memorial day, long 4th of july, then short all o' yo mama's trust fund by labor day.

be the headlight, not the deer caught in it.

never fade the fed, no matter how bad the auctions are. The worse they are, the better the set up is for being long bonds in July and august and september and october and ....... sell bonds in april next year.

predictions are so easy. it's the pulling trigggers thing without a plan that turns a headlight into a deer.

wave rust

[Michael]

Yelnick, how does this count figure in with your models? http://ewtrendsandcharts.blogspot.com/ the count is on the last two charts.
Michael :)

[joe]

Stupid Yelnick & ewavers, Market is going to blow your head

[Guy who likes BJs]

Market is going to blow your head

Sounds kinky. Sign me up.

[David]

i have posted an update for the asx 200 for any aussies out there :-)

http://www.tradeyourwayout.com/2010/03/asx-200-daily-chart-update-call-off.html

[Barood]

I developed a sentiment indicator that use quantitative methods and it is saying the top is in. Prices have diverged with sentiment for the first time since March 2009. It captured the 2007 top, and 2009 bottom and now it did exactly same divergence.

http://s62.photobucket.com/albums/h105/barood/?action=view&current=spy.png

from 2002 to 2007, it only diverged once with price in Oct 2007.
from 2007 to 2009, it did in march 2009.

I am adding an average deviation now for price and sentiment data series to sample data and will feed into SOM neural network to verify the data and deduce predictions.

I wanted to share with te this board to see if the image appeals to their eyes. I am becoming so thrilled about it after long hours of backtesting.

[grand]

nice alternate yelnick. that has been my preferred for a few weeks now.

http://waveprinciple.blogspot.com/2010/03/3510-spx-p2-incomplete.html

[yelnick]

grand, I like it - nice chart!

[yelnick]

Michael, looks like a very similar count to this ending flat concept. The chart from the comment by Grand shows it more clearly, and has a implied target at Sp1241 (where C= 1.6x A).

[trendlines]

Corrective action ahead, followed by one more high. Latest trendlines and analysis:

http://trendlines618.blogspot.com/2010/03/s-short-term-corrective-action-ahead.html

[David]

hi grand,

nice chart, that was going to be my alternate view

[Canadian Money]

The SP500 move up from the Feb. 5th low may qualify as a 5 wave count with an extended fifth wave. Elliott stressed channelling the movement to help one see the proper wave count. In this case, the movement off the Feb 5th low is now very close to an upper channel line taken off the peak of a possible wave 3 near Feb. 22.

[DG]

Since Feb 5 the market has sallied forth up a tight channel, and so far has broken in three waves, suggesting either a zigzag or an impulse.

Daily structure does not support either a Zigzag or an Impulse interpretation. None of the Fib relationships nor structure nor behavior logic are suitable for an Impulse. We could go higher, but not because of that.

[JT]

And the market continues to move higher, befuddling all of the E-Wavers who are unable to come up with a valid count since the early Feb. lows.

[yelnick]

JT, the ending flat count is valid, and supports a market run up towards 1230-1240

[DG]

JT, the ending flat count is valid, and supports a market run up towards 1230-1240

I gotta disagree, Yelnick.

The green B at the August low in that count is too short in time for anything but a Triangle.

Many of the red [i-v] waves are larger than the green B, which is a Degree issue, as we've recently discussed.

The blue (Y) is a Terminal, but doesn't end the entire pattern like a Terminal should.

The green B at the February low is too short in time for a Flat (needs to be almost the same time or up to 4X more time than wave-A)

The rally since the February low isn't an Impulse, which would be required for a Flat. Nor is it a Terminal, which could appear in the wave-C position of a Flat, to signal extreme weakness.

[upstart]

"Blow Your Head" - a tune by trombonist Fred Wesley and the J.B's from the 1970s. A piece of obscure sh-- for ya.

[David]

i have posted an spx update for those interested... market looking tired here

http://www.tradeyourwayout.com/2010/03/spx-hourly-chart-update-change-of-tone.html

[yelnick]

upstart - nice tune - you must be in a funk :-)

[yelnick]

DG, a couple of questions. You are putting me to the test of reading MEW over and over. When I do, I always wish it had been better organized, since things show up in multiple places. You could do all of us a great service by writing your version of Neely rules in a new book.

(1) about the purported ED that ends the Y wave into October: If the ED up to the Y top in Oct cannot be an ED (since B is smaller than the internal ED waves, and B is too small anyway), how can you then call it a Terminal? Terminal is Neely's parlance for an ED. We can disagree about whether an expanding ED exists - I don't want to revisit what I have laid out in four recent posts - but you do agree a Terminal with an expanding form exists, a fifth-wave extended one (or a third-wave extended one). If this is a Terminal, but doesn't end the pattern as a Terminal should, what is it?

(2) about the purported wave B of the ending flat, where in MEW does it say a flat B wave has to be > in time than the A? Not disagreeing with you, just don't have the patience to find it there. In any event, in orthodox ewave a shorter B is fine (in time).

(3) As to whether wave C is an impulse, it seems not difficult to count it as potentially a five wave impulse with a channel and an extended third wave. Wave 4 (which we may be in) needs to alternate with 2, which it may already is doing by being a longer sideways move. (I will stay on the sidelines on that, since wave 3 is tracking in the lower half of the trendlines and I wouldn't be surprised with a thrust to the upper trendline before wave 4.) Not sure why you would already conclude this falls outside of Neely's essential construction rules.

[DG]

(1) about the purported ED that ends the Y wave into October: If the ED up to the Y top in Oct cannot be an ED (since B is smaller than the internal ED waves, and B is too small anyway), how can you then call it a Terminal?

My apologies, Yelnick, I would have been more clear to say that it is labeled as if it is a Terminal, but because it doesn't end the entire pattern it can't be one. I do not mean to say that it is a Terminal. I was not entirely clear.

(2) about the purported wave B of the ending flat, where in MEW does it say a flat B wave has to be > in time than the A? Not disagreeing with you, just don't have the patience to find it there. In any event, in orthodox ewave a shorter B is fine (in time).

On MEW 11-11 where Neely defines the rules for a Common Flat and in this QOW http://www.neowave.com/qow/qow-archive-411.asp and this one http://www.neowave.com/qow/qow-archive-509.asp

"Question:
Since Triangles occur frequently, but are often misinterpreted, what clues exist (early on) to tell us a Triangle is, or is not, forming?
Answer:
Whether a Flat, Zigzag, Triangle, Diametric or Symmetrical is forming can only be determined (well in advance) by employing advanced NEoWave observations which focus on the price, time and complexity development between waves-A and B.

Under NEoWave rules, a Flat or Zigzag is possible only IF wave-B takes the same amount of time or more as wave-A (wave-B will usually consume much more time in Flats than it will in Zigzags). If wave-B takes less time than wave-A, a Triangle, Diametric or Symmetrical is forming."

(3) As to whether wave C is an impulse, it seems not difficult to count it as potentially a five wave impulse with a channel and an extended third wave. Wave 4 (which we may be in) needs to alternate with 2, which it may already is doing by being a longer sideways move. (I will stay on the sidelines on that, since wave 3 is tracking in the lower half of the trendlines and I wouldn't be surprised with a thrust to the upper trendline before wave 4.) Not sure why you would already conclude this falls outside of Neely's essential construction rules.

Check out this chart for some of my concerns. Also, check out MEW 12-11 and 12-12 for how Neely says Impulse waves should channel.

http://yfrog.com/0dspydaily3p

[yelnick]

DG, thanks. I think we are at the one of the crispest junctions between orthodox ewave and Neely. Neely must have moved his thinking beyond MEW to now say all corrections whose B waves are shorter in time than A have to be triangles, even if they start as a "5" in what orthodox ewave would see as a zigzag. How many triangles start as a "5"? None. So on a ZZ with a short B, NeoWave requires jumping through some hoops to turn a "5" into a "3" and count a triangle.

As to the ending flat formation, under NEoWave the ending flat may not fit, but it could under orthodox ewave.

We still have to figure out of the wave C is an impulse. Thanks for sharing the chart. Perhaps 3 is extending and is not as far along as your chart suggests. A smaller degree 4 may be on and out of it we should thrust up to the upper channel trendline. Then a larger 4 emerges. If this occurs, it will fool everyone, who are mostly expecting a correction or a run above 1200, not a sharp head fake up and then the drop down.

[DG]

Neely must have moved his thinking beyond MEW to now say all corrections whose B waves are shorter in time than A have to be triangles, even if they start as a "5" in what orthodox ewave would see as a zigzag. How many triangles start as a "5"? None. So on a ZZ with a short B, NeoWave requires jumping through some hoops to turn a "5" into a "3" and count a triangle.

Neely would probably say that you can't have a 5 followed by a 3 that takes less time than the 5 and, if you think you do, you need to recheck the structure of the 5. I haven't asked him this directly, but you'd have to deduce it from his rules.

We still have to figure out of the wave C is an impulse. Thanks for sharing the chart. Perhaps 3 is extending and is not as far along as your chart suggests.

I still think we are in a Neutral Triangle or Neutral Diametric from the February low. The Neutral Triangle is the one that looks most like an Impulse wave, especially from a Fib relationship perspective, since wave-C is typically 1.618X wave-A and waves A and E tend toward equality or a Fib relationship as well. A Triangle of all :3s also overcomes the issues I mentioned in the chart I linked. Even if the wave-3 of C was extending, so that my labeling would need to be redone to accommodate that, you still end up having time-related issues with waves 1 and 2 and you would also have the issue that wave-iv of 3 was larger than wave-2, creating a Degree problem. Overall, I just think it best to label the rally from the February lows as a :3. A lot of things have to fall into place for an Impulse to be "real".

I'm pretty agnostic on how far this rally could go. If my overall count is right, we're in the final segment of a Triple Combination and the upper end of the range for that pattern is 1300. And people say I am a perma-bear!