Monday, September 26, 2011

Weekend Update for September 23rd

I frequently get asked why I'm no longer blogging daily - the answer is simply you're not reading the right blog.  You can find me daily at in the GGT Forum, which I manage.  I'm only blogging here on the weekends so that my thoughts and process are available to non-members of the EV forum.

Please take the time to join me there.  There is content far beyond what I produce and the quality of the individuals there is some of the best you will find on the 'net.  Indeed, I know I lower the average IQ of contributors on the site ...


Last week was a fairly good week for me in terms of trading with my trading accounts.  Note there is no long-term bullish trend, and the bears are still in control.

What seems to be working for me right now is staying away from stocks and focusing on the SPX, the NDX, and the Russell 2K.  The broader ETFs which focus on these exchanges are far more forgiving than stocks right now, the latter of which are not giving me great signs on the long side.

I've been tying my movements to the performance of the $TICK indicator, which shows the number of stocks on the NYSE moving up in price over some period of time compared to those moving down.  Thank you Billy @ the EV site for introducing this to me (as I said above, the caliber of people there is unsurpassed).  When we accumulate, or add these values to the previous interval's value on an on-going basis, we get a good view of a trend.  For more on this start here.

Here's a presentation that paints the picture of the cumulative $TICK on the NYSE:

As with all my images, right click on the picture to open in a new window or tab.

The top presentation is of the Vanguard Total Index, VTI, which gives a good, broad view of the entire market.

The middle pane is a daily-resetting cumulative tick -- this one starts off at 0 each day and if the daily trend is up, we'll see it move upwards, and if the daily trend is down, we'll see it move down.  There's a catch to the middle pane -- it has a 500 stock/min filter on it, so only when the number of stocks is greater than 500 in either direction will the accumulator change.  This allows the noise to be rejected and for me to only view the trends that are broader.

The bottom pane is the cumulative $TICK through the end of Friday, and the white trace is the instantaneous cumulative value and the varying colors of lighter to darker purple are moving averages on the white value.

A couple of things should be evident:

  1. the broader markets did not experience significant buying or selling on Friday.  I conclude this from the nearly horizontal/range-bound white value in the lower pane.
  2. the broader markets are in a general down trend.  I conclude this from the downward-sloping values of the respective moving averages
  3. there was an attempt to move the markets upward around 11:30 a.m. Friday morning, but this move was rejected, and the move down was confirmed around 12:30.  I conclude this by the upward-moving attempt at 11:30, the reversal near that time, the second but lower-high around 12:30, and then the third attempt / failure around 14:00.
I feel that while the trend lines are all pointing down initiating new long positions is ill-advised, no matter what GGT "New Long" or "Affirmed Long" stocks say about the individual equities.  It simply is swimming up stream.

The anti-thesis to this is that we should be considering going short in the market.  Going short can be done by literally going short in a margin account, or for those who have retirement (IRA) accounts, simply looking at inversed ETFs can be a good alternative.

First, I'd start with the broad markets -- S&P500, NASDAQ, Russell 2000, etc.  There are ETFs that specifically target these broad markets, hence it pays to understand their present trend.

TradeStation tracks the stocks comprising the S&P500 with the symbol $TIKSP, and here is the 3-day view:

Here, the presentation is a bit different than in the broader $TICK display -- after about 11:30 on Friday, stocks comprising the S&P500 started to sell off in a fairly steady fashion at greater than 175 stocks/min.  Sure, there were periods where the breakouts to the upside were attempted, but in general these failed just as in the broader market view.  You can see that selling in the latter part of the day was even more pronounced and that it accelerated as the close was approached -- I take this to be that traders were unloading their positions ahead of the weekend and out of uncertainty.

The S&P500 has been relatively reluctant to drop as much as the Russell 2K or the NASDAQ -- I attribute this to the general flock to the dividend-"safety" of the constituent stocks.  This being said, the fact that we clearly have a downtrend established (albeit 2 days in length) on the S&P500 should tell you that long positions are somewhat tenuous at this point, and that you should be looking at contra ETFs:
  • SH:  -1x contra ETF on the S&P500, trades ~11MM shares/day
  • SDS:  -2x leveraged contra ETF on the S&P500, trades ~50MM shares/day
  • SPXU:  -3x leveraged contra ETF on the S&P500, trades ~18MM shares/day

Market Timing

It's obvious this past week though that we've already experienced a fairly significant down-draft.  Are we too late to enter contra ETFs?  Perhaps, but there are a number of methods to determine whether we should be on the short side or long side, and whether we are overbought (ideal area to enter shorts) or oversold (ideal area to enter longs).

The first method comes from Pascal Willain's work at the EV site.  You need to be a member in order to view this on an on-going basis, but I've permission to post a snapshot here of this market timer:

Pascal has developed a methodology to track money flow of the big players in the market, based on volume, and the top pane of the figure above shows us that money started to flow out of the market as of 9/21, as evidenced by the purple trace dropping below the 0 line.  Note, had you moved to a short position at the open on the 22nd (recall the $TICK and $TIKSP patterns on 9/22, the day you could have moved into the market), you'd be up in your position as of the close on Friday.  Also note that Friday saw a net outflow also (since the MF value is < 0), but there was some slowing of this outflow on Friday while the markets digested the drops of the past week.

The bottom pane is important -- it shows that we are nearing oversold areas and that a bounce could be in the making.  Note that we sometimes reverse sharply and move higher once we hit these levels, and also note that  the past two times we've stayed down here in this area, not moving too much.

Simply put, while I think we can still dip our toes into the contra ETF world, I think the probabilities of the positions working out from here are less than ideal, with "ideal" meaning when the markets are still overbought.  

Here is one method that I use to look at when to enter a candidate:

Shown is the display for SH, the -1x contra ETF on the S&P500.  The ribbons at the top of the figure -- the red and green ones, are the most helpful to me.

Ignoring the topmost indicator (Boingo Weekly), and starting with Bull/Bear Power gives us a good start here.  Bull power is the amount which the highs are above the 13d MA.  Here we see that it's positive, and this is bullish for the stock.  Bear power is related but different -- it is the amount which the lows are above the 13d MA.  When both Bull and Bear Power are positive we should take note -- the stock is in an uptrend.

Underneath this is the 13d Force Index, calculated 2 different ways.  The first way -- using exponential moving averages, is faster and gets us into a trade faster.  I like to confirm this with the 13d MA -- the simple moving average -- and when they both are "green" (positive), like now, we have a good setup.

Underneath this is the 2d Force Index, but the colors are reversed from the 13d FI.  This means that when the 2d FI is "red" it is positive, and when it's "green" it's negative.  As you can see, it's been "red" for several days.  A red 2d FI is blocking entry -- we need it to turn negative (green), THEN we need it to move positive again, taking out the high or the close of the previous day to show resumption of the trend.

Many contra ETFs are in this situation -- showing that they are early -- and we need them to get batted down a day with the markets rallying and THEN we can consider entry when they take out previous closes or highs.

If you drop down you see that I've also plotted the slopes of the 13d and 34d moving averages.  I typically require that these be in an up trend and that the 13d slope > 34d slope -- this helps to ensure that the equity is in in an uptrend.  Here, you can see that the 13d is above the 34d slope, and that they are both pointing upward -- momentum has been with these two.

Somewhat of a concern too is that the "smart money" appears to be leaving SH at the present time, independent of the $TIKSP pattern:

The top pane is the price series, and the middle pane is Effective Volume -- All (yellow), large (green) and small (red).  What this shows is that we had a good anticipation of this jump in SH on Wednesday -- Large Effective Volume (LEV) was diverging from the Small Effective Volume (SmEV).  Thursday saw a gap up and more importantly, the gap didn't close, which was good for the entry.  Despite this, near the end on Thursday, some selling did occur as far as LEV was concerned, and by Friday's action most of the large players had moved out of the equity.

All of this should paint a picture on how I use the tick patterns with EV to determine what is going on with a particular broad-based equity.  For SH, it appears that the time has come and gone to participate, but I'll keep an eye on the 1) decreasing $TIKSP patterns, 2) the broad-based MF indicator at the EV site, and 3) the LEV patterns for the equity.


The $TIKRL is centric on the Russell 2000 stocks:

Based on my analysis above for the S&P500, you can clearly see here that the selling in small caps has been more pronounced, and it should be somewhat clear to you that we should have entered a short /contra position in IWM / RWM respectively as early as Tuesday of last week.  Here's the view of RWM, the -1x contra ETF on the Russell 2K, using HGSI:

Same analysis as previous -- but note here too that the 2d FI has been steadily positive, essentially precluding entry.  Entry right now into RWM is not advised based on the indicators above.

Here's what Effective Volume has to say about RWM, through the end of Friday:

I was a bit late getting into RWM -- I didn't enter until the end of the day on Wednesday and it's obvious I lost some gains because of my travels and late decisions.  Nevertheless, we're seeing some LEV money flow out of RWM, and correspondingly, it would be good to wait on RWM until this situation reversed before playing it again.

So, the above presentations show that we're in a downtrend, that we're entering an oversold area, and that money is flowing out of SH (-1x ETF on the S&P500) and RWM (-1x ETF on the Russell 2K).  Elder's 2d Force Index method is holding us off on entry to either of these, as is the decreasing LEV patterns on Friday to both SH and RWM.


GGT's Price Slope Model is showing the bearishness in the market, but it is also showing the first stages of a crack in the bear ice:

It's clear by the left side of the table that we've been in a pricing downtrend since Tuesday of last week on all time frames, and if we consider the 55d and 65d slope columns only, at least since the middle of July.

On the right you can see that Friday's action resulted in ALL of the slopes of the slopes pointing upward -- this is significant because the right side of the table leads the left side, and we need "green" (positive SoS) in order to see "green" on the left side (positive slope).

On the left I've boxed the Strength Index -- it's fallen for 5 consecutive days and is now in oversold territory -- I am expecting a bounce.  We've been here before, but when we have, we ALWAYS bounce within a few days.

Here's the graphical view of the Strength Index to put it all in perspective:

You can see from the above that we're at fairly extreme levels in the oversold category so entry to contra positions is probably not the best method for Monday ....


The GGT Long-Cash Ratio (LCR) Slope Model is also showing some interesting indicators:

We've had 5 solid days of the database contracting and a falling Long-Cash Ratio (LCR), telling us more stocks are dropping below historical thresholds where they outperformed than are presently above those levels and outperforming.  I do note with some interest that the left side of the table warned us of the pending contraction, as the longer slopes started to move aggressively to the negative (red), and on the right hand side of the table, the unanimous agreement on all time frames through the 65d slope of the slope, starting on 9/19, certainly gave us a great head's up that bad things were possible.  9/20 confirmed with the 2nd day of all of the SoS being negative, and by the 3rd day there was no doubt that you should have been reviewing the contra side of the world.

Note closely though that Thursday and Friday SoS are starting to point up on the shorter time frames.  While not a strong signal -- it does not span all time frames -- this is indicative of some buying in the market AND more importantly, if we get all of the SoS periods turning up with Monday's action, we could see a bounce this week.

Time will tell.  We're shifting gears here, so I think it prudent to consider a short-term, surgical play in long ETFs if this newly established roll-over to long positions continues (as possibly confirmed by tick, Elder, and EV indicators today (Monday)).


Take a look at the following chart:

This plots the 34d EMA of the Long-Cash Ratio with the slope of the 34d EMA.  It basically is telling us that the last 5 days have been downward in database growth, and historically, when we've entered this area from above (like now), we bounce for a few days, wiggle around, then continue downward.  I have no reason to think that we'll do otherwise.  The EMA is in a clear downtrend in terms of the database expansion on the longer term so we'll continue to wallow around here for a while, bouncing back and forth.


As I opened with, there is no long-term up trend.  Short, surgical strikes improve the chances that you will win in this market.  I'll continue to post my view of the tick patterns, as well as my slope models, but be advised, there isn't a compelling reason to leave positions on for any term longer than a few days.

Good luck this week in the markets.

Remember, you are responsible for your own trading decisions, and I am not.  Please do your diligence, and please take ownership for your actions.



Monday, September 5, 2011

Weekend Update for September 2nd

Program selling on the NYSE started late Wednesday, August 31st, and continued with some acceleration into the close on Friday.  The easiest way to view this is with the cumulative $TICK pattern, which shows the 8000+ stocks on the NYSE:

As with all my images, right-click on the image to open in a new window or tab.

The top pane shows minute-bars on the SPX, simply as a point of reference.  The middle pane shows the daily cumulative tick pattern with a 250-stock filter applied.  This means that if the bar has 150 stocks either moving up or moving down it will not be added to the cumulative tick pattern in the middle pane, but if the number is 251 or greater it will.  This helps to catch the big algo moves.  The bottom pane shows various length moving averages on a non-filtered and continuous (non-resetting) $TICK pattern, and you can see that slightly more than 4 days are shown.

It should be clear to you that once the instantaneous (white) cumulative $TICK pattern failed to clear longer moving averages, and as soon as those moving averages started to slope down, we had a good short-term signal to move into the markets on the short/contra-ETF side.

Tradestation creates tick presentations for stocks that comprise the NDX-100, the SPX, and the Russell 2000.  The NDX pattern has been relatively poor all week:

The selling started on Wednesday, almost from the start, and with a 25/-25 filter applied (there is only 100 stocks), we can see that it remained on Thursday even thought the instantaneous tick pattern was relatively unchanged.  The top trace is the NDX-100 and it's clear that we lost price value on both Wednesday and Thursday, supporting the middle pane presentation.  The roll-over of the moving averages on Wednesday certainly gave plenty of warning to close any QQQ/QLD/TQQQ positions and to look closely at PSQ/QID/SQQQ.

Friday was absent of algo selling in the NDX stocks, at least large groups of stocks moving down at 250/min or faster.  The middle pane is relatively flat (until the end of the day), and a sailboat drifting in windless oceans immediately becomes the vision in my head -- we won't know what direction we're going until the wind picks up.

The Russell 2K stocks have been getting hammered too:

Interesting pattern Thursday and Friday -- the algo selling didn't start until after lunch on both days.  The filter level here is 500/-500, which means that 500 stocks / min had to be moving up/down in order to move the middle presentation.  This is a graphic example of broad-based institutional selling and I think that positions in RWM/TWM/TZA are worthy of consideration.  Given the fact that we're 2.5 days into the selling, I most likely will back off of the -3x ETFs and steer towards shorting a position in IWM rather than using a leveraged instrument (more liquid).

As has been the case for the last month or so, the stocks of the SPX are the last to sell / seem the most resilient.  I've heard all sorts of reasons why but the primary reason most likely has something to do with dividend yield of the SPX compared to the dividend yield of treasuries.  Whatever the reason, we're not seeing the broad-based, multi-day selling that we see in the other baskets of stocks:

All that being said, the SPX appears a great deal like the stocks on the NYSE in general, e.g., gentle rolling-over of the moving averages, etc.  It was only late Friday afternoon that we saw the selling algos kick in, which seems to be tantamount to throwing in the towel before the long Labor Day weekend.

While I can read a chart, I'm terrible at predicting the future.  Despite this weakness, I can say that even if we get a dramatic move upward on Tuesday it will take some work to move the tick MAs so that they are pointing upward, which all but rules out me moving long on Tuesday.  The converse isn't overly attractive either because this signal is 2.5 days old (or so) -- futures are down sharply as I write this due to a myriad of problems in Europe (ECB, Germany, Euro) and the Med (Greece) -- so I'm more inclined to launch positions in PSQ, RWM, SH or shorting QQQ, IWM, SPY (liquidity is better) than I am to move long.


GGT Price Slope Model

The PSM has moved to a decisively bearish stance:

The GGT price index fell -1.61% on Thursday and -2.75% on Friday on volume that was indicative of a long-weekend being in the crosshairs of most people.  Notable was that on Wednesday the GGT strength index stalled at 0.832, which is in bullish territory, only to fall to 0.207 on Friday, which is nearly from one end of the range to the other.  It's hard to believe but this rapid movement from end-to-end actually supports the assertion that we may bounce this week -- the change between 0.83 to 0.21 is large and to do this in 3 days is rare.  In each case that I can find in GGT's history we've ALWAYS bounced within 5 trading days.  Of course, "n" is small, so we could continue south with no regard to the past or market psychology.

The right side of the table tells the story.  We saw gradual breakdown of the slopes of the slopes of the moving averages starting on Tuesday, continuing on Wednesday (albeit weak), then the hammer on Thursday and Friday. You don't need me to tell you that your short-term pricing on your long positions probably told you to close them at the open on Friday morning.  If you held across the weekend you'll most likely pay a price for this optimism come the open on Tuesday morning if the ES continues to look as poor as it does as I write this.

With the red bar across the bottom of the table, on both sides, the PSM is basically telling us to get out of our long positions.  We have no idea if the downtrend will continue, and it's only the magic of support levels and the market's interpretation of such that can give us any hope at reversing the bleeding.  I personally do not see much support where we are at and where we can go, but I'll leave that commentary to those more skilled in such things (Billy @ and Bob English at the same location are probably the best and most patient folks to read concerning this topic).


LCR Slope Model

The LCR Slope Model doesn't look as nearly as bad as the PSM, but the metrics are completely different too.  The database has shed nearly 41% of it's long positions in just 2 days, effectively halting any advance in the model:

On the right side, while we've seen a gradual "buy-in" on a day-over-day basis of stocks moving to the long side (as evidenced by the staircase pattern of "bullish" readings), Friday was characterized by only a few of the slopes of the slopes of the moving averages moving bearish.  This shows weakness -- near threshold levels -- and overall, indecisiveness.  I prefer right-side presentations where the slope of the slopes ALL turn one way or another on a given day, and we've not seen this since back in July.

The presentation above tells me we've effectively halted any advance to the long side in the LCR and that a contraction of the database is very possible.  It only takes a price drop to move a stock from some form of LONG recommendation to CASH, so continued erosion in prices will cause a widespread contraction in the database LCR.  I've shown in previous posts that investing when the LCR is contracting is bad for our bank accounts, so again, if there were any doubt, you probably should move to the sidelines.

As evidence of this contraction, I track an indicator that I've developed that shows the day-over-day change in the number of stocks moving to the long side, or conversely, to the cash side:

We peaked earlier this week, as I noted in my blog at, and now we're in the process of pulling back.  We've pulled back only a slight amount, and given that the database strength has fallen to 0.2 (see above), it's very possible we could bounce this week, taking this indicator upward and potentially to new highs.  For now we're on a downward move, and this contraction location is the wrong place to be entering long positions.


My final graph refers back to a graphical view of the strength index, simply to illustrate the broad movement we are experiencing in strength.  I wrote about this up above in the PSM section:

I showed that we had a great probability of reversal this past week, and now we've plunged quite low.  Certainly, we can remain here, but when we move fast and furious in either direction you can see that we tend to bounce ping-pong style between bullish/bearish.  Whipsaw indeed.


If you feel the need to play the markets, I suggest you concentrate on ETFs tied to the tick patterns.  Of course, you must have access to at least $TICK in order to do this, and if you have TradeStation you can download my ELD and TSW files from the "Links" thread in my forum at  I personally think that this is a terrible time to play stocks either long or short, so tread there only if you have the confidence and trading rules in place to cut your losses and take your profits at predefined levels.

Remember, you are responsible for your own trading decisions, and I am not.  Please do your diligence, and please take ownership for your actions.