Saturday, August 20, 2011

Weekend Update - August 20th

I've decided to return here to post once a week, a request made by several of you.  My weekday blogging and market status interpretations can be found at, specifically in the GGT forum on that site.  I urge you to join that forum if you are not presently a member -- it's free and I'm continually humbled by the experience and quality of the leadership shown by the members.


I like to start with the various tick indicators, as they show the strength and direction of the algos which are driving our markets:

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

This tick chart shows the S&P500 in the top pane (orange trace), the cumulative daily tick in the middle pane (red trace) which resets to zero at the start of each trading day, and a lower pane showing a white instantaneous non-resetting tick cumulator and various moving averages on that tick cumulator.  The bar interval is 1 minute in duration, and hence you're looking at a bit more than 3 days of data.

Friday had me in the air flying back to the east coast from the west for the majority of the day, hence the action on Friday was mostly missed by me.  The SPX was convincingly being sold off on OPX day, with the white trace solidly below the 1/2-day (195 minute) moving average line.  Every time the instantaneous cumulative tick touched this value and was rejected (10:30 EDT, 12:30 EDT) itdr was a good entry point using SH (-1x ETF of the S&P500) or SDS (-2x ETF of the S&P500).  I was personally worried about carrying anything into the weekend but had I seen the nearly linear behavior of the selling which occurred in the afternoon (refer to the middle plot which is red).  This chart is not predictive but note that the selling on Friday, while continuous, was not at the same magnitude as Thursday's lever level (again, middle plot), introducing uncertainty as to whether we will continue the downward trajectory.

This next plot is the NDX:

The NDX has been a bit quicker in rolling over than the SPX, but not as quick as the Russell 2000, which you'll see in a moment.  You can clearly see the fan pattern of the moving averages rolling over late Thursday, giving you confidence at moving into an initial position in PSQ or QID.  There was strength in the NDX on Friday, which would have made you a bit nervous, but as soon as the selling started and was confirmed by the break through the 0 line (middle pane, red trace) around lunch:30 you would have had further confidence to move into the markets.

Note that the NDX finished near the cumulative tick lows for the day, which is suggestive of an early bounce when the markets open.  We'll see.

This last figure is of the Russell 2K small caps:

The R2K has been fairly clear and decisive in its' signals, and as many of you know, I entered a position in TZA on Wednesday and again on Thursday.  Somewhat mistakenly, and because I was in the air and TZA is leveraged, I sold my positions on Friday on layover in Denver.  I was disappointed to not have carried the positions in multiple accounts across the weekend with the presentation that you see above...

Of the three presentations, the R2K is leading the way down, and the moving average slopes are all solidly down-pointing relative to the other figures.  Conclude that constituent stocks in the R2K are being algorithmically sold and with the SPX being last to the party, that the selling is becoming broader.  It does not appear that you are too late to the selling party if you haven't yet participated, but ensure you watch your stops.  I would avoid leveraged ETFs at this point and only concentrate on long positions in the -1x or shorting the underlying 1x ETF (SPY, QQQ, IWM respectively).


Bob English, who is a insightful and prolific blogger at the EV site, pointed out earlier in the week that the lag of the SPX to the selling party was because the yields of the S&P500 are higher than treasuries.  This flight to safety is something that I regularly track, and you can see it here in the following graph:

The top pane shows the S&P500 and the bottom pane (blue with red moving average) shows the ratio of large cap stocks (as determined by the IWB) to the small cap index (as determined by IWM).  Clearly, with the 34d MA pointing upward on the ratio of IWB/IWM, and the ratio making new local highs, long positions and specifically, long positions in small cap stocks should be avoided like the plague.

A key early-warning sign will be when this ratio pierces the 34d moving average.  This alone is not sufficient for a pure signal, and you can see it, but we certainly want to watch for the slope of the MA to start flattening and then turning down, all of which requires that the IWB/IWM ratio fall below the MA.


We're a great distance from a confirmed long trend.  I base this on the following damage to the GGT price index model, specifically the daily change that we're seeing in the 21d, 34d, and 65d moving averages:

Shown above are the GGT price index values (left scale) and the slope of the price, given in $/day change (right scale) for the 21d, 34d, and 65d moving averages.  I've artifically drawn two red lines which show the valley and peak of the last slope values over the past couple of weeks.

You will observe that the 21d MA slope is approaching -$0.30/day average, which is just under -1% change per day.  You will also observe that as the MA duration goes up the slope values decrease, but all are negative (in the pink zone).  Day-over-day the index is losing money, and until these move positive (white area) we cannot be convinced that the waters are clear of sharks.

The more aggressive among us will want to start adding to long positions when we hit the lower red lines and bounce off of these lines to the upside -- or better yet -- never hit these lines yet reverse to the upside.  I personally would like to see higher lows on this next dip as this would indicate that we could move aggressively upwards from the turning point.  Conversely, if the lower red lines are penetrated I would be concerned about prolonged decreases in price and I would aggressively add to my shorts at that time.

I'll post the status of these levels when and if they become significant.


This next graph shows the status of the GGT database in terms of the 6 different classifications that a specific stock can hold.  The vertical axis is % in that category, compared to the total number of stocks which is 2813 at the present time.

While the scaling is hard to observe for the New Longs/New Cash values, if you look closely we're seeing more stocks moving to the New Cash side on a day-over-day basis than New Longs.  This bleeding is indicative of mass selling.  Further, more stocks are reaffirming their "Affirmed Cash" position than any other bin, again showing that in terms of trends, the trend is solidly down.  Affirmed Cash stocks are falling in price (at a minimum), and may have significant volume associated with the drop in price.

Note that the only way to determine if the falling price is coincident with aggressive, large-block selling volume is to review the individual stock or ETF candidate at the EV site.  There is also a list of stocks and ETFs that are produced daily which show divergences, and overall, are very good candidates for entry as long as the EV trends remain intact.  You need to be a paid member of the site to view this file.


This next table shows that we are deep in kimshi:

The price slope model, which is based on the prices, volume, and rate of change of stocks in the database,  has been indicating danger since 7/26, and the last rally over the past week barely saw enough strength to cause anything more than the shortest pricing MAs to move positive.  We now have 4 days  of downward-pointing "slopes of the slopes", and this is bad, very bad, for long positions.  Do not even attempt to move long until we get a row of greens on the right side of the table.

I want to draw your attention to the GGT strength oscillator, which is a value of the database strength between 0 and 1.  I've marked the previous values when we were low, and unfortunately, we're lower with Friday's action.  This somewhat causes the idea of a higher low to be tossed out, but this is only one indicator in many, so we'll have to be careful here.  I note that values this low in the past have been WONDERFUL buying opportunities if you have a long horizon.  Unfortunately, there are far more indicators which indicate we should stay on the sidelines for now.  Simply know that you should be looking for EV stocks that are dropping in price but are increasing in Large EV accumulation.

This price-slope-model chart is posted (nearly) daily in my blog at the EV site.

Related to this chart is this one:

This table is the Long-Cash Ratio (LCR) slope model table, and while we saw much green over the past two weeks on the right side of the table, the damage has been so severe that few of the slopes were able to move into positive territory (left side of the table).  The LCR value, which looks at the number of stocks rated some form of long (New Long, Affirmed Long, Long) and compared those to those that have some form of cash rating (New Cash, Affirmed Cash, Cash) is now showing that only 6% of the database is long.  While not as low as a few weeks ago (the measured low was on 8/9 @ 2.5%) if we can hold these levels AND see green start developing again on the right side of the table, we may see a floor being put in.

Stay tuned.

Monday isn't the day to go long on stocks by any stretch of the imagination.


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