Monday, January 18, 2016

Single-Digit LCR and Uncorrelated Quality Stocks

Please subscribe to this using the "Follow by Email" link to the left.  Having your email helps me see the number of folks reading this.


The Long-Cash-Ratio (LCR) has fallen into the single-digit measurement range, indicating incredibly oversold conditions.  Acceleration indicators are not plunging downward but are starting to ease and show mixed movements upward / downward depending on the analysis time frame.  It's only a short period of time before we see broad market participation regarding buying.

What is the LCR?

The LCR is an indicator that I created that compares the number of stocks in the database that are rated "Long" to those that are rated "Cash".  It is a ratio, so in today's example, we presently have 260 stocks that are in some form of "Long" and the rest (2762) are in 'Cash".  This is a ratio of 260/2760 or 0.094.  Stocks assignments are adaptive to individual market performance (price, volume, price rate of change) so the individual stocks are evaluated first, then aggregated together to form the LCR.  The global behavior of the stocks in the database in no way influences the individual stock recommendation -- this is performed in a "brick-by-brick" analysis and a big summation sign is put in front of all the stocks.  This makes the LCR superior to other methods of understanding market behavior.

Single-Digit LCR -- What does it Imply?

Simply, it suggests that this could be a good buying opportunity, although there are limited observations using the approach described below:

1) Buy the GGT index (looks most like the ETF VTI, or Vanguard Total Index) at the closing price the following day the LCR index enters single digits.
2) Hold the GGT index until the LCR index moves back above parity, that is, when the number of "long" rated stocks exceeds the number of "cash" rated stocks.

There have been 3 of these events since 2009.  All have made money.  All take a bit of fortitude because every one has dipped from the buying point to underwater before they recover.

The data above uses the GGT index from those dates.  The GGT index is an equal-weight average of all the stock prices in the database, and in general, the total number of stocks strives to maintain 3000 or so, plus or minus 100.  I've been able to keep that level since inception in 2008 so the index is sound.

The first thing that should be apparent is that the durations of holding after the buy can be quite long.  This suggests that we could continue to play in a rough area for a considerable length of time.  Note that the Duration indicated reflects calendar days, not trading days.  Nevertheless -- buying in August and holding a momentum portfolio until October is rather unusual.

Also not that you'll need to understand that every portfolio gave up some gains sometime during the holding period.  I've indicated the Drawdown -- this is the amount dropped from any local peak.

What if we bought/held GGT "long" rated stocks from those dates?  

In general, the answer is favorable.  Using HGSI ( ), it's possible to build the "what if".  Note though that although I have the stock lists that were "GGT New and Affirmed Longs" on those dates shown above, some of the stocks have either been acquired or have gone out of business.  Hence, the database today is not 100% accurate to reflect exactly what would have transpired on the dates in the past. 

This is known as survivorship bias, and you can Google it if you want to read more about what it means.

Anyways, baskets of stocks purchased/sold on the aforementioned dates in the previous table result in the following charts:

Click on the image to enlarge

Click on the image to enlarge

Click on the image to enlarge

If you click on the charts, you see the following gains:

March 2009 Portfolio:  11.2% gain
May 2010 Portfolio:  2.9% gain
August 2011 Portfolio:  4% gain

So, we have a reasonable expectation that buying in these conditions could produce a positive gain.  Note, your crystal ball is as good as mine, and the past is certainly not a guarantee of the future.

This being stated, *if* we were to consider buying, what is the group behavior of the stocks that are being considered.

GGT separates "long" stocks into three groups:

  • New Long -- previously in some form of "cash" recommendation, these stocks *just* converted to "long" recommendations on higher volume AND price.  These stocks are attracting attention in the present climate.  They may be undervalued, overvalued, or have something else going on with them -- all that I know is that they are appreciating in price and doing so on higher-than-average volume.
  • Affirmed Long -- previously in some form of "long" recommendation, these stocks are moving higher on combined higher volume and prices.  These stocks have already been firing on all cylinders, and these stocks are showing strength in the present market.  These are what some people call "leaders".
  • Long -- These stocks are in some form of long but do not have enough downward price pressure to kick them into "cash".  They could be moving upward or downward, but are moving in a low-volume environment.  Specifically, there is not enough downward forces to drop prices dramatically, and there is not enough upward demand (due to low volume) to drive prices higher.  Higher volume on these stocks would kick them up to "Affirmed Long", while severe dropping in price, with or without volume, would kick them into "New Cash".
Of the 3, the "New Long" and "Affirmed Long" stocks are of interest to me.

Here's the charts of how New Long, Affirmed Long, and a combined "New Long + Affirmed Long" index appears:

Click on the image to enlarge

This is a New Long group index.  The individual stocks are available for subscribers to my Dropbox folder (free); instructions on how to do this are below.

The first thing that jumps off the chart at me is that the "New Long" stocks have been trading below their group moving averages.  This is to be expected -- "New Long" is a new breakout behavior pattern and hence I see the huge move on Friday from these stocks on higher volume.  

Of significance is that the middle of the plot shows the 13d and 34d slopes of the prices, and these are trending upwards.  This is a good sign, and these stocks have been slowly building the last week as the market dropped lower and lower.

A key question is whether I would buy these stocks?    As a group, they all have the required characteristics, except we are at the beginning of a (possible) change in trend.  This always suggests higher risk, so buyer beware.  If I were to jump in here I'd do so only on strength, and with a fairly tight stop loss.

Here's the chart for Affirmed Longs:

Click on the image to enlarge

This chart shows a much more established trend with these stocks.  The trading-above-moving averages is hugely positive, and the slopes of the 13d and 34d are pointed upwards and expanding.  There is positive momentum with this portfolio.

Compared to the New Longs, this Affirmed Long portfolio is more established.  Whether stocks will move up at the same rate as the New Long portfolio is to be seen.  Historically, the answer is "no", but I have no idea (nor do you) about the future.

If we combine the "New Longs" and the "Affirmed Longs" we get the following plot:

Click on the image to enlarge

This chart shows the combined behavior of all the stocks with "New Long" and "Affirmed Long" recommendations.  I see nothing wrong with this chart, and if I could buy the index, I certainly would.  It shows that we have established stocks as well as a major bump on Friday on higher volume, which is amazing.  Strong stocks in a down market are noteworthy, and this is a large basket of diversified stocks.

One thing that I like to do when I'm "bucking" market trends (and the primary market trend is downward right now) is to select low-correlation stocks, relative to the major indexes.  Correlation is measurable, and stocks are correlated if they move in the same direction together.  

For example, suppose the S&P500 (SPY) moves up 1%.  If the stock under consideration also does this, meaning move up 1%, and if this occurs a large number of times, say, over the past 100 trading days or so, we can measure the correlation.  A value of 1.0 means exactly correlated.

Conversely, if we take a stock and it does not move in correlation to the SPY, meaning that as the SPY goes up or down the stock does it's own thing randomly, then we get a value of 0.0 and this means the two are not at all correlated.

If a stock moves opposite the thing it's being compared to in exact sequence (SPY moves up 1%, stock moves down exactly 1%), then the two are inversely correlated and the value is -1.0.  For ETFs, the ETF (SPY) and it's inverse (SH) are inversely correlated with each other.  The correlation matrix for these two looks like this:

SPY 1.00 -1.00
SH -1.00 1.00
So, as the theory goes, if we pick quality stocks that are bucking the trend, the stocks will continue to buck the trend as long as they continue to buck the trend.  Meaning, nobody knows how long in the future the low-correlations will continue but it's a good start.

When I do this for the basket of New Long and Affirmed Long stocks, and I filter those out that are correlated with the Russell 2000, S&P 500, NASDAQ 100, and the DJIA, I'm left with the following:

Click on the image to enlarge

The "red" on the left shows that the R2K (^RUT), the S&P500 (SPY), the NAS (QQQ), and the DJIA (DIA) are all heavily correlated with each other, as we would expect.

The remaining green shows the correlation to these indexes and were hand-picked:  if the correlation was between -0.25 < stock < +0.25 turn it green and pick it.

When I take this list and pump it into HGSI here is what I get for the past recent behavior:

Click on the image to enlarge

The uncorrelated basket of stocks looks pretty good.  Not as strong as the previous basket, but not bad.

My personal expectation is that this basket of stocks will continue to outperform the markets over the longer haul as long as they remain uncorrelated.

Again, buyer beware.  Not recommendations -- do your diligence.

The LCR Table

Jumping into the markets on Tuesday, the first trading day after the MLK holiday, is ill-advised, correlated or not.  Nothing on my LCR Table suggests that Tuesday is a day of entry:

Click on the image to enlarge

The long-series of down days in the drop of the LCR suggests that we're going to bounce.  We haven't bounced, nor is there any indication that "Tuesday is the day".  Your crystal ball is as good as mine.

On the far right we do have an initial sign of an abating downward acceleration.  The ball may be reaching the top of the apex and may just start to fall back to earth, hence reversing direction.  Who knows?  What I do see is that the right side of the table is between reds and greens, meaning that the slope of the acceleration is different depending on time frame.  This is good.  Note though that history suggests we could play around in these low numbers for awhile, so care is required.

Given the LCR table status, which I believe in with 100% of my portfolios, Tuesday is not a buying day, uncorrelated basket or not.  Correlation doesn't change over night so the same basket of stocks that I calculated today will have (more/less) the same correlation values if I run them any time this week.  The only thing that will change is the GGT recommendation.

What I need to see in the table is that we see some green on the left side of the table.  I'd like to see (minimally) the 2d and 3d turn green in 1 day, and be quickly followed by the 5d and 8d.  If the 8d moves green I'm going to start buying and will let everybody know.

 Cumulative Tick

Click on the image to enlarge

Confirming a bit of "relaxing of the selling" is the cumulative tick indicator that I use.  See previous posts to get a flavor of what you're looking at.  For those of you who follow me, you see that we came (mildly) off the lows of the day so Friday wasn't a complete blood-bath -- there was some buying into the close.

This being said, NOTHING in the CT chart looks like a buying opportunity.  I intend to sit on the sidelines on Tuesday.


Although not discussed here, I'm probably going to buy deeply out-of-the-money cash-secured puts on some quality stocks that have earnings beyond the February options expiration date. These are going to be on quality stocks that I have evaluated as "undervalued" and that I would not mind owning a even a lower price than I can get right now.  Many do not pay a dividend so I free myself of having to track the ex-div dates, etc.   Might as well generate some minor income.

No buying of stocks, correlated or not, on Tuesday.

I continue to unwind positions that are recommended as "New Cash".  There aren't many left that I'm holding.  I've set my exit criteria and will be out of the positions if they hit their stops.  For those of you subscribing to the real-time alerts, you'll get automatic notification when the positions are exited.

For the non-Collective2 accounts, I will set a 1% trailing stop loss, GTC, effective after 9:45 a/ET and will forget about these positions until they either 1) reverse to a "New Long" recommendation or 2) are sold.

For the Collective2 accounts, I will set a 1% stop loss, GTC, effective after 9:45 a/ET and that will be adjusted upward (never downward) nightly until triggered.  The C2 accounts have been battered with this last round of down markets and I'm re-evaluating stocks and entry/exit criteria.


Stock updates are posted in a daily file that I attempt to share by the following morning with all subscribers. To review the stocks that you are holding and see how I evaluate them, you need to be a member of my Dropbox.  Send an email to pduncan [ a t} v _ t (dot] e du, fixing the address of course, with the word "DROPBOX" in the subject and I'll add your email.  I attended Virginia Tech many moons ago and it is my alumni address, so it should be easy to see how to fix the address -- simply use "".  I also ask that you subscribe to this list using the link to the left, as it's the only way I can communicate with Dropbox users, if the need arises.

Subscribe to my twitter feed if you want intra-day observations:  grems8544


New Meeting Announcement

We will be holding a face-to-face meeting on 2/13/2016 for all interested parties at the following location:

Burke Centre Library
Room: Burke Centre Meeting Room 116K
Address: 5935 Freds Oak Rd, Burke, VA 22015
Library Phone:(703) 249-1520
Time: 10:00 AM to 1:00 PM
Meeting Start Time: 10:00 AM

I will attempt to stream the meeting audio, and perhaps video, as per past meetings, via GotoMeeting.  Attendance via GotoMeeting is limited to the first 25 call-ins.  The ability to do this is completely controlled by the library and I have no say in Internet access.


As with all my ramblings, you are responsible for your own investment decisions and I am not.  Please do your own diligence, and please take ownership for your actions.