PDA

View Full Version : Index Cluster Weights



grems8544
02-16-2012, 08:06 AM
Here's the latest cluster weights for the major indexes and their tracking ETFs:

12910

For those of you who want background, please read here (http://www.effectivevolume.com/showthread.php?4676-Index-Pivots-for-Monday-February-13&p=20510#post20510) and the entries prior to this link.

We have a mixed bag if you take these weights in isolation. The Russell 2K is indicating relatively poor risk/reward (RR), with all values below 0.35. When values are in the 2-4 range entry is favorable on the long side. In fact, if you look all the way to the right, the shorting RR values are "at zone".

Note what I call "zone breadth" for each of these ETFs. I define "zone breadth" as the calculated distance between the nearest support with a weighting over 3 and the nearest resistance with a weight over 3. TZA has a huge zone breadth -- 18.2% - 2.0% = 16.2% / leverage amt (3) = 5.4%. TWM has a 11.4% - 1.1% = 10.3% / leverage (2) = 5.15%. RWM is at 5.1%. These are healthy allowances for downward moves and are suggestive of favorable RR on the contra ETF side of IWM.

Once again though the S&P500 isn't confirming the Russell.

The S&P500 is caught between the monthly R1 and R2, right where John Person indicates is a normal, healthy location for a bull market. If you look to the far right, we'd need to move up +1.4% in the SPY from here in order to hit an optimal 3:1 shorting R/R point.

Further, if you look at the zone breadth for SPY, SSO, and UPRO, you can see that for the SPY we're at 1.6%, for the SSO we're at 1.7%, and for UPRO we're at 1.7%. This suggests, in isolation, that there is upside room available on the SPY and it's derivatives.

The NAS saw distribution yesterday, largely due to AAPL, and the zone breadth for the long side of the NAS is not supportive of entry. QQQ has a ZB of 0.2%, QLD is 0%, and TQQQ is negative. The shorting RR entry points are already "green", meaning we're favorable in terms of entry, so entry on any strength on the SHORT side could be a good bet. Note though that with such poor zone breadth it is quite hard to determine proper shorting R/R.

All of this is still work in progress, but is posted for your review and comment.

Regards,

pgd

TraderD
02-16-2012, 08:35 AM
Here's the latest cluster weights for the major indexes and their tracking ETFs:

For those of you who want background, please read here (http://www.effectivevolume.com/showthread.php?4676-Index-Pivots-for-Monday-February-13&p=20510#post20510) and the entries prior to this link.

We have a mixed bag if you take these weights in isolation. The Russell 2K is indicating relatively poor risk/reward (RR), with all values below 0.35. When values are in the 2-4 range entry is favorable on the long side. In fact, if you look all the way to the right, the shorting RR values are "at zone".

Note what I call "zone breadth" for each of these ETFs. I define "zone breadth" as the calculated distance between the nearest support with a weighting over 3 and the nearest resistance with a weight over 3. TZA has a huge zone breadth -- 18.2% - 2.0% = 16.2% / leverage amt (3) = 5.4%. TWM has a 11.4% - 1.1% = 10.3% / leverage (2) = 5.15%. RWM is at 5.1%. These are healthy allowances for downward moves and are suggestive of favorable RR on the contra ETF side of IWM.

Once again though the S&P500 isn't confirming the Russell.

The S&P500 is caught between the monthly R1 and R2, right where John Person indicates is a normal, healthy location for a bull market. If you look to the far right, we'd need to move up +1.4% in the SPY from here in order to hit an optimal 3:1 shorting R/R point.

Further, if you look at the zone breadth for SPY, SSO, and UPRO, you can see that for the SPY we're at 1.6%, for the SSO we're at 1.7%, and for UPRO we're at 1.7%. This suggests, in isolation, that there is upside room available on the SPY and it's derivatives.

The NAS saw distribution yesterday, largely due to AAPL, and the zone breadth for the long side of the NAS is not supportive of entry. QQQ has a ZB of 0.2%, QLD is 0%, and TQQQ is negative. The shorting RR entry points are already "green", meaning we're favorable in terms of entry, so entry on any strength on the SHORT side could be a good bet. Note though that with such poor zone breadth it is quite hard to determine proper shorting R/R.

All of this is still work in progress, but is posted for your review and comment.

Regards,

pgd

Paul,

The challenge in interpreting these conflicting edges seems to me mostly due to the different context/timeframe in which they're being viewed. I noticed that counter-trend moves occur with fairly high frequency off of major pivots/clusters, especially in the first 1-2 occurrences when price reaches these points.

At least in hindsight, being conservative with respect to expectations for the counter-trend swing move (ie taking profits when the then-prevailing pivot-based edge disappears) would have worked quite well. What that means, I think, is that it's not only crucial to enter at the right place, but also monitor the exit very carefully to guard the fledgling move and press the eject button prior to the trend swamping it.

$.02,

Trader D

grems8544
02-16-2012, 08:51 AM
The challenge in interpreting these conflicting edges seems to me mostly due to the different context/timeframe in which they're being viewed.

Agree completely.

I need a 3-D viewing scheme. I was lying in bed last night thinking something along these lines:

The cluster weighting system is a snapshot in time -- it does not tell us how we got here, nor does it tell us where we are going. It's much like viewing an instantaneous price series, but now, we have 1-8 weights and 6 levels (R3-R1, S1-S3) to consider, for a total of 48 series to consider.

I created GGT and the slope/slope of slope models because the slope of slope models LEAD the change in slope, and a transition in slope on multiple time frames generally confirms a change in trend in the overall market. It's exactly what we're seeing today in my LCR system.

Along these same lines, there has to be information contained in the sequence of where we came from in terms of clusters to where we are today. Predictive? No. Just like slopes and slopes of slopes are not predictive, they can only tell you where you have come from. BUT, just because we transition (like today) into a favorable RWM/TWM/TZA setup for entry, if the 3:1 RR (or some other magical number) is attained, what do the cluster weights look like at that point? I think a challenge we have is that cluster weights move continuously, and a system built on cluster weighting needs to have a clear entry "tunnel" to the 3:1 point, and once attained, must have a clear exit to a sell point.

Pascal and Billy raise the floor of the stops based on this, and this system generally works fine, but I think fundamentally, if we could "fast forward" to the 3:1 point and look at the new cluster supports, would that influence our entry/exit? The answer is "perhaps" if it occurs on a daily move, "probably" if it occurs within a week, and "most likely" if it occurs within a timeframe larger than a week.

I don't know.

My day job is demanding much of my time, so I can't look at this except late in the evenings after the family retires. It is something that interests me though -- I think Billy's methods are solid, but now, I want to start looking at them with this R/R and entry/exit picture in mind.

Regards,

pgd

TraderD
02-16-2012, 09:22 AM
Along these same lines, there has to be information contained in the sequence of where we came from in terms of clusters to where we are today. Predictive? No. Just like slopes and slopes of slopes are not predictive, they can only tell you where you have come from. BUT, just because we transition (like today) into a favorable RWM/TWM/TZA setup for entry, if the 3:1 RR (or some other magical number) is attained, what do the cluster weights look like at that point? I think a challenge we have is that cluster weights move continuously, and a system built on cluster weighting needs to have a clear entry "tunnel" to the 3:1 point, and once attained, must have a clear exit to a sell point.


One way to implement this idea is to pre-calculate future pivots based on a set of prospective price scenarios and attaching a probability distribution to these scenarios. How far in advance to paint these scenarios is a good question, I would start from one day in advance just to see if this holds water. The scenarios may be the current set of calculated pivots and their weights (factored with respect to where we are on the calendar) may be used to attach respective probabilities. Some more details need to be flushed out, but that can essentially define the "tunnel". You prepare to pounce where the maxima (or maximas) of the scenario distribution manifold generates good RR entry points.

Trader D

grems8544
02-16-2012, 09:30 AM
One way to implement this idea is to pre-calculate future pivots based on a set of prospective price scenarios and attaching a probability distribution to these scenarios.

Agreed. I think using a model like an options pricing model probability curve could give us a head start on the likelihood of a particular price being achieved.

EB
02-16-2012, 02:10 PM
One way to implement this idea is to pre-calculate future pivots based on a set of prospective price scenarios and attaching a probability distribution to these scenarios. How far in advance to paint these scenarios is a good question, I would start from one day in advance just to see if this holds water. The scenarios may be the current set of calculated pivots and their weights (factored with respect to where we are on the calendar) may be used to attach respective probabilities. Some more details need to be flushed out, but that can essentially define the "tunnel". You prepare to pounce where the maxima (or maximas) of the scenario distribution manifold generates good RR entry points.

Trader D

This is an interesting idea. The fact that each period's pivots is based on three variables, the high, low and close does add some complexity to painting scenarios. But the daily pivots are the lowest ranking anyway. I think it gets interesting on the few days before the end of a week, month, etc. when the following period's pivots are being set up.

TraderD
02-16-2012, 02:29 PM
This is an interesting idea. The fact that each period's pivots is based on three variables, the high, low and close does add some complexity to painting scenarios. But the daily pivots are the lowest ranking anyway. I think it gets interesting on the few days before the end of a week, month, etc. when the following period's pivots are being set up.

H/L generation does look like it would complicate things. I suppose an ad-hoc approach of generating a stochastic price series over and over can help, but it might be a chicken and egg situation (ie if you can generate an accurate enough model for price behavior, what do you need pivots for...)

grems8544
02-17-2012, 09:22 AM
Here's the cluster weights as of the close on Thursday:

12945

From a clustering point of view, IWM and it's variants are all supportive of a move upward. We do have a balanced R1/S1 weighting that is uniform for TNA and UWM, and slightly more over head resistance by 1 compared to support for IWM. You can see this favoritism by the values in the "Bal" column all being less than 1.0.

Ditto S&P500 and the DJ30.

NAS, not so much. YOu can see that TQQQ is telling us to jump in, but there is no confirmation on the lower leveraged amounts (QLD, 2x, and QQQ, tracking).

Somewhat of interest is that the zone breadth, which I define as the distance from the closest support weight > 3 to price and the closest resistance weight > 3, again closest to price.

For the R2K, we're right in the middle of the zone, with 1.4% until we start entering resistance and 1.6% until we hit support. I note that we're right between R1 and R2, which is healthy in John Person's view for a bull market.

For the S&P500, the situation is the same -- we're near the middle of the zone, and there is no bias either way.

For the NAS there is overhead resistance at quarterly pivots, so it will be interesting to see how the market deals with this. I note too that TQQQ is above qR2 by 0.8% and has a long distance to mR3, so we'll see if this is at all able to hold water.

The more conservative part of my brain looks at the Risk/Reward (column RR). When these values move to 2-3, they turn puke-green. When they are above 3, they are at the "buy point" You can see that these values for the R2K and SP500 are all below 2, so they are red -- R/R is poor for entry at the present prices. The NAS is showing some large numbers for QQQ and this is a problem I have with a simple R/R calculation -- it is non-linear the smaller the denominator gets (e.g., as you get closer to a lower support the denominator goes to 0, causing the R/R to skyrocket). The converse is true if we get close to a resistance level (R/R --> 0). The "linear" zone is from about 1 to 7 or so, so I have been watching those areas for entry.

Again, all pure mental games here. I'm not actually trading any money on these yet, but simply trying to learn what I'm seeing.

Regards,

pgd

================

Late (edit) post:

In the FWIW column, I took a snapshot of the Precious Metals Cluster weight when I received Pascal's RT signal concerning GDX (about an hour after the actual signal). Here's the figure for your review:

12946