Index Pivots for Monday, February 13

[grems8544]

Call this an experiment in real time -- I'd like to post some of the displays that I'm seeing so we can get a better handle on pivot/cluster grouping within the major indexes.

As many of you know, I've created a TradeStation indicator to implement the methods of Billy and Maxime (I borrowed some of Bob's code for the pivot calculations). Here's my presentation on the major indexes, related ETFs with their leveraged counterparts, as well as the contra ETF counterparts.

12774

Here's a quick explanation of the columns:

R2, R1, S1, S2 as shown are the strength of the two clusters above (resistance) and below (support) the present price. This updates in real time. A cluster is defined as being 20xATR in width, e.g. an exponential 20d MA is used in the ATR calculation. The value listed in the column is based upon the strength assigned to whether the pivot level is a daily (1), weekly (2), monthly (3), 50d (4), quarterly (5), semester (6), 200d (7), or yearly (8).

Colors are indicated in the R2 .. S2 columns. Bright green indicates that the resistance clusters (R1 + R2) are at least 2x smaller than the support clusters (S1 + S2). In the chart TNA fits this, as (2 + 6) * 2 < ( 9 + 13 ). Conversely, bright red is the exact opposite, where the resistance clusters are 2x greater than the support clusters. In the chart QLD fits this description: (12 + 4) > 2 * (5 + 2).

Normal green in the R1 and S1 columns means that R1 < S1 (greater support). Normal red in the R1 and S1 columns means that R1 > S1 (greater overhead resistance).

Normal green in the R2 and S2 columns means that R1 + R2 < S1 + S2 (greater support). The converse in red means greater overhead resistance for both clusters.

Puke-green in a columns means equivalency -- the clusters are balanced and a great war is about to be waged (or not). IWM fits this description. Where the R2 and S2 columns are RED yet R1 and S1 columns are puke-green the war will be fought but resistance is futile, as the R2 cluster dominates.

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

Now, with the background in place, it's time for some real observations.

First of all, there is real power when the underlying and it's leveraged counterparts all sing from the same sheet of music. That is, all the clusters tell the same story independent of leverage. We ALMOST have this with the S&P500 -- the index symbol $INX is a bit of an outlier, but the SPY, SSO (2x) and UPRO (3x) are all in alignment. These three are suggesting a combined cluster weighting of 7:12, favoring a move upward.

Unfortunately, the Russell 2000 and the NASDAQ are not confirming this.

$RUT is the Russell 2k index. The combined cluster weighting is 9:8, so we're balanced. The IWM, which I take as my benchmark, is 9:9 or balanced. UWM, the 2x leveraged ETF is 9:8, just like the $RUT. Only TNA shows significant support and suggests a potentially strong move up. Remember though -- UWM and TNA are pinned to the performance of $RUT, so I personally discount TNA as an outlier and disregard. The R2K is not favorable in either direction going into Monday's action.

The Q's present much of the same picture. $COMPX has a combined cluster weighting of 7:13, but QQQ is 11:7 and QLD is worse at 16:7. TQQQ is bearish too at 16:13. Ugliness, with a bias to the downside on the Q's. Despite this, everything keys off of the $COMPX which is pointing upward more than anything else in this group, so time will tell.

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

Hence, from this presentation, it appears that we could see some strength in the S&P500, more weakness in the R2K and a coin flip on the NAS. I look forward to tracking this for a few weeks in the open for you to see if we can see some trends emerge using the cluster/pivot techniques.

Regards,

pgd

[lisa]

paul,

just curious if you are looking at the index pivots alone or in combination with your ggt signals. if in combination, are your ggt signals supporting the pivot findings or conflicting and which would you rely more on if in conflict?

thanks,
lisa

[grems8544]

paul,

just curious if you are looking at the index pivots alone or in combination with your ggt signals. if in combination, are your ggt signals supporting the pivot findings or conflicting and which would you rely more on if in conflict?

thanks,
lisa

I've not combined GGT and pivots. The presentation below was purely of pivots with no respect to GGT.

If you go to the Dropbox folder for the ETFs on Friday, then look at the top of the Dashboard page, you'll see the Broad Index ETFs. All are long in one form or another, and SPY and QQQ signaled above average volume and price action on Friday, even though prices fell. This tells me that SPY and QQQ are still relatively strong in the larger picture.

GGT's time frame is different than the pivots. GGT's time frame could be quite long; as a stock transitions through clusters the weighting changes constantly. The two systems are different in this regard.

GGT methods, applied to an individual stock, do not tell us about overbought or oversold. Recall that GGT is asking the question: are the prices, volume, and rates of change of prices higher than historical levels when the stock has done better? Correspondingly, the fact that SPY and QQQ are "Affirmed Long" with Friday's close simply tells us that compared to history, Friday's action in those ETFs was bullish. Contrasting, John Person tells us that in an up trend a healthy range is S1 to R2 for long positions. Conversely, in a down trend, we can expect the range to be R1 to S2. Pivot analysis will tell us where we lie within this range -- GGT will not.

A good (unverified) setup would be an uptrend in general, GGT signalling "New Long", and the pivot/cluster setup indicating that we're at the lower end of R2-S1 (closer to S1). If the weighting were such that there was little overhead resistance, these could be good candidates. Definitely something to watch for.

Thanks for sparking some ideas.

Paul

[EB]

This is great work, Paul. You might want to check out some of the currencies, as pivots are used by many forex traders. I especially follow the EURUSD and its almost inverse counterpart, $DXY, the USD index.

I'll reiterate something Billy pointed out long ago about using indexes with pivots. The calculated index pivots will often diverge from slightly from actual traded instruments because the highs and lows of the index will be slightly off, as they are calculated maybe only every second. Inasmuch as a long term high or low is often one experienced under capitulation, this difference can be material on the higher time frames.

[TraderD]

Contrasting, John Person tells us that in an up trend a healthy range is S1 to R2 for long positions. Conversely, in a down trend, we can expect the range to be R1 to S2. Pivot analysis will tell us where we lie within this range -- GGT will not.

A good (unverified) setup would be an uptrend in general, GGT signalling "New Long", and the pivot/cluster setup indicating that we're at the lower end of R2-S1 (closer to S1). If the weighting were such that there was little overhead resistance, these could be good candidates. Definitely something to watch for.
Paul

Paul, this is most interesting. Is the main idea to look at an index position relative to its S/R 1&2 pivots or is it to look at its stock constituents' positions relative to their own pivots (ie, use it as a breadth/sentiment measurement in some statistical aggregate fashion)?

Trader D

[grems8544]

Here's the 2/13 EOD cluster strengths on the major indexes as well as their tracking ETFs:

12812

What a difference one day can make.

Friday's action left a bullish view in the S&P500 grouping, and rather indifferent views in the Russell 2K and NAS. See my post below.

Today saw the R2K up 1.4%, the DJIA up 0.57%, the NAS up 0.95%, and the SP500 up 0.68%.

With the close of markets on 2/13, the Russell is indicating little overhead resistance using the 20xATR calculation, and significant support. IWM weighting for resistance:support clusters is 7:19, as is UWM and TNA (note the bright green color coding). Note that IWM is 2% away from mR2 and is 1% from mR1, suggesting (perhaps) some upside room to move. This suggests a target near ~841.

In alignment is the S&P500, which is also showing a 9:19 R:S weighting, and this is supported by the leveraged instruments. The SPY is 1.5% from mR2, suggesting an upside target of ~1372.

The NAS is not set up the same as the R2K or the SPX, and it's showing a balance of 12:14 in the Q's, which are only 0.9% away from mR2.

Is the market headed up from here? Your crystal ball is as good as mine. This being said, the R2K and SP500 cluster groupings make a presentation that suggests we will continue upward, possibly dragging the boat anchor called the NAS with it, so from this presentation alone, you'd think so.

Balancing our new-found optimism is the fact that IWM's "Max Pain" for OPX is an 80 strike, which is below our present location by a significant amount. The SPY's MP point is about 130 or 131, again below our present location by a significant amount. The Q's MP point is at 60, again, a good distance below where we are.

Balance...

Think happy thoughts....

Regards,

pgd

[grems8544]

Paul, this is most interesting. Is the main idea to look at an index position relative to its S/R 1&2 pivots or is it to look at its stock constituents' positions relative to their own pivots (ie, use it as a breadth/sentiment measurement in some statistical aggregate fashion)?

Trader D

From a practical sense, I look at where we are within the S/R levels. Note though that multi-timeframe pivots make this more complicated (quarterly as a support and monthly as a resistance, for example), so I think that it's not as clear cut as John Person writes. Dis-aggregating back into one time frame doesn't necessarily help here either, as you could be anywhere within the S/R range depending upon the time frame chosen.

This is where I think the power of the weights of each time frame matters, so my work in this area has been trying to identify if there is truly an edge. Note that simply placing a limit order at the 3:1 buy point when a cluster pair is 1:2 S/R weighting isn't necessarily a sure-fire win strategy for major tracking indexes, as it seems to be quite random in performance. I'm still playing with this though, and in certain trending markets, it does seem to work nicely.

As to the 2nd part of your question, I've not looked at the constituent stocks of an index to see where they lie. Obviously, my indicator doesn't care if the input is the tracking index ETF or a stock, and there is nothing preventing looking at all the stocks and seeing which ones have the most favorable weighting structures. For example, if you take my "leaders universe" (12850), the best stocks going into the open on Tuesday are:

12849

Note that these are sorted using the column all the way to the right in ASC order. This "bal" column is simply the ratio of the Resistance columns to the Support Columns.

Also note that for a 3:1 price R/R within the range between the lower support cluster and the upper 1st resistance cluster, the column "RR" provides you a map of where you are. Noet that many of these are between R2 and R3, and again, according to John Person, this is SHORTING territory, not going long territory in an up trend, so this is NOT an exact science.

What is compelling to me though is to watch stocks with a yearly support and monthly overhead resistance, and see that we're only a fraction of a percentage away from the optimal "3:1" R/R. Look at TGI, TSCO, and SSS: each is within 0.4%, 0.3%, or 1% respectively of an optimal 3:1 buy point. This suggests that we could place a limit order at the level indicated in "BuyPt" and be done with it.

Conversely, with the upper stocks listed, and knowing that they are between MR3 and MR2, we could place limit orders to Sell to Open at the "ShortPt" level indicated. GILD is 5.3% away from this, FAST is 3.0% away, etc. You get the idea.

All work in progress, so I would not do ANYTHING with these stocks. This is simply a thought experiment at this time.

Regards,

pgd

[grems8544]

This is great work, Paul. You might want to check out some of the currencies, as pivots are used by many forex traders. I especially follow the EURUSD and its almost inverse counterpart, $DXY, the USD index.

Interesting. Here's the pair, and looking at it on a 5m chart, it's instructive to watch the prices move through the clusters.

12851


I'll reiterate something Billy pointed out long ago about using indexes with pivots. The calculated index pivots will often diverge from slightly from actual traded instruments because the highs and lows of the index will be slightly off, as they are calculated maybe only every second. Inasmuch as a long term high or low is often one experienced under capitulation, this difference can be material on the higher time frames.

This is the most likely explanation why the raw index does not match the tracking index ETF. From this I conclude that since we can only trade the tracking ETF, that the raw index clusters are less important and could be removed from the presentation with no impact on performance.

Thanks Bob.

pgd

[TraderD]

From a practical sense, I look at where we are within the S/R levels. Note though that multi-timeframe pivots make this more complicated (quarterly as a support and monthly as a resistance, for example), so I think that it's not as clear cut as John Person writes. Dis-aggregating back into one time frame doesn't necessarily help here either, as you could be anywhere within the S/R range depending upon the time frame chosen.

This is where I think the power of the weights of each time frame matters, so my work in this area has been trying to identify if there is truly an edge. Note that simply placing a limit order at the 3:1 buy point when a cluster pair is 1:2 S/R weighting isn't necessarily a sure-fire win strategy for major tracking indexes, as it seems to be quite random in performance. I'm still playing with this though, and in certain trending markets, it does seem to work nicely.

As to the 2nd part of your question, I've not looked at the constituent stocks of an index to see where they lie. Obviously, my indicator doesn't care if the input is the tracking index ETF or a stock, and there is nothing preventing looking at all the stocks and seeing which ones have the most favorable weighting structures. For example, if you take my "leaders universe" (12850), the best stocks going into the open on Tuesday are:

Note that these are sorted using the column all the way to the right in ASC order. This "bal" column is simply the ratio of the Resistance columns to the Support Columns.

Also note that for a 3:1 price R/R within the range between the lower support cluster and the upper 1st resistance cluster, the column "RR" provides you a map of where you are. Noet that many of these are between R2 and R3, and again, according to John Person, this is SHORTING territory, not going long territory in an up trend, so this is NOT an exact science.

What is compelling to me though is to watch stocks with a yearly support and monthly overhead resistance, and see that we're only a fraction of a percentage away from the optimal "3:1" R/R. Look at TGI, TSCO, and SSS: each is within 0.4%, 0.3%, or 1% respectively of an optimal 3:1 buy point. This suggests that we could place a limit order at the level indicated in "BuyPt" and be done with it.

Conversely, with the upper stocks listed, and knowing that they are between MR3 and MR2, we could place limit orders to Sell to Open at the "ShortPt" level indicated. GILD is 5.3% away from this, FAST is 3.0% away, etc. You get the idea.

All work in progress, so I would not do ANYTHING with these stocks. This is simply a thought experiment at this time.

Regards,

pgd

Paul, this chart is very useful and is roughly along the lines of my thinking since I first learned about pivots. I see pivots as junctions on a roadmap where price can take decisions as to where to go next, except that the signs at the junctions come at different sizes (small=daily,large=yearly, etc.) and hence convey varying level of importance.

On top of that, the signs aren't always in effect depending on what trip (=timeframe) one is taking. The longer trips mostly adhere to the larger signs. Since at any given time the driver is surrounded by cars that take different trips which together make up "traffic", I'm thinking this may be best handled by a statistical portfolio which implements "pivot arbitrage" where market bias is determined either by an external factor or by some statistical aggregate of the individual stocks' positions relative to their prevailing pivot cluster environment (that's what I was referring to by the second part of my question).

Like you, I am not sure what is the right way to mix R/S pivots from different timeframes, the weighted sum of closest ATR-sized cluster (or variation thereof) is as sensible as anything else. In summary, the concept is based on the hypothesis that "some pivots can be wrong some of the time but most pivots are right most of the time". I'd be interested to find out what the R2-R3 stats look like as a function of an external breadth indicator that sets the context for the trend.

Trader D