Setting up a gold bull, part 2

We have seen last time what was needed to set a gold bull market in motion. Lets follow through on the last week analysis and see where are we in the process and can we still say that this thing is developing in the right direction.

SLV

Two days before the SP500 generated its 92-96 true buy signal the SLV generated its true 35-39 signal. This signal came as a result of a failed 16-20 sell while overcoming a long term resistance from the April 2011 top and, also very important, while being on a bearish 92-96 sell that couldn't produce any downside.   I concluded that it seems the resistance has been broken and the odds are that the SLV is going to accelerate north. The odds worked favourably and SLV moved up almost 9% since then. I expect it to move further up to over 30 and stay there for a 92-96 buy signal. 

SP500

Beside three failed true 92-96 signals in the last two months the SP500 kept stubbornly grinding higher to generate yet another true 92-96 signal. It has happened on the 96th day after the April 2012 high and at the April high level. Given the timing, the price and overbought technicals it looks like a top. On the other hand, considering the prevailing negative sentiment and the action of the SLV chart I think there is a good chance that after some consolidation this chart breaks north to test the 2008 all time high.
The other option is that precious metals are disconnecting from the SP500 and they will rally while the SP500 is going into a dive. Considering the state of the dollar chart I think that this option is less probable.

GLD

The GLD has bought the 92-96 and has reached the April down trend line at the same time. It is trying to break out and it will if it closes above 162. However , it doesn't have to happen straight away, a steady sideways move for up to two weeks will do the trick. Selling and buying back the 92-96 along the way would be just perfect. 

UUP

For 65 days from its inception the UUP true bull 92-96 buy didn't manage to produce any decent upside. The 16-20 buy failed twice, same as 35-39 buy. This is not confirming behaviour and I expect this chart to drop further to test the 92-96 support. In the short term (2-3 weeks) it could go sideways/up to touch the 16-20 from the downside before it resumes sliding down.

HUI

The stage three of a true bull set-up has been achieved for the HUI, the 92-96 buy. There is a disclaimer though, beside coming very close to it it couldn't clear June high of 464.76 but I expect it to be done in the next couple of days.

The HUI setup is a bit different when compared to the SLV and the GLD.
The SLV clearly broke out of its May 2011 top down-trend but it has to go further to reach the 92-96 echo before it can try to break free.
The GLD has reached the down trend line and the 92-96 echo simultaneously and trying to break out.  It has bought the 92-96 but still needs to close over 162 to break the trend line. It would be nice to see it happen in the first two weeks of September.
The HUI, on the other hand, bought the 92-96 but is still far away from its down trend line which is positioned around 485. I think that proper positioning for the HUI break out requires it to reach this 485 level before a serious consolidation takes place. This resistance line should then stop the advance until the next 16-20 buy signal comes into the picture. The scenario that we should be cautious about is if we go strait down for the next ten days to below 422.47 without breaking 464.76. In this case there is a good chance we would get a double 16-20/92-96 buy but that will open a serious possibility for a triple sell next time we reach HUI 450. I don't like this possibility because structurally it is bearish.
The conclusion is that once the HUI 465ish has been broken it will confirm that the initial leg up out of the bottom was an impulsive wave. That is in accord with the premise that the May low is a long term low. The ensuing long term rally should last at least 710 trading days.

On the premise that this leg up should set the stage for a gold bull I think that the HUI has some unfinished business around 485. For that reason I am still 50% long. I have restructured the holdings to get rid of shares that reached the targets and replaced them with those with more potential. If I am wrong and we go straight down to 16-20 buy I will sell when we reach the current level again and then wait to see if we are going to avoid a possible triple sell. If I am right I will reduce my position to 25% long at HUI 485 (optimally 495.70) and wait for the 16-20 buy to go long 100%.

Setting up a gold bull market

Long time ago I have written an article on the structure and dynamics of the gold bull market as I see it through the glasses of the SKI indices. The proper setup has a few stages:

1. Proper index alignment: 16-20 bellow, 35-39 in the middle and 92-96 on top. To achieve this order the price has to go down for an extended period of time, usually a couple of months.
2. Forming a bottom. The bottom is signalled with a failed 16-20 sell signal.
3. The third stage is a confirmation of the long term bottom. It is manifested as a rally that has the structure of an impulsive initial wave up. A proper confirmation rally will buy the 92-96 echo because without this buy the true bull is not possible. I want to keep it simple so I will not go into technicalities of this requirement.
4. The next stage is a breakout. It is made of a quick 92-96 sell and a rebuy. This event is a proof of strength of the impulsive wave and the measure of its quality is the absence of a 35-39 sell.
5. After the successful breakout the price moves up while the 92-96 is dropping hence they are forming a ‘scary’ picture where it seems like the price is lacking a firm support. This picture represents the scepticism that should be predominant sentiment throughout the whole move until the bull is widely recognised.

Let’s now analyse some charts while having this description in mind.

Aussie gold shares index

So far this chart is satisfying the criteria to setup a bull market. It had had a 21 month long correction that has put indexes in the right order (stage 1). Along the way it had numerous failed true 16-20 buy signals until eventually the one at the bottom succeeded. The sell signal that ensued failed as it should (stage 2). This signal was followed by a 35-39 buy executed yesterday. This pattern signalled a bottom but we need the price to rise to the 92-96 index/echo and buy it to confirm the bottom and to confirm the impulsive nature of this initial rally. Quite a long way to go compared to other gold indices like HUI but the Aussies were hit the hardest and they have to work the most for the proper recovery. If the 92-96 is going to be bought on this move up, I think it will be around $5500 and in a form of an inverted head and shoulders pattern. That will reconcile very well with the traditional TA support/resistance lines as it can be seen in the picture.

All of my holdings are in Aussie stocks and I am very happy since they are in profit between 20% and 40%. For most of them there is another 10-20% rise before they will be able to confirm the bottom.

HUI SKI

The HUI is similar to Aussie gold stocks except the bottom happened in May and the initial rally was stopped at the 35-39 buy signal. It then had to sustain a drop to a higher low to be able to rearrange the indexes in the right order and then buy the 35-39 again and reach the 92-96 echo. We need that extra push now to buy the 92-96 to be able to setup a true bull market. The resistance exists between $443 and $460 and will be interesting to watch how the price is going to behave while interacting with this range. An immediate push up over $460 and stay over it for 3 days will buy the 92-96 and almost warranty attack on $480-490 before a serious correction.

GLD SKI

While the stocks are leading the rally up the GLD is resting in a sideways correction. Needles to say this is traditionally the right way to start a sustainable gold rally. The price is hovering above the short term indices and approaching the moment of decision. The 92-96 is already set properly for the bull market, we just need the 92-96 signal. By mid-September the time will be up and I expect the volatility to increase dramatically after that.

SLV SKI

The same as GLD except it seems like it has already overcome the resistance from the May 2011 high and is about to move quickly to $30 to test the 92-96 resistance. Check how it, since the 28^th June low, has been almost glued to that long term support line. While doing it, it was also working the short term indices so we have the failed 16-20 buy signal and now the 35-39 buy executed yesterday. I think this should accelerate now to challenge the 92-96 echo soon.

S&P 500

Ok, here is the problem. What to make out of this chart??? I have analysed it in previous updates to come to the conclusion that it doesn’t look good. It has generated three failed true bull signals on the way up which normally indicates a rollover top. Surprisingly it has managed to move even higher and reach the top of the previous medium term rally and is about to give the fourth true bull signal. The problem with this is that this time it happens that the 92-96 echo has the right structure to allow for this signal to survive because it will be in a sharp fall for about two months. The 16-20 and 35-39 will act as a support and there is a chance that we will see higher prices from here. If tonight S&P generates a 92-96 true buy signal it will happen exactly on the 96th day after the April high. These kinds of coincidences always mean something important. Is it possible that we see some kind of a melt up in the S&P price while gold is setting up a true bull and then they disconnect and gold goes up while S&P goes down? I am thinking of this possibility just because I cannot really see gold setting up a proper bull if the S&P crashes and burns right here. To remind you the UUP (dollar) is on a true bull but it is not performing well. Maybe the scenario is this: S&P melts up and buys for a true bull, gold sets up a true bull while the dollar is testing its own. Then, when everyone gives up on the S&P crash (which is predominant sentiment today), we finally see it burning while gold and dollar stay on their bull signals and move up in tandem.

It has been ten years I have been trading gold and gold stocks and this is probably the most interesting setup I have seen so far.

I am 50% long, sitting on a solid profit and waiting to see if we are going to move higher to test the 92-96 echo. If and when the buy is generated I will reduce the position to 25% and try to go 100% long at the end of the next correction. On the other hand, if the S&P crashes here and now and the gold follows I have my trailing stops in place and will let them take me out of the market.

Another positive week ahead?

Some minor bullish development for gold can be spotted among the charts I follow.

Australian Gold Index

Aussie gold chart is interesting to watch because it topped before the other did (Nov 2010) and it went deeper in red than the other did. On the monthly chart it is more oversold than it has ever been since 1997 including the 2008 crash. All my trading  positions are in Aussie stocks, currently 50% long. At present it is on a true 16-20 buy and is about to sell for profit if today closes around $5000. I am not going to sell a single share before I see the 35-39 buy and probably rise to challenge the 92-96 echo. $5433 is the major level and it needs to break above it to establish new long term bull trend.

UUP dollar long

Since I reported last time on this chart its action was mostly sideways but with short to medium term bearish bias.  The true bull signal is not under threat but the behaviour around the 35-39 is not confirming so far. When it seemed it was going to avoid a double sell it dropped sharply to violate the 35-39 echo and sell it for a double sell. It, then, tried a recovery and two days rally produced a new 35-39 buy but it wasn't long lasting because on Friday it collapsed again and it now seems it will sell to confirm a medium term bearish period that should lead to the 92-96 support test. I interpret this behaviour as not confirming the true 92-96 buy from 50 days ago. It is still not too late but this bearish behaviour should stop and it must produce a convincing 35-39 buy that is going to last and lead to a new high. For this to happen Monday and Tuesday should keep it above $22.50.

In my last report I thought that chances were 60:40 for the bullish scenario to develop but now I think this is going further down to test the 92-96 echo. That test would be in line with a bottom in gold and is going to decide if this chart has a potential to transform into a real bull.

S&P 500

Since my last report S&P has had another failed true bull signal but it maintains the up trend and it seems it will be able to follow the 92-96 echo to the top around $1420. What happens after that is the question? To me it still looks like topping out but if it can be at 1420 mark in about a week the structure of the 92-96 echo can give it another true bull signal but this time with a much better chance to keep it alive even if a significant correction ensues. This is  especially interesting in conjunction with the SLV chart which could give a significant buy signal just at the same time when S&P reaches the previous high and produces a true bull signal.

SLV

The SLV chart has shown some serious resilience in the last 2 weeks by avoiding breakdown and slowly building a base for a move higher. It has sold a 16-20 echo but keeps grinding higher so it is now, 10 days into the signal, almost 3% in green. See how all is happening while breaking the resistance line from the Apr 2011 high and touching breaking the 35-39 resistance. This looks bullish and if it buys the 35-39 echo in a week time at the same time when S&P reaches $1420 it could be a signal that the S&P is not going to top out as everyone is expecting. 

GLD

The GLD is moving sideways and breaking the immediate resistance from the Feb high but more importantly it seems stuck in-between the 35-39 as a support and the 92-96 as a major resistance. This can last only till the middle of Sept when it will be forced to break through the 92-96 into the bull territory or to follow it down and break bellow $150. 

The development over the last two weeks seems positive for the bulls. Grinding slightly higher or sideways for another month is setting up for a major decision event in a middle of Sept 2012. I am happy to keep my 50% long position at least until the prices of different indices reach the 92-96 echo. The stops for all my positions are now in place and they are set below the recent lows. For Aussie shares that is mostly July lows. For those interested I have positions in NCM, RSG, OGC and IGR.

Good luck to everyone.

My approach to markets using SKI indices

My approach to markets using SKI indices

The way I am using the SKI indices is materially different from Jeff's. My approach is not through the statistics. I have been a creator of many systems based on statistics and have found that a major down side of this approach is that when the market changes direction or there is a new force in the market, that significantly changes the market behavior, it takes too long before it is reflected in averages. Usually, mechanical systems based on pure stats fail in the most critical moment when the best opportunities arise or when the market changes the gear. Simply put they need time to adjust to new conditions. If we take the current bull market in gold as an example, with so many unprecedented events recently, we will probably need 30 years before all what is happening now became a part of the SKI stats. I do not have that time.

I had been looking for years to find a tool that doesn't have this fault and I found it in the SKI concept. I use the word ‘concept’ because I would like to distinguish between a broad abstract idea of the SKI mechanics and the SKI system which, for me, is an implementation of this concept. There is one concept and many possible implementations/trading systems that can use this concept to create a tradable system.

The SKI concept comprises three most important features that you can wish for in a TA indicator:

1)     It is self-adaptable to changes in market volatility. This is because, to produce signals, the price interacts with original price values rather than with average values. The consequence of this is that the SKI indices don't need time to adjust to new market conditions! If the life run low of 2008 was unprecedented, as it was, the next time the price interacts with it in, for example, 92-96 index/echo it will interact with this exact low and its exact amplitude and volatility, not with an average of all life run lows over the last 30 years. For this reason we could assume that any system based on the SKI indices is self-adaptable to market conditions hence very robust and resistant to volatility changes.

2)     It can integrate multiple time frames. SKI concept implements the cyclical nature of markets and humans through 16-20, 35-39, 92-96, 218-222… indices that I call 'echoes'. I don’t like the word index because somehow it doesn’t reflect its most important quality: it is the original price just shifted forward in time. Combining these echoes allows us to construct complex market images with multiple time frames overlaid in one clear picture. I do not know for a better tool to define if the market is in a bull or bear formation.  MACDs, MAV crosses, RSIs… work to some extent but because of their nature (average) are slow and prone to failure in volatile markets.

3)     It eliminates noise while producing a signal. The concept of 5 back prices is a concept of weekly implemented into daily time frame to eliminate the noise. It is brilliant and works very well.

Here is how I work. I deploy SKI and other indices to a desired set of data. I mark traditional support/resistance lines. Then I look for important levels and time periods. An important level is where support resistance line meets SKI echo or two or more SKI echoes converge or a combination of any of these. An important period is where price faces an echo that represents important event from the past (important breakout, top, bottom, trend...). After I have decided where the important levels/periods are I watch BEHAVIOR around them. The third element is the overall sentiment that prevails through media and forums I read. That is it. I do not need 30 years of stats to analyze an unprecedented situation like we had many times since the LR low. The SKI indices are telling me what are the important points and the behavior around them is telling me what TO DO not WHAT IS GOING TO HAPPEN. Mix in a little bit of risk management and there you go.