The TSI Trader. How to trade the stock market using the True Strength Index

Darn it if I didn't wake up today an hour after the NYSE began trading and miss out on having some serious fun. It's my birthday today and besides, I don't have to go to work in the summer so, well, I'm sure you understand what motivated me to snooze. But the old cliche still rings true and I'm going to show you the opportunity I unfortunately chose to lose.

The opportunity had to do with trading something gold related like Direxion's Gold Miner Bull/Bear 3X ETFs (NUGT /DUST). The True Strength Index (TSI) indicator was handing out the trades on a silver platter this early morning, requiring no new tricks but the use of the basic techniques we have talked about for the past two years.

However, I am going to show you something quite interesting about the TSI indicator that I have recently figured out. If you read on we'll be to that discussion before you know it.

So anyway, here is the 1 hour chart of gold futures (GC) and the first trade this morning was a SELL using the negative divergence technique. Gold screamed right past its previous price high of 1597.3 (I use this bar because it correlates to the TSI high shown directly below) and reached 1602. Well, that wasn't going to work. Additionally, the lowest indicator panel continues to show that money is coming out of gold, not into gold. So then gold drops and drops and drops.

Click on any chart to ENLARGE

Great trade and I slept right through it. But then, gold drops so far that it blasts right past its previous low of 1582.7 and reaches 1581.3. Good heavens. Now we have a positive divergence BUY signal.....time to cover the short position and go the other way. Oh yeah, I was still asleep.....or more likely just barely awake.

Anyway, on this chart and the one that follows I would like to show you that *quite interesting* something I promised to discuss if you kept reading.

You'll notice I am showing a pair of moving averages on price and configured their colors to identify their respective up/down direction movement.

And, you'll also notice that the location of the trend line break BUY/SELL signals torpedo price right about or just before these two moving average begin to change direction.

Here is another chart of gold (GC) but this time on the 4 hour time frame. I wanted you to see another example of the moving average behavior vs. TSI trend line break signals. Also note: the moving averages chosen are the same as my TSI setting (7 and 4).

It is clearer to see what I am talking about on this 4 hour chart (above) as price is not wiggling around as much as on the 1 hour chart.

This *quite interesting* discovery still does not explain the mystery of why the trend line break technique works as incredibly well as it does, but it does somewhat explain with clarity what is actually going on at or around the time of the signals.

I have asked some really smart mathematicians to explain to me why the trend line break works as well as it does. Unfortunately, they have no idea. Seriously.

But maybe you do. And if so, PLEASE tell me!

Best always,

John
tsiTrader@gmail.com

This article will begin with a close up look at the recent rallies off the May 16th low of both gold futures (GC) and Market Vectors Gold Miner ETF (GDX), proceed with some fascinating data regarding the time ratio of all C-wave tops with respect to gold's repetitive ABCD pattern, and conclude with 5 charts focused on the US Dollar Index (DX) from 2008 and forward. The sum of this discourse will likely leave the reader with the notion that gold has probably not yet bottomed, the US Dollar has probably not yet topped and the reader will gain an awareness of the time frame that may be necessary before these metrics are behind us, rather than before us.

Also, this article will demonstrate a True Strength Index (TSI) indicator technique that I have used successfully for a long long time, but don't think I have ever written about. This technique will show you how you can use a specific TSI trend line to project the timeline for a long rally's conclusion (peak) far in advance of the actual occurrence.

So let's get started with this daily chart of the gold futures contract (GC) and make a few quick observations.

Click on any chart to ENLARGE

Here we see that the May 16th low of 1526.7 has been followed by one rally that eventually retraced itself by 78.6%, and now we are in a second rally that on its 6th day has managed to retrace itself by 61.8% then bounce to closer to a 50% retracement by the close. The TSI has given timely BUY and SELL signals, and at present is reading just above ZERO and no new signals are in place.

Next is the daily chart - GDX.

GDX tells much the same story as gold. First rally off the May 16th low makes a slightly stronger showing with a 61.8% retracement vs. gold's 78.6%. The second rally has, in agreement with gold, now made a 61.8% retracement and then bounced up some at the close last Friday. The TSI has been spotting the BUY and SELL locations well with the trend line break technique and currently reads just below ZERO at -2. This reading is technically a SELL signal using the ZERO crossover technique of identifying BUY and SELL signals.

At this point, things don't look real bad, but they are not great either.

Both GC and GDX have now made a succession of 3 lower highs, gold has not closed a single day above its descending 200 dma in 47 trading sessions, GDX has been below its descending 200 dma for 90 trading sessions and has not been able to close above the descending lower megaphone trend line for more than a day or two before being pushed back below.

So this brings us to my research on the time ratio of C-wave tops with respect to the entire ABCD pattern in which they are found. If you are wondering why this item of trivia was of interest for me to discover, the question is this: is there a common relationship between the C-wave top and the amount of time, a ratio of the entire ABCD pattern, that the D-wave needs to conclude its corrective work? In other words, are all D-waves similar in their time ratio with respect to the ABC-waves which precede it?

To arrive at the answer, I used a weekly chart of GC and counted and counted and counted. If I miscounted something by a week it was because my eyes got blurry and I really just wanted to just know the general answer to my question and not try to earn a Ph.D. in counting. Three college degrees is more than enough for me, thank you.

Here is what I found out:

June 2002 C-wave topped on week 24 D-wave bottomed on week 32 C-wave topped at 75% of this ABCD	February 2003 C-wave topped on week 26 D-wave bottomed on week 35 C-wave topped at 74.2% of this ABCD
The early 2004 C-wave had a double top. The second top was 50 cents higher than the first top. Here are both ways of looking at this:	January 2004 C-wave topped on week 39 D-wave bottomed on week 58 C-wave topped at 67.2% of this ABCD March 2004 (2nd top) C-wave topped on week 52 D-wave bottomed on week 58 C-wave topped at 89.5% of this ABCD
November 2004 C-wave topped on week 29 D-wave bottomed on week 39 C-wave topped at 74.4% of this ABCD	May 2006 C-wave topped on week 65 D-wave bottomed on week 70 C-wave topped at 92.9% of this ABCD
March 2008 C-wave topped on week 92 D-wave bottomed on week 123 C-wave topped at 74.8% of this ABCD	RECAP 2002 - 2008: 75.0%, 74.2%, 67.2%, (89.5%), 74.4%, 92.9% and 74.8%

Clearly, 2006 at 92.9% was an anomaly. The double top 2004 situation was also unusual because it was, well, as I mentioned, a double top (duh!). Otherwise, 75% sure looks like the 'typical' placement of a C-wave top within an entire ABCD pattern.

So with that perspective, let's look at the data for the current ABCD.

September 2011
C-wave topped on week 150
D-wave bottomed on week 166 (using December 1523.9 low)
C-wave topped at 90.0% of this ABCD

90.0% seems a bit too early to me, so let's see where we would be if next week provides the true D-wave bottom. Remember, to accomplish this gold would have to trade intra-day below 1523.9 sometime this upcoming week.

C-wave topped on week 150
D-wave bottoms on week 194 (next week)
C-wave topped at 77.3% of this ABCD

Hey, now we are getting close to that 75% time ratio that is more in line with gold's past ABCD patterns.

And just where would we be if we use 75%, you ask? Well, just a short ways before the end of August, is the answer. The C-wave topped on week 150 and when/if we get to week 200 for a D-wave bottom, that makes the 75%.

This conjecture is just so much nonsense if there is no reason to believe that gold has not already bottomed.

Can you think of any valid reasons to seriously entertain the notion that gold may very well have not bottomed?

Aside from all my bean counting and trying to offer some respectable rationalization, I can indeed.

My notion has to do with a proven catalyst for gold's price behavior - in both the up and down directions - and that catalyst is the US Dollar Index (DX). In fact, the outcome of what I am about to show you is truly going to make all the difference regarding where gold goes from here.

So let's continue now with a daily chart of the DX from the 2008 time period to the present and get a feel for what this thing is capable of doing and figure out what it looks ready to do next.

The highlights I wish to bring to your attention include the magenta colored slightly descending trend line connecting past tops and cruising directly overhead at around 87.00, the possibility that the dollar has been in the consolidation phase for the past couple of months in what looks like a T1 continuation pattern with a price projection some 400+ basis points higher, and in the lower indicator panel we view the history of the buck's ability to rise above its 200 dma before being brought back to earth.

The takeaway here is that overhead resistance is at 87.00, a T1 pattern has made half its move, consolidated, and ready to finish the other half of its move towards a target of 86.50+, and the dollar, currently just 4% above its 200 dma is no where near overbought if it should decide to start rocketing higher. In fact, the completion of the T1 pattern would put the dollar right about at 10% above its 200 dma and combined with the overhead (magenta) trend line would likely send this rally to the showers if/when it gets there.

Here is a look at the dollar weekly chart also beginning with 2008, but this time with some TSI thoughts.

This is a view that is a little less cluttered and puts the magenta colored down trend line as well as the T1 pattern setup in perspective. Those reader with a sharp eye for how to use the TSI will note that currently we have the potential - with a very small basis point move of less than 10 - to make a higher high on price while obviously making a lower high on the TSI indicator. This is a negative divergence SELL signal.

But I'll tell you this. Considering we could be talking about a fully fueled T1 pattern breakout to the upside, if the dollar has that plan in its head it will just run over the negative divergence SELL signal. This happens on extremely charged highly volatile rallies, as we see was the case in early 2010 (orange rectangle). The normal expectation is what followed this specimen in March 2010 (blue rectangle). Here, the negative divergence indeed caused price to pull back for 4 weeks. This in turn created what was needed to draw a trend line break BUY signal when price resumed its spirited rally.

Finally, let's take a closer look at what the dollar index did in 2008 and 2010 and how it is poised for further action in 2012.

Here is the daily 2008 dollar using a Fibonacci time ratio measurement. This stuff always amuses me because I am amazed at what I find. The first top is exactly 61.8% of the second top - how about that? And the TSI nails both tops with incredible accuracy.

Oh, here is that technique I promised to tell you about .... circled with green. The two low points that are surprisingly close together create the trendline that I think of as the 'master TSI trend line' for the entire rally. When the TSI finally crosses back down through this trend line, it is always game over. Another day I will devote an entire post to this subject with lots of examples and then you will more fully understand how to spot it in your own work.

Here is the 2010 chart. Again, the 61.8% peak that precedes the final exciting rally, and some other stuff. This time the TSI misses the intra-day rally high as it was late 3 days. Oh darn.

Last chart - here is the current dollar action. We got a new trend line break BUY signal on this past Friday. There are a pair of underlying TSI trend lines that, when crossed, will spot the top. That does not appear to be likely any time soon. The Fibonacci time ratio came out differently on this example. I ended up using the January rally top as the 38.2%, the late February low as the 50%, and that left me with the 100% measurement landing on August 27, 2012.

In summary, gold has been below its 200 dma for so long that it is behaving as though it is still in a D-wave and not in the early daily cycles of a C-wave. Gold has not (yet) retraced at least 50% of its C-wave, as all other preceding D-waves. A price in the $1400 range would take care of that. And finally, the history of C-wave tops measuring near the 75% time marker of the ABCD pattern will not be paralleled until sometime in late August, at the soonest.

The US Dollar Index appears to have been consolidating a T1 pattern that projects it 400 or more basis points higher (which 'coincidentally' is where a major trend line awaits). Also, the dollar received a daily chart TSI trend line break BUY signal on Friday. And, the dollar index is not overbought in terms of its price relationship with its 200 dma, and though admittedly not terribly compelling, a case using the Fibonacci time ratio can be made for a late August top in the dollar.

Will all of what I have written come to pass? Will any of it?

Heck if I know.

But you and I still have to make investing decisions and live with the consequences, either way. At the moment, I plan to sell anything that can be liquidated at close to break-even, and hold the rest - come hell or high water, as they say.

I wish you are great week and honestly hope I have this entire thing all wrong. I would much rather see my positions go up in value, than the opposite. Keep in touch, OK?

John
tsiTrader@gmail.com