TS Zigzag indicator (TSZI)

�2014 Sergey Tarassov

This tool can be used as an indicator that helps to specify turning points. I recommend to use it together with Welles Wilder Relative Strength Index (my favorite TA indicator). Look at these two indicators together, RSI(14) and TS Zigzag indicator (TSZI) calculated for S&P500:

So you can use this indicator exactly as RSI indicator, i.e. it allows to specify oversold/overbought zones accordingly and reveal potential turning points.

What is new in TSZI?

1) This indicator is based on totally different approach, kind of cyclical analysis. This is important because it allows to look at the stock marked from a different point of view.

2)  This indicator is prolonged into the future for some time period (see forecast horizon option):

3) It allows to decrease one of  side effects of RSI indicator: when the stock market is in trend mode, RSI may be in overbought/oversold mode for a very long time, till the trend ends, like on this picture:

TSZI - does not indicate overbought signals there.

How to calculate this indicator?

Run "TI" -> "TS Zigzag Indicator" nodule. Start with setting the minimal zigzag swing that will be used for calculations. In this example for S&P500 (2001-2014), 5% zigzag reveals 166 turning points (CIT); the average distance between succeeding turning points is 28 days, so the forecast horizon of this indicator is 28 calendar days, i.e. TSZI looks into the future for 28 trading days ahead:

Click "Calculate", and this indicator will be displayed in the Main screen. Pay attention, the calculations here are very sophisticated plus we apply non future leaks algorithms there, so it needs some time to calculate.

We also recommend to vary the critical value for overbought/oversold zones:

It may be different for different financial instruments.

Explanation

This is one of the most interesting indicators I met in my life. Firstly, this indicator is based on cyclical analysis in its extremely simplified form. This idea is so simple that during 20 years that I do cyclical analysis for stock markets I have not pay proper attention to this approach. From another side, this oversimplified indicator has some critical nuances that I could simply miss them if I would not do cyclical analysis within the last 20 years.

Ok, here is the description of this idea. Below is a piece of S&P500 price chart shown, and we want to know the probability of price reaching its top on October 30, 2014:

Let us do timing using analogies in the past. Look at this: the distance between our date (October 30, 2014) and the last important top (September 19, 2014) is 41 calendar days (CD). The program looks for the analogies in the past and shows that 41 CD distance between TOPs  sends us back to the years 2011 and 2001:

It tells us that in 2011in July-August we had 41CD distance between two succeeding TOPs, here it is:

Also, in 2001, in October-December, the distance between TOPs is 40 trading days (1 day accuracy is used):

So for October 30, 2014 we have two analogies in the past if we compare two succeeding TOP turning points.

We can include into our consideration other analogies. As another example, take the distance between LOW on October 15, 2014 and October 30, 2014. It is 15 calendar days:

The program shows that in November-December 2007 the distance between succeeding BOTTOM and TOP was 15 calendar days as well: 

In total, we have 3 analogies in the past for October 30, 2014; this date has three confirmations.

For the next trading day, November 3, 2014, we have four analogies in the past:

This is how this indicator works: it simply shows the amount of analogies in the past. 

The new possibility to make a forecast beyond the last price bar gives this technique a new dimension.  Look at this example:

 On January 28, 2014, TSZI entered oversold (blue) zone. Our asset is becoming undervalued, so we are expecting BOTTOM. We also watch the forecast, and it shows that within one week TSZI will leave blue oversold zone: 

It means that within a week it should be clear what SNP500 will do. It could continue its drop or a reversal BOTTOM point will occur. This is what has happened in the reality:

You see, TSZI predicted very accurate the RSI movement within the oversold zone.  Also it is important to mention that to calculate this example we used NON FUTURE LEAKS algorithms. The program did not use the price history beyond January 28, 2014 to calculate indicators.

The simplicity of the basic idea was compensated by very delicate and tricky algorithms that release that idea into ready-for-usage indicators. 

Firstly, we should remember that a filtered zigzag is an indicator with future leaks (we can exactly specify that some turning point has occurred when it is confirmed by an opposite movement; it means we receive a confirmation of the turning point too late, when we can not make any profit of it. We cannot have both - a confirmation of the turning point and a profit trade - at the same time). Calculating TSZI indicator, we avoid that problem applying a special non future leaks algorithm. (I want to emphasize the importance of avoiding future leaks - just in case somebody may decide to develop a similar indicator using standard scripts from any TA based software.)

Also, some other tricky nuances are hidden under this option:

So far, these nuances are for private usage only. They are related to technologies developed and used under Terra Incognita project of Timing Solution.

November 3, 2014

Toronto, Canada

Sergey Tarassov