Basically you are doing two things wrong currently.

The good news is that I can help you with both of them.


One is not getting the exit right.

You need to minimize human interaction to be successful. You may need to leave one side to be handled by the programs entirely.

The following is one of the most advanced piece of software I have.

This cropper has directional logic fitted derived from RSI8 projections and even monitors for the behavior of the Green River and Mr. Maroon. It adjusts the 30-min RSI8 cross-back-crop value based on the direction it figures, and only closes out positions that are in gains by at least 10 pips.

It was of course invented with EUR/USD being the test subject.


I would go into details of the code in my book, that may be as far as a year out from this point in time.


The second thing is the entry, the “Anchorage”. Sometimes I think of it as a lift call button: “we want to go down now” – they pressed the button.

It is basically a telltale sign of an initial buy/sell having taken place.

You compare 30-minute fractals on one side to find an RSI2 divergence (i.e. higher high with a lower RSI2 reading).

This condition sets the area beyond the divergence block to be the greatest entry area.


(The yellow and blue ribbon arrows are plots from the directional logic.)

The off white swing line down is the wave1 that comes from outside the Anchorage.

Would I write an automated trading routine based on  the Anchorage? You bet I would.


Overall, I wanted to get your thinking to start down the path of “the condition projects the level” rather than trying to guess a line based on fancy.


“This was a clash…”

The Perforation

To turn my RSI8 system into a truly mechanical one, I now plot the lines and have a perforation margin as well, 8 pips away. The perforation should not be violated or you must reverse your direction.

As part of the system, you can instate rules such as a new high/low read sets a new break out point.

Such a reset happened on Friday, and was overrun the same day – and even back tested later on.


Basically, if you were long already, you simply had to stay long. There are only three things that can reverse the price back down now:

1. Another 82+ reading, that sets the  RSI limit (and the perforation), and later price makes it into this field with an RSI, that is out of overbought (<77).

2. Price makes a run for it, it tags the 1.18-1.1820 current sell field, and RSI does not get overbought (<77).

3. The opposite of the following picture happens: you start getting Sync-Async sequences, but on the upside.


Currently I only call Async highs qualified if they are at least 6, but not more than 10 bars away from the Sync high.

As for perforation in my life, just yesterday I decided on wanting to write a book on computer aided trading.


Computer Aided Trading

Welcome to the ERA of the Hybrid trading, dear reader. There is a never ending quest of finding new edges to boost bottom lines. We, the sole traders are starting to use computers in a more active manner, we entrust more and more of the decisions to programs for we have come to know intimately, that we are very limited in our abilities.

The programs are there when you cannot be, they can interact instead of you, and if they are set right, they can prevent losses or make gains on their own initiative.

There is one more thing that seldom anybody talks about: the advantage of making automated trading routines is that it is an active conversation, where you, the programmer would end up improving in skills, you would be able to find many of your own personality flaws, and your thinking would be stimulated into fining treasure mode.



I implemented my ideas of peak and bottom RSI undercuts today in the form of plots, as follows:

for (i = 500 ; i >= 0; i–) {
if(RSI8[i+1]>highrsi && RSI8[i]<highrsi && iFractals(Symbol(),0,MODE_UPPER,i+1) ){
ObjectCreate(“txr 30″+High[i+1], OBJ_TEXT, 0, Time[i+1], High[i+1]+.0003);
ObjectSetText(“txr 30″+High[i+1], “85”, 32, “Arial Black”, Orange);

ObjectCreate(“txr 30″+High[i], OBJ_TEXT, 0, Time[i+1], High[i+1]+.0014);
ObjectSetText(“txr 30″+High[i], “r8/S:”+DoubleToStr(NormalizeDouble(High[i+1]+60*Point,5),4), 32, “Arial Black”, Orange);

if(RSI8[i+1]>highrsi+5 && RSI8[i]<highrsi+5 ){
ObjectCreate(“txr 30″+High[i+1], OBJ_TEXT, 0, Time[i+1], High[i+1]+.0023);
ObjectSetText(“txr 30″+High[i+1], “90”, 32, “Arial Black”, Orange);

ObjectCreate(“txr 30″+High[i], OBJ_TEXT, 0, Time[i+1], High[i+1]+.0054);
ObjectSetText(“txr 30″+High[i], “r8/S:”+DoubleToStr(NormalizeDouble(High[i+1]+360*Point,5),4), 32, “Arial Black”, Orange);

if(RSI8[i+1]<lowrsi && RSI8[i]>lowrsi && iFractals(Symbol(),0,MODE_LOWER,i+1) ){
ObjectCreate(“txr 30″+Low[i+1], OBJ_TEXT, 0, Time[i+1], Low[i+1]-.0003);
ObjectSetText(“txr 30″+Low[i+1], “15”, 32, “Arial Black”, Orange);

ObjectCreate(“txr 30″+Low[i], OBJ_TEXT, 0, Time[i+1], Low[i+1]-.0014);
ObjectSetText(“txr 30″+Low[i], “r8/B:”+DoubleToStr(NormalizeDouble(Low[i+1]-60*Point,5),4), 32, “Arial Black”, Orange);


if(RSI8[i+1]<lowrsi-5 && RSI8[i]>lowrsi-5 ){
ObjectCreate(“txr 30″+Low[i+1], OBJ_TEXT, 0, Time[i+1], Low[i+1]-.0003);
ObjectSetText(“txr 30″+Low[i+1], “10”, 32, “Arial Black”, Orange);

ObjectCreate(“txr 30″+Low[i], OBJ_TEXT, 0, Time[i+1], Low[i+1]-.0014);
ObjectSetText(“txr 30″+Low[i], “r8/B:”+DoubleToStr(NormalizeDouble(Low[i+1]-120*Point,5),4), 32, “Arial Black”, Orange);




where lowrsi is 15, highrsi is 82

The Kijun-sen Experiment

26 sample… The industry standard.

True Blue = Kijun-sen

Yellow bands – 5 pips away in trending mode, 16 pips away when flat lining – see code below.

Once you find the fake out, it is easy to take a ride with simple set of rules. The most basic one is: until price gets back on the same side again.


The second thing you can do is to count the violations back into the field after price went outside. This can take you closer to maximum gains.


Could we assume that the third touch would come with a crack, and price is in a wave 5 up?


One more thing. People think of head and shoulders patterns as picture perfect things with neckline and all. Totally wrong. The magenta “A-1561” print on the bottom was a right shoulder. The two bottom “S” prints are the left shoulder and the head. If you held shorts and wanted to take profits, the undercut of the first “S”ync candle’s low was the perfect place – the undercut of the last “below 15” RSI8 reading, and the number would had been a guess. 10-12 would had worked this time. Other times there is 15 pips or more. The point is, this would had been done as an anticipation, and sometimes this would mean missing out on a continuation.

Now, closing shorts and going long are two different things. When the “A”sync low was printed, you had – in this instance – 9 hours to build a long position at & below 1.1561.

Good eye, Suthmo for finding the reversal.

for (i = 300 ; i >= 0; i–) {
if ((iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2==(iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,9,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,9,i)))/2) {Kijun_Plus[i]=(iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2+.0016; Kijun_Minus[i]=(iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2-.0016;}
else if ((iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2==(iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i+2))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i+2)))/2) {Kijun_Plus[i]=(iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2+.0016; Kijun_Minus[i]=(iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2-.0016;}
else if (Close[i]< (iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2) {Kijun_Plus[i]=(iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2+.0015; Kijun_Minus[i]=(iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2-.0005;}
else if (Close[i]> (iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2) {Kijun_Plus[i]=(iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2+.0005; Kijun_Minus[i]=(iHigh(Symbol(),0,iHighest(Symbol(),0,MODE_HIGH,26,i))+iLow(Symbol(),0,iLowest(Symbol(),0,MODE_LOW,26,i)))/2-.0015;}



When Volatility Dries Up

This mostly applies on the upside.

When you see the 3rd “Sync” swing high in a short period, chances are that you are in a wedge. There are 2 different shapes that can make price reverse. A head and shoulders and a wedge.

Fig. 1


3 “S” on the top – check. The red and the blue lines are standard Tenkan-sen and Kijun-sen (Ichimoku indicator).  Of course they lag a bit, but you can get the sense of the volatility just by looking at it. The wedge is over when the Kijun-sen gets volated. You should go short just below it.

Fig. 2


3 “S” check. Break below Kijun-sen, perfect.

Fig. 3


This one shows why you should not try to guess the top, and only go short upon the blue line’s break. (After the 3rd “S” of course).



Lessons in S-A Sequence

Do you have any N-Sync jokes? I can’t think of a single one.


After the second S, the break of the swing low starts the selling with A-Sync appearing as a lower high.


After the second S, the break of the swing high starts the buying with A-Sync appearing as a higher low.



Not every fractal matters.


Not every A-sync Fractal would resist, only those following a Sync fractal closely.


Async points should not be violated. The weight of the 30 minute candle is where you may be able to get in.


A tie off is when price leaving a swing point in a hurry (1H+ drop/rise).


To maximize your potential, you must travel. These indications help you finding the turns.

The priority flips when they manage to push the price over the Green River by more than 40 pips (Guard Rail – Chartreuse thick) with opposing configuration (Mr. Maroon on the other side.)


I am a traveller.
My harbors are the Sync-Async sequences not exceeding 10(U)/12(D) candles in between them.

I am looking for
– a short entry above the E16
– a long entry below the E16