Is The Bottom In?

Watch my hand, I’ll do it in slow motion.

On the bottom is the CI chronometer / metronome. The question it is asking repetitively is about the bottom when the radioactivity is red. Now, why was that fractal landing on the zero line not a Head?

Because the price arrived from inside a previous range.

Why was this a dynamite head?

The dynamite had a corresponding target – a “160 degree” full swing on the Range Pendulum.

You knew this? Very well, come sit in the front!

The First Blood print was where they dumped a larger short position into the German PMI today. They were not careful enough and they moved the market tipping their hand. I actually could spot this in the price movement.

     ///first cut up

   if (NBEU[i]!=EMPTY_VALUE && NBED[i]!=EMPTY_VALUE)
   if (iLow(symbol,0,iLowest(symbol,0,MODE_LOW,100,i))<NBED[iLowest(symbol,0,MODE_LOW,100,i)] && High[i]>iLow(symbol,0,iLowest(symbol,0,MODE_LOW,100,i))+480*Point && Low[i]<iLow(symbol,0,iLowest(symbol,0,MODE_LOW,100,i))+480*Point
      && (High[i]<taxi[i]-70*Point || (High[i]>iMA(symbol,0,32,0,MODE_EMA,PRICE_OPEN,i)+50*Point && Low[i]<taxi[i])) && High[i]>taxi[i]-500*Point
   ){
////zero dn   
      if (iFractals(symbol,0,MODE_LOWER,i) && Low[i-1]>Low[i] && Low[i]>NBED[i]-330*Point && 
        (
         (High[i]>NBED[i] && Low[i]<NBED[i])  || (High[i+1]>NBED[i+1] && Low[i+1]<NBED[i+1]) || (High[i+2]>NBED[i+2] && Low[i+2]<NBED[i+2])
         || (High[i+3]>NBED[i+3] && Low[i+3]<NBED[i+3])  || (High[i+4]>NBED[i+4] && Low[i+4]<NBED[i+4]) 
         ) 
         &&
         (
          (Low[i+22]<NBED[i+22]-330*Point && NBED[i+22]==NBED[i+25]) || (Low[i+23]<NBED[i+23]-330*Point && NBED[i+23]==NBED[i+25])
         )
         && NBED[i+23]!=EMPTY_VALUE
      ){
        ObjectCreate("Obellix"+IntegerToString(i), OBJ_TEXT, 0, Time[i], Low[i]+80*Point); 
     ObjectSetText("Obellix"+IntegerToString(i), CharToStr(77), 38, "Wingdings", clrDarkGreen);
     if (NBED[i+23]==EMPTY_VALUE)  ObjectSetText("Obellix"+IntegerToString(i), CharToStr(77), 38, "Wingdings", clrYellow);
     
      ObjectCreate("Targeted"+IntegerToString(i), OBJ_TEXT, 0, Time[i],NBEU[i]+((NBEU[i]-NBED[i])/10*4.5)+50*Point );  
                ObjectSetText("Targeted"+IntegerToString(i), "TGT: "+DoubleToStr(NormalizeDouble(NBEU[i]+((NBEU[i]-NBED[i])/10*4.5),4),4), 16, "Arial Black",  clrPurple);
   }

The yellow horizontal line below is the goldilocks level.

As you can see, it splits the gray box in half – halfway between what a typical Wave 1 can manage and what a Wave 3 is likely to achieve.

This is a qualified Wave 1. You can see the hook back up on the stochastic losing the embedding. If this does not get to 76 within 46 hours, you know what you must do. The undercut could be anything up to about 45 pips, so scale in. 10, 20, 30 etc.

The optimal 18-sample stochastic reading for a successful turn is between 9.5 and 4.4. The previous day went too oversold, the number was plotted in white.

Why doesn’t this one say CAP for capsize attempt? I just told you, the previous day went deeply oversold. Today is a low, but not THE Low.

Too much acceleration. Have to re-balance with a divergence.

 if (st18[i]<9.5) { ObjectSetText("Bingo"+IntegerToString(i), "CAP", 26, "Impact", DarkGreen); momentum[i]=1;
            cap[i]=1;
            if (st18[ArrayMinimum(st18,10,i+1)]<4.4) {ObjectSetText("Bingo"+IntegerToString(i), "Cos,+90,-DVG", 26, "Impact", LimeGreen); c77[i]=true;

Here’s an example from earlier. From the CoverOverSold low to the swing high there was 113 pips.

The swing low printed 79 pips lower than the Cos low.

You always seem to get some additional move the next day. This one had 88 pips from the Cos low (this is why I chose 90).

This one had 2 more days up and 157 pips in total.

There would always be a lower low after a Cos +90.

I’ve shown you this before.

The market first fails at the 20SMA and later at the 9EMA before making the final wave down.

So if I had to guess, I would say there’d probably be a rendezvous with the 9 tomorrow or in the coming days. Yes, all new legs down started above the E-44.
Projected distance numbers on the bottom. 4H RSI2 is going to print a continuation divergence at the high.