One for all. General exeprt. - page 5

New comment
 
Williams says at least 10% per month, which is not bad in principle =)

2 solandr:
Maybe they just hide the fact that they are billionaires ;))
I think they already have such an EA written somewhere, I am too lazy to write it myself.

2 ExpertTrader:
It seems to me difficult to write an Expert Advisor based only on Alligator, because the computer still needs to explain how to clearly identify the trend (yes, there is Gator, but there you can only talk about a corridor compared to trends, which means you have to analyze the history, etc.).
If we do it the way Billy describes, then it is the same whether the Alligator's jaw is wide open or not, and the signal for entering is fractal penetration, and then apply other measurements.


In general, it is difficult not to agree with Bill's arguments (psychologist, after all =)))
 
I can't stand it. I will enter the conversation at this stage. 2 rebus regarding "In fact, the bar contains the most important thing - the vector of market changes." Maybe it's more correct to say that the bar contains the most important thing... On the other hand, a bar in general is an imperfect instrument, given some fractality of the market. And on this basis, we can talk about the imperfection of most tools like AO and the like. Such is the point of view.
 
2 njel:
How do you see a bar as a tool? Like "Close price is less than open price"? Is it possible to get something definite out of a bar? On the other hand, a sequence of bars is more than one bar. And "considering some fractality of the market" (I would say non-linearity of the market) Williams just developed Fractals.

I think that if you strictly follow Bill's proposed strategy, the profit should be there. (If you look at the quotes history, it seems to really work). Unfortunately I tried it little on demo, but maybe someone will confirm, that it's true (or refute it ;( ).
 
2 m1rr0r:

well, what's a bar not a tool? Closing less than opening, shows highs and lows for the period... for example :) . But that's not the point at all. When you develop a strategy, it's not just the result of running it that counts. The lion's share of all working systems is represented by the ones that are out of date. Tell me how many strategies you need to generate, so that on history, they would give you profit, and I will generate them for you. 5 quid per strategy. if more than 100, then + 5 strategies for free.... If you have more than 100, you get 5 strategies for free. There is nothing to complain about, because everything is logical. What is the basis of Williams' theory? Well, fractals, so what?
 

Here's what I've come to in the meantime.
I suggest to use MACD.
Signals:
1. Buy: When the indicator crosses the signal line from bottom to top, the trend, defined by the same indicator, should be directed upwards. The trend is plotted based on two points: from the previous signal to the current one (shown in the picture with a black line).
2. Close a long position: the indicator crosses the signal line from top to bottom (the signal line becomes higher).

Opening and closing of short positions is mirrored.

 
>I suggest using the MACD.

Simply genius! :o)))
 
ExpertTrader писал (а):

This is what I have come to in the meantime.
I propose to use MACD.
Signals:
1. Buy: When the indicator crosses the signal line from bottom to top, the trend, defined by the same indicator, should be directed upwards. The trend is plotted based on two points: from the previous signal to the current one (shown in the picture with a black line).
2. Close a long position: the indicator crosses the signal line downwards (the signal line becomes higher).

Opening and closing of short positions is mirrored.


//+------------------------------------------------------------------+
//|                                     MACD Double Signal Trend.mq4 |
//|                      Copyright © 2006, MetaQuotes Software Corp. |
//|                                        https://www.metaquotes.net/ |
//+------------------------------------------------------------------+
#property copyright "Copyright © 2006, MetaQuotes Software Corp."
#property link      "https://www.metaquotes.net/"
 
#property indicator_chart_window
#property indicator_buffers 5
#property indicator_color1 Blue
#property indicator_color2 Yellow
#property indicator_color3 Blue
#property indicator_color4 Yellow
//---- buffers
double buf0[];
double buf1[];
double buf2[];
double buf3[];
double tempbuf[];
 
//+------------------------------------------------------------------+
//| Custom indicator initialization function                         |
//+------------------------------------------------------------------+
int init()
  {
//---- indicators
   SetIndexStyle(0,DRAW_ARROW,EMPTY, 2, Blue);
   SetIndexBuffer(0,buf0);
   SetIndexArrow(0,241);
 
   SetIndexStyle(1,DRAW_ARROW,EMPTY, 2, Yellow);
   SetIndexBuffer(1,buf1);
   SetIndexArrow(1,242);
 
//   SetIndexStyle(2,DRAW_ARROW,EMPTY, 1, Blue);
   SetIndexBuffer(2,buf2);
//   SetIndexArrow(2,251);
 
//   SetIndexStyle(3,DRAW_ARROW,EMPTY, 1, Yellow);
   SetIndexBuffer(3,buf3);
//   SetIndexArrow(3,251);
 
   SetIndexBuffer(4,tempbuf);
//----
   return(0);
  }
//+------------------------------------------------------------------+
//| Custom indicator deinitialization function                       |
//+------------------------------------------------------------------+
int deinit()
  {
//----
   
//----
   return(0);
  }
//+------------------------------------------------------------------+
//| Custom indicator iteration function                              |
//+------------------------------------------------------------------+
int start()
  {
   int   counted_bars=IndicatorCounted();
   int   limit=Bars-counted_bars;
   if( counted_bars>0) limit++;
   for(int i=0; i<limit; i++)
      {
      double MACD1=iMACD(NULL,0,12,26,9,PRICE_CLOSE,MODE_MAIN,i+1);
      double MACD2=iMACD(NULL,0,12,26,9,PRICE_CLOSE,MODE_MAIN,i+2);
      double signal1=iMACD(NULL,0,12,26,9,PRICE_CLOSE,MODE_SIGNAL,i+1);
      double signal2=iMACD(NULL,0,12,26,9,PRICE_CLOSE,MODE_SIGNAL,i+2);
      
      if (MACD2<signal2 && MACD1>signal1) 
         {
         buf2[i]=Open[i];
         tempbuf[i]=MACD1;
         }
      if (MACD2>signal2 && MACD1<signal1) 
         {
         buf3[i]=Open[i];
         tempbuf[i]=MACD1;
         }
      }
   i=limit;
   int count=0;
   while (i<Bars && count<2) 
      {
      if (tempbuf[i]!=0)count++;
      i++;
      }
   int iB=0;
   int iS=0;
   while (i>0)
      {
      if (buf2[i]!=EMPTY_VALUE)
         {
         if (iB!=0 && tempbuf[i]>tempbuf[iB]) buf0[i]=Open[i];
         iB=i;
         }
      if (buf3[i]!=EMPTY_VALUE)
         {
         if (iS!=0 && tempbuf[i]<tempbuf[iS]) buf1[i]=Open[i];
         iS=i;
         }
      i--;
      }
//----
   
//----
   return(0);
  }
//+------------------------------------------------------------------+

Right?
 
rebus:
I came up with the idea when the test visualisation came out. I paid attention to the way price behaves within a single bar. Basically, you don't need to open a lower TF to see what is happening. Not always, of course, but very often. That's why I thought: what if I calculate the average price of the same ticks within one bar (any TF) and compare it with the bar reference centre (High+Low+Close)/3? Or better yet, compare it with the OC of the previous bar. The result should be an offset or a vector, where the price goes. The most interesting thing is that this vector can already be used before the current bar closes. Approximately, this is the idea. I have not had a chance to try it yet. I forgot all my business with the Championship, and now I have to catch up with it.
Give me your criticism. I'll be sure to experiment myself, as soon as I've got some rest. At the moment I'm using the daytime OC and its levels. Our ancestors were clever, I tell you! :) Therefore I will dig in this direction anyway.

The essence of this idea is supposed to be implemented by the ordinary moving average. With a period of 1. The MA price calculation methods are quite diverse. I think the MA method can be selected according to the above criteria. By calculating the MA on the previous bar, we obtain the MA line which will be a vector of price changes. Did I get it right ? The Expert Advisor by difference in values on candlesticks is simple.
Indeed, only experimentation can show the viability of the theory.
 
>See, for example, how Rosh approaches price research. There's nothing to complain about because everything makes sense.

Send me a link to where to look.
 
I've cooked up a "turkey" here, so check it out. Perhaps on the basis of this idea you can make a profitable Expert Advisor...

The indicator is built using three moving averages. If you plot those MAs on the chart you will get something similar to the Alligator.

1. If MA Close is higher than MA (H+L+O+C)/4, while MA (H+L+O+C)/4 is higher than MA Open - BUY.
2. If MA Close is lower than MA (H+L+O+C)/4 and MA (H+L+O+C)/4 is lower than MA Open - SELL.
If MA Close is higher than MA (H+L+O+C)/4 - CLOSE.
4. If MA Close is lower than MA (H+L+O+C)/4 - CLOSE LONG POSITION.

The Indicator shows these signals using two lines:
1. The green line - shows buy/sell signals. If zero line is crossed from below to above - BUY, from above to below - SELL.
2. The red line indicates signals for closing a position. At zero line crossing from below upwards - CLOSE SHORT POSITION, from above downwards - CLOSE LONG POSITION.



Indicator parameters:
1. period - period of averaging for moving average calculation.
2. ma_method - averaging method. Can be any of values of Moving Average methods.
Files:
mamy.mq4  2 kb
123456789101112
New comment
Reason: