Market phenomena - page 10

New comment
 
Avals: Alexey, have you checked on synthetic returns based on the volatility of a real instrument?

I.e. you propose to check dependences of values (Hi-Lo) of a real instrument, but not its returns (by the way, my returns are equal to Clo-Open, i.e. not exactly equal to usual)?

If that's the case, it's not difficult.

2 anonymous: And the book is great and very much on topic, thanks a lot!

 
Mathemat:

I.e. are you suggesting to check the value relationships (Hi-Lo) of the real instrument, not its returns (by the way, my returns are equal to Clo-Open, i.e. not quite equal to the normal one)?

If exactly so, it's not difficult.


Let's take the EURO H1. We take one real bar, look at its Volume and generate one synthetic bar on its basis. For example, we randomize 10 - 10 times +1/-1 and get a new bar. And so for all real bars. Here is a script for offline substitution of real instrument history for a random one:

#property show_inputs

extern datetime startdate=0;//дата и время начала 
extern datetime enddate=0;//дата и время конца 
extern bool ToLastData=true;// генерировать до последнего доступного времени 
extern string instrument="EURUSD";//инструменты 
extern string genperiod="60";
extern int koefVola=8;

int start()
  {       if (ToLastData) enddate=iTime(instrument,StrToInteger(genperiod),0);
          if (startdate==0) {
            startdate=enddate-60*60*24*250;
            Print(TimeToStr(startdate)); 
          }  
          
          int ExtHandle=FileOpenHistory(instrument+genperiod+".hst", FILE_BIN|FILE_READ|FILE_WRITE);   
          if (ExtHandle>0) {          
               FileSeek(ExtHandle,148,SEEK_SET);           
               bool sign=true;
               MathSrand(TimeLocal());
               while (sign){

                 datetime t=FileReadInteger(ExtHandle,LONG_VALUE);
                 double o=FileReadDouble(ExtHandle,DOUBLE_VALUE);
                 double l=FileReadDouble(ExtHandle,DOUBLE_VALUE);
                 double h=FileReadDouble(ExtHandle,DOUBLE_VALUE);
                 double c=FileReadDouble(ExtHandle,DOUBLE_VALUE);
                 double v=FileReadDouble(ExtHandle,DOUBLE_VALUE);
                 double lastprice;
                 if ((t>=startdate) && (t<=enddate)) {
                   o=lastprice;
                   h=lastprice;
                   l=lastprice;
                   for (int i=0; i<v*koefVola;i++){
                     int rnd=MathRand();
                     int tick=0;
                     if (rnd<16384) tick=-1; else tick=+1;
                     lastprice=lastprice+tick*Point;
                     if (lastprice>h) h=lastprice;
                     if (lastprice<l) l=lastprice;
                   }//for
                   c=lastprice; 
                   FileSeek(ExtHandle,-44,SEEK_CUR); 
                   FileWriteInteger(ExtHandle, t, LONG_VALUE);
                   FileWriteDouble(ExtHandle, NormalizeDouble(o,Digits), DOUBLE_VALUE);
                   FileWriteDouble(ExtHandle, NormalizeDouble(l,Digits), DOUBLE_VALUE);
                   FileWriteDouble(ExtHandle, NormalizeDouble(h,Digits), DOUBLE_VALUE);
                   FileWriteDouble(ExtHandle, NormalizeDouble(c,Digits), DOUBLE_VALUE);
                   FileWriteDouble(ExtHandle, NormalizeDouble(v,0), DOUBLE_VALUE);        
                   FileFlush(ExtHandle);
                 } else lastprice=c;
                 
                 if (FileIsEnding(ExtHandle) || (t>=enddate)) sign=false;
               }//while
          }  else  return(0);         
   FileClose(ExtHandle);
   return(0);
  }//start

highlighted in the code how a single bar is generated

Just a number of dependencies and "phenomena" on real data that synthetics based on normal distribution do not have and synthetics based on real volatility. I.e. they are based on known properties of volatility.

 
Sweet:
Sorry, where can I read about this?

https://www.mql4.com/ru/search?keyword=%D0%BF%D0%B5%D1%80%D0%B5%D1%82%D0%B5%D0%BA%D0%B0%D1%8E%D1%89%D0%B8%D0%B5+%D0%BF%D0%B0%D1%82%D0%B5%D1%80%D0%BD%D1%8B
 
joo:

It is necessary to find out at what distances between the bars this memory is most clearly observed, and with this information it is possible to effectively build TS based on NS.

Andrey, what makes you think it is the distances that are important? Or maybe the values?

 
Sweet:
Childishly. Based on your research. Elliott's theory, isn't it a myth?


A bit of a "stick-in-the-mud" with your permission. My understanding is that it is a myth (to say the least), yes, quoting is a complex but not random process (this is really encouraging), there are very strong non-linear relationships. But this does not change the main thing - the price trajectory (what the lines, levels, grids, arcs and so on are applied to) actually does not characterize the process itself in any way. In other words - one cannot graphically find these dependences.

Elliott Waves are a dozen lines (constructed according to completely MUTE rules) + intuition + intuition + intuition + intuition (intuition in the period). This is not a business, not to mention the fact that every wave builder builds waves in his own way (and experience has nothing to do with it, not really). You can make money - of course, but once you've earned it, don't stay in the market, run away from it, otherwise it will take it all back :o).

PS: On top of that, long term memory just fucking increases the complexity of the system, and leaves no winners ... (look at the history of any championship, year after year ...)

 
TheXpert:

Andrey, what makes you think that it's the distance that matters? Or maybe values?

Or maybe the meaning, I don't know. I just grabbed hold of this information - 'long-term memory'.

Meanings you say? - I don't know, I don't know. And if the EURUSD price had started 1000 pips lower than when it entered the market? - I don't think anything would change, at the current moment the price would simply be 1000 pips lower.

Or, do you mean something else?

 
Farnsworth:


A bit of a "sticking point" with your permission. My understanding is that it is a myth (to say the least), yes, quoting is a complex but not random process (it really is encouraging), there are very strong non-linear relationships. But this does not change the main thing - the price trajectory (what the lines, levels, grids, arcs, etc. are applied to) actually does not characterize the process itself in any way. In other words - you cannot graphically find these dependencies.

Elliott Waves are a dozen lines (built according to completely MUTE rules) + intuition+intuition+intuition+intuition (intuition in the period). This is not a business, not to mention the fact that every wave builder builds waves in his own way (and experience has nothing to do with it, not really). You can make money - of course, but once you've earned it, don't stay in the market, run away from it, otherwise it will take it all back :o).

PS: On top of that, long term memory just fucking increases the complexity of the system, and leaves no winners ... (look at the history of any championship, year after year ...)

Hi, nice to meet you....)))) You're probably right. It's hard to argue with anything, except how to explain it, it's a fluke, but it repeats in my 6th year....)))

 
ZetM:

Hi! Nice to meet you....))))


And I'm glad to meet you and that you continue your active participation in the forum (I remember you wanted to go on a philosophical holiday) :o)

You're probably right. It's hard to argue with something, except how to explain it, it's random, but it repeats with me for 6 years....)))

Somehow you "agree to disagree", :o) The thing is, with a 50% chance everything repeats itself, any structure in the market, the art of figuring out when. If you have trained your neural network (moSk :o) to identify such plots, then that's just great.

 
Farnsworth:



It's summer, I live by the sea, it's beach season, the "philosophical holiday" continues....)))

I see...)) You're an intelligent person. That's why you can't be backed into a corner and there's no need to try, it's safer for yourself...))

 
joo:

Or, do you mean something else?

Extremes, say.
1...34567891011121314151617...75
New comment
Reason: