Im Back! With a new strategy! <3 kelton

medisoft · March 1, 2013, 12:06pm

Looks well. The change should avoid the multiple exit/entry within one box

andydoc:

Thanks, medisoft, how about this:


//+------------------------------------------------------------------+
//|                                      kelton-renko-2013022801.mq4 |
//|                               Copyright 2013, Dr. Andy Thompson. |
//|                                 mailto://[email protected] |
//+------------------------------------------------------------------+
#property copyright "Copyright 2013, Dr. Andy Thompson"
#property link      "mailto://[email protected]"

//--- input parameters
extern int       SlowEMA=650;
extern int       FastEMA=120;
extern int       SlowRSI=250;
extern int       FastRSI=120;
extern double    Size=1.00;
bool flgReenterLong=FALSE;
bool flgReenterShort=FALSE;


//+------------------------------------------------------------------+
//| expert initialization function                                   |
//+------------------------------------------------------------------+
int init()
  {
//----
   
//----
   return(0);
  }
//+------------------------------------------------------------------+
//| expert deinitialization function                                 |
//+------------------------------------------------------------------+
int deinit()
  {
//----
   
//----
   return(0);
  }
//+------------------------------------------------------------------+
//| expert start function                                            |
//+------------------------------------------------------------------+
int start()
  {
   //loop through current positions and if any in our symbol then manage them (need to amend to magic number later)
   int total =OrdersTotal();
   for(int pos=0;pos<total;pos++)
    {
     if(OrderSelect(pos,SELECT_BY_POS)==false) continue;
     if(OrderSymbol()==Symbol()&&Time[0]>OrderOpenTime())
     {
     ManagePosition(OrderTicket());
     return(0);
     }
    }
   int Trade=0;
   //If Entry conditions satisfied then set direction of trade 
   if(Bid>iMA(NULL,0,SlowEMA,0,1,0,0)&&iRSI(NULL,0,SlowRSI,0,0)>50)Trade=1;
   if(Bid<iMA(NULL,0,SlowEMA,0,1,0,0)&&iRSI(NULL,0,SlowRSI,0,0)<50)Trade=-1;
   //If Reentry conditions satisfied then set direction of trade
   if(flgReenterLong&&iRSI(NULL,0,FastRSI,0,0)>iRSI(NULL,0,SlowRSI,0,0)&&iRSI(NULL,0,FastRSI,0,1)<iRSI(NULL,0,SlowRSI,0,1)&&Bid>iMA(NULL,0,FastEMA,0,1,0,1)) Trade=1;
   if(flgReenterShort&&iRSI(NULL,0,FastRSI,0,0)<iRSI(NULL,0,SlowRSI,0,0)&&iRSI(NULL,0,FastRSI,0,1)>iRSI(NULL,0,SlowRSI,0,1)&&Bid<iMA(NULL,0,FastEMA,0,1,0,1)) Trade=-1;
   //if we have a trade direction then send the order
   double Price=Ask;
   if(Trade==-1) Price=Bid;
   if(Trade!=0) OrderSend(Symbol(),(1-Trade)/2,Size,Price,5,0,0);
   //cancel the reentry flags if reentry conditions no longer exist
   if(flgReenterLong&&Bid<iMA(NULL,0,SlowEMA,0,1,0,0)) flgReenterLong=FALSE;
   if(flgReenterShort&&Bid>iMA(NULL,0,SlowEMA,0,1,0,0)) flgReenterShort=FALSE;
   

//----
   return(0);
  }
//+------------------------------------------------------------------+
//| expert manageposition function                                   |
//+------------------------------------------------------------------+
int ManagePosition(int ticket)
  {
//----
  OrderSelect(ticket,SELECT_BY_TICKET);
  bool RSIClose = FALSE;
  switch(OrderType())
  {
  case 0:
  if(iRSI(NULL,0,FastRSI,0,0)<50) RSIClose=TRUE;
  if((Bid<iMA(NULL,0,FastEMA,0,1,0,0)&&(Open[0]>iMA(NULL,0,FastEMA,0,1,0,0)||Close[1]>iMA(NULL,0,FastEMA,0,1,0,0)))||RSIClose)
  {
  OrderClose(ticket,OrderLots(),Bid,10);
  if(!RSIClose) flgReenterLong = TRUE;
  return(0);
  }
  case 1:
  if(iRSI(NULL,0,FastRSI,0,0)>50)RSIClose=true;
  if((Bid>iMA(NULL,0,FastEMA,0,1,0,0)&&(Open[0]<iMA(NULL,0,FastEMA,0,1,0,0)||Close[1]<iMA(NULL,0,FastEMA,0,1,0,0)))||RSIClose)
  {
  OrderClose(ticket,OrderLots(),Ask,10);
  if(!RSIClose) flgReenterShort = TRUE;
  return(0);
  }
  }
//----
   return(0);
  }
//+------------------------------------------------------------------+

The amendment is in the condition for calling the ManagePosition function.

medisoft · March 1, 2013, 12:20pm

Well, i had various failed trades on gbpnzd last week and this week. But yesterday I entered one trade that seems to be good. It has 82 pips now. (short)

Also I entered a trade today on usdjpy (long) that has 60 pips in profit.

About the trend lines idea, it seems very good idea for me. But it could be simpler than the trend line.

In the chart pipcompounder placed, the trend line was offsetted by about 1 renko box from the trend. That means if we are in a downtrend, when two green renko boxes are closed, we close our sells and buy. If two red boxes are closed, we close our buys and short.

pipcompounder · March 1, 2013, 12:33pm

Do you think the GBP\NZD is a good pair for this idea?

I like the counting the boxes idea, it seems looking at a lot of charts would help with settings.

medisoft · March 1, 2013, 1:26pm

well, with Kelton’s settings, each box is 15 pips, so with high pip average daily range pairs, like GBPNZD, we can get enough boxes to trade.

Small range pairs, like eurchf are useless with 15 pips, because it seldom make a move big enough.

andydoc · March 2, 2013, 1:39am

Counting boxes is brilliant - I should have spotted that it does away with the need to draw an offset trendline.
Going back to an earlier point, regarding calculating virtual renko bars in an EA, this would make back testing easier, and finding optimal settings per pair like boxsize, MA and RSI settings and boxcount for this new reversal idea. That being said, I have found with my provider Alpari at least, that the 1M history available is limited and I am thinking of a new installation with either TDS (which allows back testing on offline charts in strategy tester) or tick data from elsewhere for visual back testing. My Alpari data is so short my RSI250 is not yet drawing!

Kelton · March 2, 2013, 10:03pm

can someone explain the counting boxes thing… I dont understand it.

Also who can make a custom indicator containing 505 ema through 795 ema with thick lines and orange color? It would be an
indicator of almost 300 seperate ema’s right next to each other

Kelton · March 3, 2013, 12:29pm

Also can anyone tell me how to get another copy of the original mt4 indicators… I need another copy of the Relative strength index indicator that will allow me to set the period as well choose if I want to apply it to the close price or open price.

I cant find it anywhere

butchfx · March 4, 2013, 6:23am

I have the same problem as Andydoc above, namely the Renko indicator does not work properly on Alpari. Can someone please point me in the direction of a Metatrader platform that allows sufficient back data to load on a 1M chart?
Thanks.

pipcompounder · March 4, 2013, 8:24am

If a trend starts and goes lets say 12 boxes after your entry, then if we were to use a trendline that is offset just a bit, so that it wouldn’t close the trade on just one box worth of retrace, and use the trendline to close the trade, it is the same as counting the number of boxes the trendline is offset by, because the boxes always go in the same angle. So, if we set it up to only make a decision when there is 3 boxes in a row, (or whatever number we find to work the best) then we may have better trend confirmation, or the best trade exit when in a trade.

For instance, if we hypothetically find that historically when there’s 4 boxes in a row (going the same direction), that it will continue 80% of the time to total more than 8 boxes, then we could make a rule that doesn’t allow trades to be open or closed unless there is 4 boxes in a row. There will then be 20% losses where they hit a 4 box SL anyways, but our winners will cause the general idea to be profitable…make sense?

andydoc · March 4, 2013, 10:46am

Won’t add much to excellent description of box counting except to state the obvious: that it is effectively a stepped trailing SL of x * boxsize, stepped at boxsize, where x = number of boxes!

Custom EMA difficult due to limits on numbers of lines and buffers in a custom indicator in MT4. (Unsure of MT5)

medisoft · March 4, 2013, 2:19pm

Ibfx has enough data for majors and aud and nzd pairs

andydoc · March 4, 2013, 6:32pm

Of course the Renko indicator is only accurate if when the indicator is attached none of the 1m bars in the history are bigger than the box size which during times of high volatility they may be. It falls foul of the absence of intrabar tick data. Without sufficient history it is impossible to tell how severe the effect will be
The use of third party tick data could assist…

mixtrader · March 4, 2013, 11:11pm

I am a Chinese,now living in Chungking.trading forex in china is in an embarrassed position,between legal&illegal,for the government carry out currency control.
Well,thanks for ur strategy.could u please tell more in detail,like shipment space ?wwhatanmanagement

mixtrader · March 4, 2013, 11:20pm

I wanna know about the money management&what percentage do u use on the opening transaction in your strategy.Sometimes I think i am on the right side&open too heavy,but the price goes oppositely.

medisoft · March 5, 2013, 1:05am

I use 1%. Never more than that.

mixtrader · March 5, 2013, 1:28am

very good,a conservative way! Sometimes I even use up to 5%when I think I am right,but pretty dangerous.

medisoft · March 5, 2013, 9:13pm

Well if you read yourself you should note that if you don’t think to be right, then you should not place the trade. Is like thinking you are going to get sick if eat an unknown mushrom but ignore that and eat it anyway. You can get very sick or ever die if the mushrom results to be poisonous.

medisoft · March 8, 2013, 1:34pm

I want to show you how is this technique going.

This is a screenshot of my demo testing account with my own EA, that is almost equal as the one posted here, with only some beauty details.

Look at the “Beneficios” column, lots, and lots of pips there!

killerno · March 8, 2013, 4:43pm

Hi medisoft,

would you share the EA to test or it is not public?
I could not find any previous EA in this topic.
thanks

medisoft · March 8, 2013, 5:32pm

it is public, but it is not ready yet. I need to test a little more to know if it has not bugs.

I need to see it taking profits and use stop loses to know that it is ready.

and yes, there is an EA here in the thread, I think Anydoc made it (forgive me if I’m wrong!)