POOLER/SAN Stunning generosity

[Maurik]

1) trade ranging pairs especially. The larger the volatility of a pair, the larger the grid size we use. So eg trading AUDNZD 10 pip steps before re-entry should be no problem. When trading EURUSD you might want to be a bit more careful and set the grid size a bit higher.

2) Increase the grid size (reentry point) when market starts trending against you.
Base that on
a) absolute pips moved or
b) percentage of margin used or
c) absolute drawdown or
d) percentual drawdown


Cheers, Maurik

Maurik, hi, what do you define as ranging pairs. I am testing a modified grid trader and are using AUDCAD, AUDSGD, are you using other pairs? Also how many would you trade at one time with the 0.01 lot size and a $50K account.

rgds

Peter[/quote]


Hi Peter,

Just to make sure:
Ranging pairs move in a range - trend less. (But you probably got that)
From a practical point of view a ranging pair is a pair that does not move an enormous amount of pips before retracing. Check out AUDNZD. That one is my favorite. AUDCAD is not bad as far as I can tell. People say EURGBP is ranging too, but the range is relatively large, so there you might want to use a somewhat larger grid size.

If I would trade a 10 pip grid, just to make sure, I would only trade 1 pair. If I was going to trade multiple pairs (say four to five) I would increase the grid size depending on the trending behaviour of the pair to 20 to 50, always using the progressive grid size that I mentioned in my earlier post. You will need a larger retracement, but in my experience it will come. Be ready to suffer some drawdowns for days. (DD over 1000 euro is not uncommon if you trade multiple pairs) Also, mind that you will lose some money to swaps.

In case you need me to look at your system or any of your code, I would be happy to.

Good luck,

Maurik

[Pooler]

Thanks everyone so far. Given me much to chew on.

Believe it or not but the idea for this first started when my darling "clueless about trading" teenage daughter said to me one day as I was telling her about Bulls/Bears etc etc "So if you don't have any stop losses you can never lose any money".

After my initial reaction this stuck with me and here we are.

Remember with PSan there are never losses (as incredible as that sounds) because no matter if 10 OR levels are opened if TP is eventually filled then when you add up all the trades you still made your original profit. The problem is the strain on the account while getting there.

I first thought that Bobs recovery method would be the answer because he also doesn't use SL's but its not that simple. Bob's 1.2.6 Recovery kicks in to recover a loss--with PSan there is no loss--just Draw down.

Which is why I suggested some sort of hedge. But if the idea here is capital preservation then maybe the answer lies in a suitable money management strategy without denying yourself the profit potential.

Much to ponder on a Friday afternoon while waiting for NFP Appreciate all the input

[maxpain]

I always look out for ur messages Steve, much appreciate what you do. The knowledge and tools you share are absolutely priceless. And many thanks to Pooler, u rock mate.

Gonna test this and try and see what fits in best. Will report back if i find anything interesting.

[Maurik]

Thanks everyone so far. Given me much to chew on.

Believe it or not but the idea for this first started when my darling "clueless about trading" teenage daughter said to me one day as I was telling her about Bulls/Bears etc etc "So if you don't have any stop losses you can never lose any money".

After my initial reaction this stuck with me and here we are.

Remember with PSan there are never losses (as incredible as that sounds) because no matter if 10 OR levels are opened if TP is eventually filled then when you add up all the trades you still made your original profit. The problem is the strain on the account while getting there.

I first thought that Bobs recovery method would be the answer because he also doesn't use SL's but its not that simple. Bob's 1.2.6 Recovery kicks in to recover a loss--with PSan there is no loss--just Draw down.

Which is why I suggested some sort of hedge. But if the idea here is capital preservation then maybe the answer lies in a suitable money management strategy without denying yourself the profit potential.

Much to ponder on a Friday afternoon while waiting for NFP Appreciate all the input

Hi Pooler,

You are spot on about money management.

Regarding 1.2.6. recovery
Actually, the 1.2.6. recovery works as a hedge as well... all trades are kept open until TP is reached, which is normally somewhere between the second and third trade. The only thing Bob does - as far as I can tell (hopefully I paid enough attention) - is close the first trade after the market has moved x pips aginst the third trade, and open a fourth one. Until now, I haven't been in this position trading the Beast

I trade different systems now. The most important one is a countertrend grid, which hedges losing trades, takes profit relatively quickly and uses an automatically expanding gridsize.

Since I stopped using Stop Losses, I haven't lost a dime. The key to all this is proper Money Management. As long as I trade a 10k account like I only had 1k to lose, and a 50k account, like I only had 5k to lose (or even less), I am doing fine. At least so far . Everyone who is against Martingale is basically right: continuously doubling your bet until you win will kill you at some point. However, some simple math can really limit our risk. As long as we are not too greedy and willing to ride out a period with some drawdown a "Semi-Martingale" approach can work. At least for me.

TTYL.

Maurik

[johnlucas]

Hi everyone,

and many thanks to Pooler for this gift!
The concepts of the Asian session and of the ranging movements during that session have, in my opinion, a solid technical foundation.

Anyway, if I were able to code, I would direct my efforts towards the implementation of a hard stoploss, maybe using the ATR/volatility of each pair.

This is because, as someone may have already experienced, (and I surely, unfortunately, have), it happens that, once in a while, a certain pair starts to travel in one particular direction without looking back and without any meaningful retracement.

In those instances, a "level 10" size or any particular way of "recouping" losses due to an erroneous trade's direction would be of little help.

Anyone trading the markets for a sufficient lenght of time should have already heard of the "Black Swan".
Without a hard SL, I think that one's could deplete the account for much less.

Just my 2 cents,

thank you and have a nice weekend,

John.

[Pooler]

Thanks all,

Here's that last GU trade after NFP. See what I mean re OR levels acting as S/R levels. Which kinda makes sense as they are derived from the Average Daily Range.

Just per info. For three months (April, May June) I demo-traded a $ 1000 account with 25 pairs with the aim of making $15 per day. My goal was 10 pairs to fill each day starting at .01 lots making an av of $ 1.50 per day (per pair) x av. of 20 trading days = $ 300 per month which is 30% growth per month. I ended on $1921 after the 3 months then upped my starting lot size to .02 and promptly *Eish* blew my account.

Therein I suppose is the lesson...what is the correct money management discipline here.

[mozdukemod]

I do not post but must state publicly how impressed I am with the generosity displayed by this board's members. Thank you all for letting me be a part of this community.

[dlewisfl]

Firstly many thanks for the generosity from a fellow South African.
First issue, is this thing is building huge log files?
2012.11.02 06:03:10 FXPooler EA v2.1 GBPUSD,M15: subTotalClosed running
Not sure what that means or why it needs to be printed, but there are thousands of lines saying the same thing when loaded.

Again thanks.

You need to do something about the Print statement or your logs will be huge. Near the bottom of the code, in the "subTotalClosed" function, either delete the Print statement or add a // to the front of it to comment it out. It should look like this:

//Print("subTotalClosed running");

Then re-compile.

[Baluda]

I found this code issue:

Code: Select all

void CalculateADR()
{
   int i=0;

   R1 =  (iHigh(NULL,PERIOD_D1,1)-iLow(NULL,PERIOD_D1,1))/Point;
   for(i=1;i<=5;i++)
      R5    =    R5  +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
   for(i=1;i<=10;i++)
      R10   =    R10 +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
   for(i=1;i<=20;i++)
      R20   =    R20 +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
...

Variables R5, R10 and R20 are NOT initialized. Please change to this:

Code: Select all

void CalculateADR()
{
   int i=0;
   R1 = 0;
   R5 = 0;
   R10 = 0;
   R20 = 0;

   R1 =  (iHigh(NULL,PERIOD_D1,1)-iLow(NULL,PERIOD_D1,1))/Point;
   for(i=1;i<=5;i++)
      R5    =    R5  +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
   for(i=1;i<=10;i++)
      R10   =    R10 +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
   for(i=1;i<=20;i++)
      R20   =    R20 +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
...

and recompile.

Paul

[dlewisfl]

I found this code issue:

Code: Select all

void CalculateADR()
{
   int i=0;

   R1 =  (iHigh(NULL,PERIOD_D1,1)-iLow(NULL,PERIOD_D1,1))/Point;
   for(i=1;i<=5;i++)
      R5    =    R5  +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
   for(i=1;i<=10;i++)
      R10   =    R10 +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
   for(i=1;i<=20;i++)
      R20   =    R20 +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
...

Variables R5, R10 and R20 are NOT initialized. Please change to this:

Code: Select all

void CalculateADR()
{
   int i=0;
   R1 = 0;
   R5 = 0;
   R10 = 0;
   R20 = 0;

   R1 =  (iHigh(NULL,PERIOD_D1,1)-iLow(NULL,PERIOD_D1,1))/Point;
   for(i=1;i<=5;i++)
      R5    =    R5  +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
   for(i=1;i<=10;i++)
      R10   =    R10 +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
   for(i=1;i<=20;i++)
      R20   =    R20 +  (iHigh(NULL,PERIOD_D1,i)-iLow(NULL,PERIOD_D1,i))/Point;
...

and recompile.

Paul

They are initialized, the problem is he defined them as global variables near the top of the code. Since they are only used in this function, they should be defined within the function as local variables. So, you can move their definition from near the top of the code to this function. it should look like this:

void CalculateADR() {
//Calculates the average daily range as Pooler wants to use it.
//Code from Average_Daily_Range.mq4 indi
int i = 0,
R1,
R5 = 0,
R10 = 0,
R20 = 0,
RAvg;
R1 = (iHigh(NULL,PERIOD_D1,1) - iLow(NULL,PERIOD_D1,1)) / Point;
for (i = 1; i <= 5; i++)
R5 += (iHigh(NULL,PERIOD_D1,i) - iLow(NULL,PERIOD_D1,i)) / Point;
for(i = 1; i <= 10; i++)
R10 += (iHigh(NULL,PERIOD_D1,i) - iLow(NULL,PERIOD_D1,i)) / Point;
for(i = 1; i <= 20; i++)
R20 += (iHigh(NULL,PERIOD_D1,i) - iLow(NULL,PERIOD_D1,i)) / Point;
R5 /= 5;
R10 /= 10;
R20 /= 20;

There are other variables defined globally that should be defined locally, as well as other inefficiencies (e.g. not using arrays and copy/pasting code unnecessarily, the number of times he reads through the order pool each tick, using the error message function in the source instead of using an include, needlessly defining variables). The source file for this EA could easily be ~500 lines instead of the 1400+ and be much more efficient.