Shelley's multi-pair basket trader

[Kenboi]

I have just copied this post from Bob on FF...very interesting comments he is making about the slope indicator!

http://www.forexfactory.com/showthread. ... ost5722211

The slope will change in its strength but not in its angle. In other words if it is above 0 it is above 0. As it continues to move up the bars represent the angle upward so they increase as the angle increases. This is what lets us know how strong the move is. The key though is the whether the angle is moving up or down. The reason I like the indicator so much is it catches the subtle change in direction that you dont see by eyeballing the chart. For example we think price is ranging but it will show a slight angle down which gives us the hint that it is going to go that way. We are prepared to trade down when others are sitting and wondering which direction to go next. That along with it being below the weekly open and then the crossing of the pivot is what sets this system apart from the rest. We are in when others are wondering, is this it?? That is the advantage the slope gives you. That slight angle down we see now, that later will show bigger bars in a back test as the angel increases. We catch the new direction quickly. If we are wrong and it switches on us, well we see price go above the pivot telling us something maybe wrong. It is this quickness in seeing these subtleties that give this system its uniqueness and ability.

and this post:

http://www.forexfactory.com/showthread. ... ost5722213

For example we now have new monthly pivot lines. If the euro decides to go up we will be one of the first to know because. Price will go above the weekly, then monthly pivots and the slope will angle up and the bars will be above the 0 line. When others wonder how long or strong the move will be, we will know. It went above 4, we know there is one good trade here. Look it went above 8 this is going to last a while. OR it hit 4 and then went below, temporary, dont do another buy. That is very powerful information. When others are guessing and arguing fundamentals, etc. we will know what to do next. (barring outside interference from banks and big boys which no indicator can tell you)

I think this is exactly what we have been seeing the last couple of weeks!

shelley, good stuff. I begin to suck myself into bob's 10.2 system because in the last 2 weeks of progress, the positive outcomes were so concrete and so predictable that I almost believe it's a hair off from Holy G. Thanks for posting

[phil_trade]

Hi Steve

Coming back from hard work !

For TMAtrueConfirmation, you ask TimeFrame, but don't use it in GetTmaTrue, so just take it from chart !

Code: Select all

void GetTmaTrue(string symbol, int shift)
{
   //This code replaces an iCustom call to Tma True. It is provided by Paul (Basluda).
   //Many thanks Paul. You are a star.
   
   double IB_TMA = iMA( symbol, 0, xi_Period, 0, MODE_LWMA, PRICE_CLOSE, shift );
   double ld_Range = iATR( symbol, 0, xi_ATR.Period, shift + 10 );

   TmaUpper = IB_TMA + (ld_Range * xd_ATR.Mult);
   TmaLower = IB_TMA - (ld_Range * xd_ATR.Mult);
}//void GetTmaTrue(string symbol, int shift)

should be this, no ?

Code: Select all

void GetTmaTrue(string symbol, int shift)
{
   //This code replaces an iCustom call to Tma True. It is provided by Paul (Basluda).
   //Many thanks Paul. You are a star.
   
   double IB_TMA = iMA( symbol, TmaTrueTimeFrame, xi_Period, 0, MODE_LWMA, PRICE_CLOSE, shift );
   double ld_Range = iATR( symbol, TmaTrueTimeFrame, xi_ATR.Period, shift + 10 );

   TmaUpper = IB_TMA + (ld_Range * xd_ATR.Mult);
   TmaLower = IB_TMA - (ld_Range * xd_ATR.Mult);
}//void GetTmaTrue(string symbol, int shift)

[quanttraderjoe]

Unfortunately forward testing takes time & most often one month is nowhere near enough time to decide how stable a strategy is.

Very true, Shelley. I wish I could test it in Tradestation, but unfortunately TS can't test multi-symbol strategies like this.

I'm finishing to translate in Delphi for ForexTester to backest multi-pair with good historical data. Will post infos asap.

Philippe

hello Philippe,

really appreciate your hard work and effort in backtesting this multi-pair ea!
hope we can get good results using the tma m30 slope filter to reduce the odds of getting caught out by the retracements. but the issue of missing multiple trade opportunities remain when we use the filter.

would it be advisable for the ea to check every 5mins or so for favorable conditions (m30 tma d1 tma slope at 0.8) and subsequently trigger the basket assuming there r no open baskets? that way we would not miss any high momentum baskets. would this be the best of all worlds?

any thoughts guys?

[phil_trade]

hello Philippe,

really appreciate your hard work and effort in backtesting this multi-pair ea!
hope we can get good results using the tma m30 slope filter to reduce the odds of getting caught out by the retracements. but the issue of missing multiple trade opportunities remain when we use the filter.

would it be advisable for the ea to check every 5mins or so for favorable conditions (m30 tma d1 tma slope at 0.8) and subsequently trigger the basket assuming there r no open baskets? that way we would not miss any high momentum baskets. would this be the best of all worlds?

any thoughts guys?[/quote]

I'm testing first, the Shelley original version for determining best Open Session Hour and best statistical TP upon 3 years 2009-2011 and then forwarding test over 2012 to confirm

I shall add other filters after, because it's hard work to translate MQL to Delphi.
So waiting for idea guys...but not to much please

[quanttraderjoe]

"I'm testing first, the Shelley original version for determining best Open Session Hour and best statistical TP upon 3 years 2009-2011 and then forwarding test over 2012 to confirm

I shall add other filters after, because it's hard work to translate MQL to Delphi.
So waiting for idea guys...but not to much please "

wow...this is GOLD Philippe!

thanks again! great work!

[spotdespot]

hello Philippe,

really appreciate your hard work and effort in backtesting this multi-pair ea!
hope we can get good results using the tma m30 slope filter to reduce the odds of getting caught out by the retracements. but the issue of missing multiple trade opportunities remain when we use the filter.

would it be advisable for the ea to check every 5mins or so for favorable conditions (m30 tma d1 tma slope at 0.8) and subsequently trigger the basket assuming there r no open baskets? that way we would not miss any high momentum baskets. would this be the best of all worlds?

any thoughts guys?

This is essentially what I suggested yesterday for an adaptation of Shelley plus Gary but I think it could just open trades as the lower timeframe condition is met (all original Shelley conditions already met). Any basket treatment of the trades are best handled by MPTM as the trades would often be triggered individually. I think the concept has a lot of merit but no-one replied so I didn't follow up.

[SteveHopwood]

and add prefix/suffix eg.:

Code: Select all

extern string	PairsToTrade = "AUDCADm,AUDCHFm,AUDNZDm,AUDJPYm,AUDUSDm,CADCHFm,CADJPYm,CHFJPYm,EURAUDm,EURCADm,EURCHFm,EURGBPm,EURNZDm,EURJPYm,EURUSDm,GBPCHFm,GBPJPYm,GBPUSDm,NZDUSDm,NZDJPYm,USDCADm,USDCHFm,USDJPYm";

Just so everyone knows, all my multi-pair bots add pair suffixes automatically. There is no need to add them manually.

hi Steve, unfortunately i didnt work for me until i changed pairs manually.
oddly enough, EA was giving me "ordersend error 131". And no, my trade lot size was not wrong. My crim allows 0.01, and it was set to 0.01. Even if i've changed it to 0.1 or 1.0 it would not help me and would flood me with alerts until i added suffix to the pairs which solved the problem.

just my story, may help others too.

Curious. The code to do so is there and works for me. As you can see from the lack of comment about this, it works for everyone else. Strange.

[SteveHopwood]

Hi Steve

Coming back from hard work !

For TMAtrueConfirmation, you ask TimeFrame, but don't use it in GetTmaTrue, so just take it from chart !

Code: Select all

void GetTmaTrue(string symbol, int shift)
{
   //This code replaces an iCustom call to Tma True. It is provided by Paul (Basluda).
   //Many thanks Paul. You are a star.
   
   double IB_TMA = iMA( symbol, 0, xi_Period, 0, MODE_LWMA, PRICE_CLOSE, shift );
   double ld_Range = iATR( symbol, 0, xi_ATR.Period, shift + 10 );

   TmaUpper = IB_TMA + (ld_Range * xd_ATR.Mult);
   TmaLower = IB_TMA - (ld_Range * xd_ATR.Mult);
}//void GetTmaTrue(string symbol, int shift)

should be this, no ?

Code: Select all

void GetTmaTrue(string symbol, int shift)
{
   //This code replaces an iCustom call to Tma True. It is provided by Paul (Basluda).
   //Many thanks Paul. You are a star.
   
   double IB_TMA = iMA( symbol, TmaTrueTimeFrame, xi_Period, 0, MODE_LWMA, PRICE_CLOSE, shift );
   double ld_Range = iATR( symbol, TmaTrueTimeFrame, xi_ATR.Period, shift + 10 );

   TmaUpper = IB_TMA + (ld_Range * xd_ATR.Mult);
   TmaLower = IB_TMA - (ld_Range * xd_ATR.Mult);
}//void GetTmaTrue(string symbol, int shift)

Cheers Phil. It should indeed.

[Zypip]

I have tested Shelley today again with the following setup and results:

- Original Shelly : +500 pips ish
- Shelley with TMATRUETF 60M : +250pips ish
- Shelley with slope conf TF 60M : +100pips ish

If I could be allowed one wish it would be to have the slopeconf TF trigger at 0.4/-0.4 instead of the current 0.8/-0.8 (actually user definable would be fantastic for fine tuning). From my preliminary analysis it would increase the number of trades without too much impairing on the retracement filtering.
Still using v1r with Gary's setfile with jsl disabled.
Will keep my comparison settings running to see which one will fail first and if one can survive future significant retracements. I may add another Empty4 session using SCTF and TMATRUETF set at 30M.

-

I have added this and V 1u is attached. The inputs are SlopeConfirmationBuy/Sell.

This also includes Gary's JSL fix. I attach it here for the coders and for anyone who has found a way to play with this in weekend backtesting. I shall upload it (with anything else that comes up) to post 1 in time for Sunday/Monday opening.

Thanks Steve! I will demo and report asap. Cheers

[Zypip]

I have just copied this post from Bob on FF...very interesting comments he is making about the slope indicator!

http://www.forexfactory.com/showthread. ... ost5722211

.... The key though is the whether the angle is moving up or down. ...

Very interesting Shelley, thank you for pointing this out.