Peaky Fifo

[cozybooks]

Peaky Fifo is looking good. 5 Minute Sell account demo got all messed up (nothing to do with the EA). 5 Minute Buy account if performing beautifully. I was waiting for one of the pairs to hit three trades to constitute a basket. This finally happened with GBPCHF (NanningBob's favorite pair, as I recall).

Trading .01 lots and moving up .01 with each trade. I purposely set the basket TP to $5,000 so that it would just run without closing. When I saw trades of .01, .02 and .03 that were all open, with the basket running almost $16.00 in profit, I changed the basket TP to 10.00, and whoosh, the three trades closed.

As an aside, the 5 minute demo is up 345 pips ($50.00 because of the bigger lot sizes on the second and third trades), with three single trades open at the moment.

Steve and Gary, what a great job!

It would be interesting to watch three demos run side-by-side for a while, one set to increment each trade, one set to start at the highest lot size and move down, and one set to start in the middle.

I've set up Peaky Fifo to run live in a 15 minute TF just on AUDUSD so far. It's been stuck all year between .73 and .7750, so if it stays there it will be a nice opportunity to make some pips. In case it doesn't want to behave itself, I'll put a buystop grid starting at .7830 and a sell stop grid starting at .7270 just in case in breaks out of its range, while still letting Peaky Fifo do her thing for the long term.

Side note to U.S. traders: If you don't (or forgot to) use the varied lot sizes or are using one of Steve's buy or sell stop scripts and you are using Oanda, you can still use the Oanda Desktop trade manager to change the lot sizes of your trades as long as they are open orders and haven't filled yet. So if have .01 lots, they will show up as 1,000 each in the Oanda trade manager. You can change them to 1001, 1002, 1003, etc. just in case you want to close them out of order at some point later on.

Happy trading!

[cozybooks]

Demo update - I've been running Peaky Fifo on a 5 Minute demo on 17 pairs - buy trades only (I have a sell trade only one that got all messed up and needs to start over). It was up early, then in a fair sized hole.

I have varied lots=true, starting at .01 and bumping up .01 with each order.

Very interesting to watch it pop along. Currently at +90.14 in closed trades on 545 pips, with -96.08 in the hole on open trades, running -859 pips. Number of pips is inconsequential with the different lot sizes. I'll let it run through July and August and see how it fairs.

Have fun!

[cozybooks]

Peaky Fifo 5 Minute Demo - Buy Only.

I'm letting this plug along to see if it can find it's way to profit. One of my questions has been whether Peaky (regular or Fifo version) could handle the lower timeframes.

Starts with .01 lots, with .01 increment per trade. Right now i am at a five trade max per pair. $ 718 loss in open trades, $ 269 profit in closed trades. Really, that's not bad at this stage. I will give it some room to see how it does.

Happy trading!

[cozybooks]

So my 5-Minute demo was recovering nicely until Windows 10 decided to mess up what I was going. All for the best, as I really wanted to dedicate the demo CPU time to the following:

I've started a new long term demo for Peaky Fifo - 1 Hour TF, 20 pairs, starting at .01 lot per $4,000, and incrementing .01 with each trade, nice and conservative. 1,440 bars for Sixths. TP of 105 pips, no SL of course. I'm running this live with the exact same settings on a just a few pairs, but I want to see how it will perform on twenty pairs over 6-12 months time. I'll pop back in from time to time as long as the thread stays alive to let you know how it is going. I think this should do very, very well. If all goes well on the live pairs and they churn out profits, I will slowly increase the number of live pairs going as well.

One last setting: FollowAdverseMarketwithStopOrders=true, but I am setting the MarketDistancePips to 200, which is a fairly high number.

Happy trading!

[spgandau]

I have incorporated this part of Peaky FIFO into my own code. It is a nice feature! As I was looking at it, it seems that there is a possible logical "bypass" that might surprise a user....Example: if the user sets a number value for SundayStartHour and sets the 'TradeSundayCandle=false', the EA will continue to execute... because 'if (h < SundayStartHour)' is checked before 'if (!TradeSundayCandle)' and may return 'true', thus bypassing the if(!TradeSundayCandle) altogether

This is true for Sunday, Monday, Friday and Saturday.

ORIGINAL CODE SNIPPET (WITH PARTS REM'D)

Code: Select all

bool SundayMondayFridayStuff()
{
   //Friday/Saturday stop trading hour
   int d = TimeDayOfWeek(TimeLocal());
   int h = TimeHour(TimeLocal());

   if (d == 5)
      if (h >= FridayStopTradingHour)
         return(false);

   if (d == 6)
      if (h >= SaturdayStopTradingHour)
         return(false);

   //Sunday start hour      if (!TradeSundayCandle)
   if (d == 0)
      if (h < SundayStartHour)
         return(false);

   //Sunday candle
   if (d == 0)
      if (!TradeSundayCandle)
         return(false);

   //Monday start hour
   if (d == 1)
      if (h < MondayStartHour)
         return(false);

   //Monday candle
   if (d == 1)
      if (!TradeMondayCandle)
         return(false);
//............
//............
   //Friday candle
   if (d == 5)
      if (!TradeFridayCandle)
         return(false);

   //Saturday candle
   if (d == 6)
      if (!TradeSaturdayCandle)
         return(false);

   //Got this far, so we are in a trading period
   return(true);

}//End bool  SundayMondayFridayStuff()

Proposed MODIFIED CODE CODE SNIPPET (WITH PARTS REM'D)

Code: Select all

bool SundayMondayFridayStuff()
{
   //Friday/Saturday stop trading hour
   int d = TimeDayOfWeek(TimeLocal());
   int h = TimeHour(TimeLocal());
//................
//................
   //Sunday candle
   if (d == 0)
   {
      if (!TradeSundayCandle) return(false);
      //Sunday start hour
      if (h < SundayStartHour) return(false);
   }
   //Monday candle
   if (d == 1)
   {
      if (!TradeMondayCandle) return(false);
      //Monday start hour
      if (h < MondayStartHour) return(false);
   }
//............
//............
   //Friday candle
   if (d == 5)
   {
      if (!TradeFridayCandle) return(false);
      // Friday stop hour
      if (h >= FridayStopTradingHour) return(false);
   }

   //Saturday candle
   if (d == 6)
   {
      if (!TradeSaturdayCandle) return(false);
      // Saturdy stop hour
      if (h >= SaturdayStopTradingHour) return(false);
   }
   //Got this far, so we are in a trading period
   return(true);

}//End bool  SundayMondayFridayStuff()

I don't know who wrote the original code, so if I have missed something obvious please let me know...and delete this.

Cheers!
spgandau

[SteveHopwood]

I have incorporated this part of Peaky FIFO into my own code. It is a nice feature! As I was looking at it, it seems that there is a possible logical "bypass" that might surprise a user....Example: if the user sets a number value for SundayStartHour and sets the 'TradeSundayCandle=false', the EA will continue to execute... because 'if (h < SundayStartHour)' is checked before 'if (!TradeSundayCandle)' and may return 'true', thus bypassing the if(!TradeSundayCandle) altogether

This is true for Sunday, Monday, Friday and Saturday.

ORIGINAL CODE SNIPPET (WITH PARTS REM'D)

Code: Select all

bool SundayMondayFridayStuff()
{
   //Friday/Saturday stop trading hour
   int d = TimeDayOfWeek(TimeLocal());
   int h = TimeHour(TimeLocal());

   if (d == 5)
      if (h >= FridayStopTradingHour)
         return(false);

   if (d == 6)
      if (h >= SaturdayStopTradingHour)
         return(false);

   //Sunday start hour      if (!TradeSundayCandle)
   if (d == 0)
      if (h < SundayStartHour)
         return(false);

   //Sunday candle
   if (d == 0)
      if (!TradeSundayCandle)
         return(false);

   //Monday start hour
   if (d == 1)
      if (h < MondayStartHour)
         return(false);

   //Monday candle
   if (d == 1)
      if (!TradeMondayCandle)
         return(false);
//............
//............
   //Friday candle
   if (d == 5)
      if (!TradeFridayCandle)
         return(false);

   //Saturday candle
   if (d == 6)
      if (!TradeSaturdayCandle)
         return(false);

   //Got this far, so we are in a trading period
   return(true);

}//End bool  SundayMondayFridayStuff()

Proposed MODIFIED CODE CODE SNIPPET (WITH PARTS REM'D)

Code: Select all

bool SundayMondayFridayStuff()
{
   //Friday/Saturday stop trading hour
   int d = TimeDayOfWeek(TimeLocal());
   int h = TimeHour(TimeLocal());
//................
//................
   //Sunday candle
   if (d == 0)
   {
      if (!TradeSundayCandle) return(false);
      //Sunday start hour
      if (h < SundayStartHour) return(false);
   }
   //Monday candle
   if (d == 1)
   {
      if (!TradeMondayCandle) return(false);
      //Monday start hour
      if (h < MondayStartHour) return(false);
   }
//............
//............
   //Friday candle
   if (d == 5)
   {
      if (!TradeFridayCandle) return(false);
      // Friday stop hour
      if (h >= FridayStopTradingHour) return(false);
   }

   //Saturday candle
   if (d == 6)
   {
      if (!TradeSaturdayCandle) return(false);
      // Saturdy stop hour
      if (h >= SaturdayStopTradingHour) return(false);
   }
   //Got this far, so we are in a trading period
   return(true);

}//End bool  SundayMondayFridayStuff()

I don't know who wrote the original code, so if I have missed something obvious please let me know...and delete this.

Cheers!
spgandau

This is my code. Nice spot.

[cozybooks]

10 Days into August and my Peaky Fifo test on the 1-hour timeframe is going nicely. TP is 105 pips on all charts. Running on 20 pairs. Four trades closed for a 420 pip ($ 40.47) profit.

I have six open trades currently running at dead even (well, 35 cents total profit at the moment )

How can you not like Peaky in all of its forms. As I clear old trades off my live accounts, I will slowly replace them empty spots with Peaky Fifo. Once I get far enough to have live results, I'll post them as well.

Steve, you area double Superstar!

[cozybooks]

Gary, Peaky Fifo is running flawlessly so far. Thanks for all of the hard work on this! On a few pairs, I'm setting TP to 1000 pips and letting them run, looking to wait to close until price goes all the way to the opposite sixths line and generates an opposite trade signal.

As a manual addition to this, I have the HGI indi running with just the Trend Big arrow showing on the chart. Once Peaky has made its first Sell trade, I'll add another sell trade manually any time I see the Big Red sell arrow, and vice versa obviously for long trades. So far this looks pretty good. I'll try to keep track of it as it goes along to see if it proves as worthwhile as it seems.

Thanks again, to both you and Steve!!

[rosst]

10 Days into August and my Peaky Fifo test on the 1-hour timeframe is going nicely. TP is 105 pips on all charts. Running on 20 pairs. Four trades closed for a 420 pip ($ 40.47) profit.

I have six open trades currently running at dead even (well, 35 cents total profit at the moment )

How can you not like Peaky in all of its forms. As I clear old trades off my live accounts, I will slowly replace them empty spots with Peaky Fifo. Once I get far enough to have live results, I'll post them as well.

Steve, you area double Superstar!

Would you mind posting your set - I must have made a mistake in my settings somehow as am having losses mostly.
Thanks!

[cozybooks]

HI Rosst,

My set file is attached. I am trading the one hour tf, I am using varied lots that start at .01 and increase by .01 with each new trade.

It might depend on when the demos started, but you can compare this against what you are doing and see if there are any other differences. My demos just started in August, so I don't have any multiple trades yet.

Good luck!