FollowTheLeader System | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Please suggest a simple & effective Method/System of "Trading System" Forum to newbie 182 replies

The System Behind the System, by Merlin Jeffries 16 replies

200 Ema system - help me build a system ! 19 replies

Knife system, Vegas 4h system related questions. 4 replies

Would you trade this system? New system #2. 10 replies

Attachments: FollowTheLeader System
Exit Attachments

FollowTheLeader System

  • #41
  • Aug 11, 2009 7:58am Aug 11, 2009 7:58am
wow i just want to say i like this so much
i was thinking using previous High Low but using open/close is better
Thank you for sharing this mate.
  • #42
  • Aug 11, 2009 8:04am Aug 11, 2009 8:04am
Definitely welcome. But if you can help by testing this manually or using the EA and see if we have THE SAME results, even for Jan 2 to Jan31, 2009 only.
  • #43
  • Aug 11, 2009 8:08am Aug 11, 2009 8:08am
Quoting 10pipsman
Definitely welcome. But if you can help by testing this manually or using the EA and see if we have THE SAME results, even for Jan 2 to Jan31, 2009 only.
Could you teach me your method of backtesting?
so i can start doing it
i will try use mb trading metatrader 4 its using Newyork time.
  • #44
  • Aug 11, 2009 8:20am Aug 11, 2009 8:20am
I just did it manually, it's Ronald the programmer who's trying to code the EA and tested it via Strategy Tester.

UPDATE:
Further manual testing shows it seems to work best on H4 timeframe than on D1. H1 timeframe is not good, and so is the other lower timeframes.
  • #45
  • Aug 11, 2009 9:48am Aug 11, 2009 9:48am
Thanks for sharing system. I have a question how do we trade Dojis? or we just dont take trades in that case. Thanks

Quoting 10pipsman
I just did it manually, it's Ronald the programmer who's trying to code the EA and tested it via Strategy Tester.

UPDATE:
Further manual testing shows it seems to work best on H4 timeframe than on D1. H1 timeframe is not good, and so is the other lower timeframes.
Yea it's Life//
  • #46
  • Aug 11, 2009 10:22am Aug 11, 2009 10:22am
Quoting biz0101
Looks impressive. The drawdown doesn't seem to be that high, definitely within my allowed drawdown zone. Have you tried to use adaptive lot sizes? ie, when system is performing well, increase position size, when less, decrease position size.
no, i'd like to see the potential from naked money management, that is, 0.1 lot always, if that's promising, then we think about money management.

thanks for your question.
  • #47
  • Aug 11, 2009 10:27am Aug 11, 2009 10:27am
As I said, applied Ronald's set, and also change:

usetakeprofit=true
takeprofit =100
timeframe =1440

then i got my result.

if use Ronald's set, it seems loss on either 1hour or daily trading. however, this is not due to the flaw of the original idea. I guess, no offense, that Ronald's EA doesn't really perform correctly for SLed/Reentry algorithm as 10pipsman suggests.

10PipsMan:

i'm at work, can't do EA running now, but ur idea seems golden to me. please bear with my slow way. i'll work on it after work.

Quoting 10pipsman
Maybe we can let the EA run as how this is designed to be first.

Then make the enhancements as an option? Like take the trending side only?


@charvo, we don't seem to have the same result?

I can email to you the excel sheet, and email me your day to day result maybe, so we can compare notes manually?
  • #48
  • Aug 11, 2009 10:30am Aug 11, 2009 10:30am
Quoting charvo
Ronald's EA doesn't really perform correctly for SLed/Reentry algorithm as 10pipsman suggests.
What part is wrong?
  • #49
  • Aug 11, 2009 11:23am Aug 11, 2009 11:23am
Looking back to 2002 it looks like this is a big loser without any other type of filtering. over 3000 pip loss.
  • #50
  • Aug 11, 2009 11:24am Aug 11, 2009 11:24am
2009.08.0600:001.699081.702851.674981.6806366068UP0.00430.01840.0000
2009.08.0700:001.68061.683251.664981.6685566771DOWN-0.01840.00000.0121
2009.08.0900:001.667281.668941.666451.667892108DOWN-0.01210.0000-.0006
2009.08.1000:001.6681.671671.642941.6474167192UP0.00060.02060.0000
2009.08.1100:001.647391.652111.642981.6494860196DOWN-0.02060.0000-0.0021
Just a sample from my spreradsheet
Totals
Long / Short
-0.0516 -0.3401 -0.3917
  • #51
  • Aug 11, 2009 11:28am Aug 11, 2009 11:28am
Hi, Ronald, Firstly, thanks a lot for your code.

I am not sure if it is wrong or not, really, however, just to clarify:

look:

if yesterday is a "down" day, then we short today. but today's intraday price goes back to yesterday's open, then today's position should be stoploss. then we reentry a long (that's my guess), but the EA seems still reenter short. if so, we shouldn't stoploss in the first place.

if what i described above is not only with my test, then that's why i suggest using yesterday's Hi and Lo instead of open and close. but that'll involve a lot of coding changes.

Just my observations though. Ronald, your programming rocks!


Quoting Ronald Raygun
What part is wrong?
  • #52
  • Aug 11, 2009 11:38am Aug 11, 2009 11:38am
Quoting charvo
I am not sure if it is wrong or not, really, however, just to clarify:

look:

if yesterday is a "down" day, then we short today. but today's intraday price goes back to yesterday's open, then today's position should be stoploss. then we reentry a long (that's my guess), but the EA seems still reenter short. if so, we shouldn't stoploss in the first place.

I coded the EA to follow this particular rule about re-entry.

Quoting 10pipsman


Always trade with the direction of the last candle. So, if the SL of the day was hit, but it came back and went to the direction of the previoius candle, we will open a new position again. But if it is a reversal, we dont open a new position.

Reversals are only possible if the previous candle is the opposite direction of the current trade.
  • #53
  • Aug 11, 2009 5:51pm Aug 11, 2009 5:51pm
To avoid confusion, I am posting this explanation/sample again. I am not good at words, but if a picture can paint a thousand words then, this is it:


Let the SL be hit in 2 ways - either it closes in different kind - bear to bull (Jan 5), or it went past yday's open price (Jan 6)

http://img17.imageshack.us/img17/593/gudaily.jpg
Illustration:

Jan 2 OHLC -
1.4565 1.4576 1.4377 1.4546

Jan 5 OHLC
1.4546 1.4738 1.4433 1.4699

Jan 6 OHLC
1.4698 1.4992 1.4503 1.4916



Take note of Jan 2 candle, it is a BEAR. So for Jan 5 we open SELL trade, SL at Open price of Jan 2 (1.4565). Jan 5 opened at 1.4546, reversed and went up, hit our SL. -19 pips + the 3 pip spread, net loss of 22 pips for the day. We DO NOT open BUY position on this day because previous candle was a bear.

Next day, we BUY (based on Jan 5 bull candle), SL at Open of Jan 5 (1.4546), again it went to a low of 1.4503, thus hitting our SL and a loss of 153. But lo and behold, it went back up again, and went pass our previous day's close. Since we are on a BUY trend, we opened a new position. Thus ending that trade with a nice +218, for a net profit of +59pips for the day.

We only open a SECOND TRADE for the day, if it is on the same trend.

I am sure this "naked" trading style needs a lot of improvement. But right now, lets stick to the basics, and see how it runs in D1 time frame. Like I said, lets test this on D1 first, but further testing shows it seems to be better on H4 timeframe.

For the days of Jan here are the results: Net gain of +1407 pips

2009.01.02
2009.01.05 (19)
2009.01.06 59
2009.01.07 175
2009.01.08 118
2009.01.09 (54)
2009.01.12 244
2009.01.13 319
2009.01.14 (115)
2009.01.15 (92)
2009.01.16 96
2009.01.19 (102)
2009.01.20 493
2009.01.21 (32)
2009.01.22 (32)
2009.01.23 73
2009.01.26 (79)
2009.01.27 134
2009.01.28 114
2009.01.29 (68)
2009.01.30 178


@Ronald, I think instead of looking for good entry/exit point, we might find a good money management instead. Say the lots opening might be based more on the equity available, like using $1 for every $1000?!? But this will only work if indeed this FTL system will hold in the long term. Sounds good?
  • #54
  • Aug 11, 2009 5:53pm Aug 11, 2009 5:53pm
Quoting stevecartwright
Looking back to 2002 it looks like this is a big loser without any other type of filtering. over 3000 pip loss.

Thanks for the feedback, maybe you can run the first month of 2009, so we can see if it is done as how it was supposed to?

It sucks to know it lost so much, but if it does, well ...
  • #55
  • Aug 11, 2009 6:07pm Aug 11, 2009 6:07pm
A while back, I loaded all the daily data for the major pairs I could get and looked at the odds for either trading in the same direction as the last or reversing direction. The odds are pretty much a coin flip as to the direction of the next bar, with a slight edge that the next bar will reverse. The odds for a reversal go up a little more with two consecutive bars in the same direction.

I haven't checked yet, but I guess the more bars in the same direction the higher the odds of a reversal.

Dan
.
  • #56
  • Aug 11, 2009 6:14pm Aug 11, 2009 6:14pm
^nice observation maybe. But please provide us with the numbers.

The idea is to catch the trend by following the previous candle, and be tight and conservative when it reverses on us, by providing a strict SL.
  • #57
  • Aug 11, 2009 6:27pm Aug 11, 2009 6:27pm
I like your approach is very similar to what I've been looking at. Below are the stats I tallied for the odds of a pair's direction on the next day.

Pair Total days Same Reverse
AUD/CAD442 213229
AUD/JPY2972 14661506
AUD/NZD441 228213
AUD/USD3117 15131604
CAD/JPY441 200241
CHF/JPY2975 14631512
EUR/AUD673 306367
EUR/CAD442 207235
EUR/CHF3025 15211504
EUR/GPB3136 15811555
EUR/JPY2987 15231464
EUR/USD3137 14901647
GBP/CHF3075 15671508
GBP/JPY2849 14131436
GBP/USD3117 15391578
NZD/USD2939 14581481
USD/CAD3120 15171603
USD/CHF3116 14641652
USD/JPY2846 14011444

Same direction 49.21%
Reverse Direction 50.79%
.
  • #58
  • Aug 11, 2009 6:34pm Aug 11, 2009 6:34pm
Quoting redpine
I like your approach is very similar to what I've been looking at. Below are the stats I tallied for the odds of a pair's direction on the next day.


Same direction 49.21%
Reverse Direction 50.79%

Nice, can you kindly do the same for the H4 timeframe?

The stats seems so close to 50/50, that is why if only we can let the win ride, and cut our losses, then we might have a winner indeed?!?

Besides, if we can throw in a nice money management, who knows? This might just be a winner? hahahaha ... wishful thinking?!?
  • #59
  • Aug 11, 2009 7:39pm Aug 11, 2009 7:39pm
Quoting 10pipsman
Nice, can you kindly do the same for the H4 timeframe?

The stats seems so close to 50/50, that is why if only we can let the win ride, and cut our losses, then we might have a winner indeed?!?

Besides, if we can throw in a nice money management, who knows? This might just be a winner? hahahaha ... wishful thinking?!?
I'll run some numbers for the the H4 timeframe. I've previously looked at the one hour time frame and the results are similar.

This is a tough nut to crack. I've looked at various options. The best I found is to only follow the leader when the leader is following the trend. If you use TP and SL levels of about 200 pips, you got a winner (backtest numbers are good). The problem with the approach is you'll have to stomach some big drawdowns.

To me the question of what the next bar will do comes down to the age old question of whether the market is trending or ranging. If you can quantify that, then I think this is a real winner. Always reverse in ranging markets and follow the leader in trending markets.
.
  • #60
  • Aug 12, 2009 3:11am Aug 12, 2009 3:11am
On a manual approachthe stoch. indicator @ 5.3.3 high/low smoothed seems to act as a reasonable filter on 4h but even some of the loosing positions could product some pips with a TSL. Just a thought
Koyco
0 traders viewing now