12_100_9 Trending System | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Hedge Wedge Trending System 15 replies

Simple Trending System 19 replies

5 Point System (for trending markets) 16 replies

Scaling In System for Trending/Ranging Currencies 24 replies

Part Time Wealth Building System - "High-Low Trending" 9 replies

Attachments: 12_100_9 Trending System
Exit Attachments

12_100_9 Trending System

  • #41
  • Edited 10:42pm Aug 19, 2007 10:27pm | Edited 10:42pm
Quoting tkimble
OK. This is a great trending system that is very simple.

1. Create a one hour chart and any pair but the more volatile pairs are more profitable.
2. Add the MACD indicator using the 12, 100, 9 parameters.
3. Buy when the fast average passes over the slow average.
4. Sell when the fast average passes under the slow average.
5. Use 50 PIP stops.

That is the entire system. 100% mechanical. Simple but very effective...

Happy Trading,
Tyrone
Tyrone,

Thanks for volunteering your system.

Just for correctness, you talk about the "fast average" and the "slow average". If you're using the default MACD on MT4 (don't know if you are or not), when the fast MA (12) and the slow MA (100) cross then the histogram changes from positive to negative. I believe your signal is when the histogram passes above or below the *signal line* -- which is simply a MA (9) of the histogram values. Correct me if I'm wrong (in which case your signal is for a zero-line MACD cross). If I'm right (and it is indeed the histogram/signal cross which is your trigger), I would question how much you know about how MACD works, or perhaps you just got your terminology confused...?

- Could you explain how this system will perform any differently than any other MACD-based system?
- Why is the "100" EMA value different and why does it perform better than the default "26" EMA?
- How will this perform in a sideways market? (see picture)
- Do you have any performance data for this system?

I ask these questions for two purposes:

1) Because I'd like to know how well this system will perform, and you should be the best authority on it!
2) Because people here tend to blindly accept systems without having an understanding of how they work (I'd wager a bet that the majority of people reading this thread right now have little understanding of how something as popular and ubiquitous as MACD even works, despite its relative simplicity), and I'd like for those who might fall into this category to think a little bit about why something may or may not work, before spending valuable time chasing a "holy grail".

I'm not hijacking, or trying to put down your system--just trying to get answers and to get people to think. Thanks for posting your method!
Attached Image
  • #42
  • Aug 20, 2007 7:40am Aug 20, 2007 7:40am
Quoting JoshDance
Tyrone,

Thanks for volunteering your system.

Just for correctness, you talk about the "fast average" and the "slow average". If you're using the default MACD on MT4 (don't know if you are or not), when the fast MA (12) and the slow MA (100) cross then the histogram changes from positive to negative. I believe your signal is when the histogram passes above or below the *signal line* -- which is simply a MA (9) of the histogram values. Correct me if I'm wrong (in which case your signal is for a zero-line MACD cross). If I'm right (and it is indeed the histogram/signal cross which is your trigger), I would question how much you know about how MACD works, or perhaps you just got your terminology confused...?

- Could you explain how this system will perform any differently than any other MACD-based system?
- Why is the "100" EMA value different and why does it perform better than the default "26" EMA?
- How will this perform in a sideways market? (see picture)
- Do you have any performance data for this system?

I ask these questions for two purposes:

1) Because I'd like to know how well this system will perform, and you should be the best authority on it!
2) Because people here tend to blindly accept systems without having an understanding of how they work (I'd wager a bet that the majority of people reading this thread right now have little understanding of how something as popular and ubiquitous as MACD even works, despite its relative simplicity), and I'd like for those who might fall into this category to think a little bit about why something may or may not work, before spending valuable time chasing a "holy grail".

I'm not hijacking, or trying to put down your system--just trying to get answers and to get people to think. Thanks for posting your method!
Just to let you know, I backtested this strategy on E-Signal, over the Euro Futures tick data for the whole of 2004 and 2005. Euro futures is strongly correlated with EUR/USD for obvious reasons. This system is profitable straight out of the box.

I had better add that TKimble is not a fan of backtesting for reasons I completely understand therefore I did not post the results on here.
  • #43
  • Aug 20, 2007 2:19pm Aug 20, 2007 2:19pm
rogerha, what pairs performed well with your backtests and did you use any TP or trailing stops for exits?
thanks

Quoting rogerha
Just to let you know, I backtested this strategy on E-Signal, over the Euro Futures tick data for the whole of 2004 and 2005. Euro futures is strongly correlated with EUR/USD for obvious reasons. This system is profitable straight out of the box.

I had better add that TKimble is not a fan of backtesting for reasons I completely understand therefore I did not post the results on here.
  • #44
  • Aug 20, 2007 5:59pm Aug 20, 2007 5:59pm
Quoting rogerha
Just to let you know, I backtested this strategy on E-Signal, over the Euro Futures tick data for the whole of 2004 and 2005. Euro futures is strongly correlated with EUR/USD for obvious reasons. This system is profitable straight out of the box.

I had better add that TKimble is not a fan of backtesting for reasons I completely understand therefore I did not post the results on here.
Did you mechanically test it? Or visually? TKimble, if you wouldn't mind, and rogerha, if you don't mind, can we see the tests? That would be helpful.

Here's the thing rogerha--I am Mr. Optimistic--I WANT systems like these to work!!! But here's the problem, besides the fact that backtesting really just doesn't simulate live trading nearly well enough to matter--this system is nothing more than an MA crossover system, with a weak TP strategy. When the 12 crosses above the 100, buy. When it crosses below, sell. Even adding the 9 period MA does nothing more than add lag to existing lag, as all MAs are lagging.

I can eyeball the chart, and see how a trade might go very well--in fact, on a volatile pair like GBPJPY, a daily trade can easily get you 500 pips... you will get several like this. Maybe you will be somewhat profitible for several months--then all of a sudden you will get whipsawed to death and lose all of your profits when the market goes nuts, and the MAs can't keep up with the price. I would like to believe a simple MA crossover system can work, but one never has in anything even remotely long term, especially with a weak TP strategy.

I'm not saying don't try this strategy, don't get me wrong!--I'm just saying that after you try it and realize that it doesn't work long term, understand a system's strong points and weaknesses before you invest a lot of time into it. I'm not saying MACD isn't useful, because some successful traders use it--but like every single indicator that exists, it is based on price, and will never give you a clear vision of price, and will always lag behind it. MACD and RSI and Stochs and all this stuff can complement your trading system, but if you make it by itself your trading system, you won't succeed, even with good money management.
  • #45
  • Aug 20, 2007 8:13pm Aug 20, 2007 8:13pm
Quoting JoshDance
Did you mechanically test it? Or visually? TKimble, if you wouldn't mind, and rogerha, if you don't mind, can we see the tests? That would be helpful.

Here's the thing rogerha--I am Mr. Optimistic--I WANT systems like these to work!!! But here's the problem, besides the fact that backtesting really just doesn't simulate live trading nearly well enough to matter--this system is nothing more than an MA crossover system, with a weak TP strategy. When the 12 crosses above the 100, buy. When it crosses below, sell. Even adding the 9 period MA does nothing more than add lag to existing lag, as all MAs are lagging.

I can eyeball the chart, and see how a trade might go very well--in fact, on a volatile pair like GBPJPY, a daily trade can easily get you 500 pips... you will get several like this. Maybe you will be somewhat profitible for several months--then all of a sudden you will get whipsawed to death and lose all of your profits when the market goes nuts, and the MAs can't keep up with the price. I would like to believe a simple MA crossover system can work, but one never has in anything even remotely long term, especially with a weak TP strategy.

I'm not saying don't try this strategy, don't get me wrong!--I'm just saying that after you try it and realize that it doesn't work long term, understand a system's strong points and weaknesses before you invest a lot of time into it. I'm not saying MACD isn't useful, because some successful traders use it--but like every single indicator that exists, it is based on price, and will never give you a clear vision of price, and will always lag behind it. MACD and RSI and Stochs and all this stuff can complement your trading system, but if you make it by itself your trading system, you won't succeed, even with good money management.
JoshDance,

I back tested only a month. I looked at it on the Daily, Hourly, and the 15M charts. I also compared it to the default settings. What I found was the default settings would signal an earlier entrance than TK's settings, however, his settings smoothed out the MACD line eliminating some false signals (i.e. the default settings would cross the trigger line where TK's weren't as erratic.)

I normally use MACD for divergence (or Histogram divergence.) That being said, I am going to forward test his system on my demo account for the next 2 to 3 weeks and see if it is something I can integrate into my trading style. If it performs well on the demo, I will try it on my live account. Currently, I'm thinking more along the lines of using it for the long term trend on the Daily chart.

Whether I decide to use it or not, I greatly appreciate that TK has put the time into developing tools to increase profitability, and he is willing to share his work with the rest of the traders here on the forum.

Keep up the good work TK.
  • #46
  • Edited 1:04am Aug 21, 2007 12:53am | Edited 1:04am
Answers to the backtesting questions:

The system was mechanically tested over 1 year (2004) of Euro Futures data. This is not Forex data, but the movement and scale of Euro Futures is almost identical to that of EUR/USD, and in my experience is a good test for a strategy. The reason I use this data is because I have 2 years of full tick data for this instrument for 2004 and 2005.

The system as tested trades the cross of the MA's (or in this case the cross of the Histogram above/below 0), with a 50-Pip stop loss on a 15-minute chart.

Orders are Market orders *on the next bar AFTER the cross*, so slippage is taken care of to some degree in there. A good trader will be able to improve on these results purely because of the timing of entry.

Commission/Spread is not take into account in this backtest.

Exit is either at stop loss or stop and reverse at the next cross.

If the trade exits at stop loss it will wait for the next cross before entering rather than reversing at stop loss.

Attached is the HTML file (zipped) with the backtest, I will attempt to post screenshots below for those that struggle with the attachment.

But here is the important factor, everybody should listen to TKimble. This strategy will give you a positive expectation. It will not make you rich by default (no strategy will) and it does not necessarily mean you can trade it successfully. The only reason I backtest a strategy is to evaluate my expectation of that strategy, from there each individual must assess whether the strategy is likely to be profitable for them in live trading !

TKimble - I hope I have interpreted your rules properly ! Thanks for sharing this system.
Attached File(s)
File Type: zip tkimble_backtest.zip   120 KB | 620 downloads
  • #47
  • Aug 21, 2007 12:56am Aug 21, 2007 12:56am
1st Screenshot
Attached Image (click to enlarge)
Click to Enlarge Name: tkimble1.png Size: 106 KB
  • #48
  • Aug 21, 2007 12:57am Aug 21, 2007 12:57am
2nd Screenshot
Attached Image (click to enlarge)
Click to Enlarge Name: tkimble2.png Size: 125 KB
  • #49
  • Aug 21, 2007 12:58am Aug 21, 2007 12:58am
3rd Screenshot
Attached Image (click to enlarge)
Click to Enlarge Name: tkimble3.png Size: 45 KB
  • #50
  • Aug 21, 2007 12:59am Aug 21, 2007 12:59am
The E-Signal code:

function preMain() {
setPriceStudy(false);
setStudyTitle("TKimble");
setCursorLabelName("Upper Donchian", 0);
setCursorLabelName("Stop", 2);
setDefaultBarFgColor(Color.blue, 0);
setComputeOnClose()
//setColorPriceBars(true);
}

var xMACD = null;
var lotsize = 1;
var nStop = null;


function main(){

xMACD = macdHist(12,100,9);



if (Strategy.isLong() && high() >= nStop){
Strategy.doSell("Long Stop", Strategy.STOP, Strategy.THISBAR, Strategy.ALL, nStop);
}

if (Strategy.isShort() && low() <= nStop){
Strategy.doCover("Short Stop", Strategy.STOP, Strategy.THISBAR, Strategy.ALL, nStop);
}


if ( !Strategy.isLong() && xMACD.getValue(0) > 0 && xMACD.getValue(-1) < 0 ) {
Strategy.doLong("Long Signal", Strategy.MARKET, Strategy.NEXTBAR, lotsize );
nStop = close()-0.0050;
}

if ( !Strategy.isShort() && xMACD.getValue(0) < 0 && xMACD.getValue(-1) > 0 ) {
Strategy.doShort("Short Signal", Strategy.MARKET, Strategy.NEXTBAR, lotsize );
nStop = close()+0.0050;
}

if (Strategy.isInTrade() == false ) {
nStop = null;
}


return xMACD

}
  • #51
  • Aug 21, 2007 1:11am Aug 21, 2007 1:11am
The following are on 60 minute charts (before somebody asks me).
Attached Image (click to enlarge)
Click to Enlarge Name: tkimble4.jpg Size: 58 KB
  • #52
  • Aug 21, 2007 1:12am Aug 21, 2007 1:12am
60 Minute chart
Attached Image (click to enlarge)
Click to Enlarge Name: tkimble5.jpg Size: 64 KB
  • #53
  • Aug 21, 2007 1:12am Aug 21, 2007 1:12am
60 Minute Chart
Attached Image (click to enlarge)
Click to Enlarge Name: tkimble6.png Size: 62 KB
  • #54
  • Aug 21, 2007 2:22am Aug 21, 2007 2:22am
thanks rogerha
  • #55
  • Aug 21, 2007 3:41am Aug 21, 2007 3:41am
Hi Roger

Did you implement your tests on 15 min?

The strategy is for 1hr time frame, and on 1hr it would eliminate a lot of more chops for sure
Regards
  • #56
  • Aug 21, 2007 4:02am Aug 21, 2007 4:02am
Quoting howard
Hi Roger

Did you implement your tests on 15 min?

The strategy is for 1hr time frame, and on 1hr it would eliminate a lot of more chops for sure
Hi Howard,

Yes I did, I also did a 1hr chart (see above), and you are absolutely correct in your assesment, 1hr is more profitable due to considerably less trades/whip.

15min however increases the volume of trades which is helpful for proving a concept when backtesting (at least in my mind).
  • #57
  • Sep 29, 2007 2:17pm Sep 29, 2007 2:17pm
Isn't back testing data useless if you don't account for the Spread? It seems like that can either make or break a strategy.

keep up the great work guys!
  • #58
  • Sep 29, 2007 2:22pm Sep 29, 2007 2:22pm
Has anyone tested this strategy on a live account, I would be interested to hear the results

Cheers,

Pete
  • #59
  • Sep 30, 2007 5:40am Sep 30, 2007 5:40am

hallo guys,
stopping by this thread and noticed that it hav a very pleasing enviroment...gonna enjoy this thread

"Analysis Is Hard But Trade Is Easy"
  • #60
  • Sep 30, 2007 3:34pm Sep 30, 2007 3:34pm
Hi all

Can anybody pls help me.

For the MACD Trending system. I plotted the 1hour chart with the 12 100 9 parameters and I've even bolded/thickend the fast moving line which is 12. But my issue is at the line crosses, the fast line always looks crossing above the slow line, regardless whether the mkt is uptrending or downtrending. I've spend a lot of time studying the movement.



Best Regards
Neyi
0 traders viewing now