Quick and Dirty EOD | Page 27 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

EOD trading system 1,037 replies

EOD -1 day hold 213 replies

Quicker and Dirtier EOD 117 replies

Best EOD strategy 5 replies

The meaning of EOD? 3 replies

Attachments: Quick and Dirty EOD
Exit Attachments

Quick and Dirty EOD

  • #521
  • Aug 21, 2012 3:38am Aug 21, 2012 3:38am
Quoting forexdan
Here's a screen shot of the back test results, EURUSD, 1-1-2011 to 1-1-2012.

There could be something wrong with my history data, or maybe it's just Metatraders backtester. IDK. Results are often iffy.

Still, I tend to use positive back test results as a starting point to forward demo testing
9% doesn't seem worth it but I just noticed the max drawdown is around 2%! I'd be comfortable with a max drawdown nearer 20% so there is definitely scope to risk more per trade here.
  • #522
  • Aug 21, 2012 3:59pm Aug 21, 2012 3:59pm
Quoting mikeh3855
I have a cross to the upside on the monthly chart and a pull back candle on the recent daily uptrend on the CADJPY. My broker's daily candle closes at GMT -5. I'll have a buy stop order placed above the swing high candle just to the left of the pullback candle. There might be a little resistance right there. The stop is 48 pips away at the bottom of the signal candle. The TP is 100 pips. 1:2

I like the original 1:2 TP idea. In any good 3 week trend you see 3 or 4 trades in there that become successfull. Taking profit out of each one of those is...
Well, that didn't go to well.
  • #523
  • Aug 22, 2012 12:49am Aug 22, 2012 12:49am
Quoting mikeh3855
Well, that didn't go to well.
It's not always going to and it's definitely harder at this time of year. Remember that oil is in an uptrend so cadjpy should in theory be rising too.
  • #524
  • Aug 22, 2012 4:16am Aug 22, 2012 4:16am
Quoting spyderman
If you're using MT4's strategy tester I would be very skeptical about the results, it's notoriously inaccurate. Perhaps you could post the results for others to see and interpret.
There are ways how to go around MT4 strategy tester flaws. One way is to import tick data and use Birt's scripts for that matter. Another way, is just to import M1 data which is more than enough for the EOD EA's. To make sure if results are accurate we can run the same EA live for a couple of weeks, and then afterwards run a backtest with the same settings on historical data for the comparison. If results do not differ a lot, then backtesting proves to be reliable enough in this case at least.

Here are the links to the results of different versions of EA based on Hugh Bruss EOD rules. I have also modified Yanyun G. original code and added some money management, see attachments (I gave it a name of HughBrissEOD). Here are the results on myfxbook.com, that were run on 10 years of data for EURUSD:

1. HughBrissEOD with no MM(Yanyun G., JuliuY) +2.71% Gain

2. HughBrissEOD with MM(Yanyun G., JuliuY) +67.69% Gain

3. HughBriss-ExForx-no-mm (Yanyun G., ExForx) +5.3% Gain

4. HughBriss-ExForx-with-mm (Yanyun G., ExForx) +5.24% Gain

5. HughBriss-dawn2 (forexdan) -2.17% Gain

I noticed that ExForx EA was performing better than all other versions, so I have changed a little the ExForx code by adding different Money Management (EOD2 in attachment). Here are the results:

6. HughBriss-EOD2 (Yanyun G., ExForx, JuliuY) +47.84% Gain

If you are guys are interested I can run these EA's on different pairs, and maybe even start Forward Testing.

Cheers!

Julius
Attached File(s)
File Type: mq4 HughBrissEOD.mq4   10 KB | 232 downloads
File Type: mq4 EOD2.mq4   10 KB | 258 downloads
  • #525
  • Aug 22, 2012 4:20am Aug 22, 2012 4:20am
Quoting ExForX
Sure have a look at it. Mine produced some profits but could be better. I downloaded the standard EURUSD history prices from my broker and has 90% modeling quality. Not 99% but that should provide us with an idea if this strategy is any good or not. The fact that it's profitable already means something to me, but as I said, could be optimized.

Not sure over-optimisation is always a good thing to do though. I also agree that some "discretion" is needed in trading but strangely enough that's frowned upon by a lot of people who believe you should...
90% or 99% for EOD it does not make a big difference. EA can also be used as a signal, one can run it on a demo account and setup up alerts when it makes a trade or a pending order. When it can be checked manually and if it is a good trade, order can be taken on a live account. That is how guys on CIRC thread do. Yes, if not everything at least some things can be semi-automated.
  • #526
  • Aug 22, 2012 4:22am Aug 22, 2012 4:22am
Quoting Hugh Briss
Excellent, thank you. I think at some stage I am going to get my old friend FerruFx to looks at this one. He is an awesome coder and if it can be automated exactly to the rules then he can do it.I do of course gratefully appreciate your efforts.
Hugh, no need to trouble FerruFx, unless he really has so much time. We already have four coders on the thread. If you have exact rules it would be our pleasure to program a code that would behave exactly on those rules.
  • #527
  • Aug 22, 2012 4:25am Aug 22, 2012 4:25am
Quoting Hugh Briss
Why? I have manually backtested lots of eod systems and then run an ea through the strategy tester and got very similar results. I think with eod data it is more reliable.

Can anyone suggest a more reliable way to back test systems? Amibroker? Tradestation? Someone else?
Hugh, actually a year ago I and my partner were running parallely the same EA both on TradeStation and MT4. I used the same quotes for both platforms, and it was possible to synchronize them that they would do almost exactly the same trades and give identical results. Although because of some idiosyncrasies, sometimes it was time consuming, and maybe not necessary. The good thing about Tradestation is that it runs optimization ten if non hudrend times faster than MT4.
  • #528
  • Aug 22, 2012 4:42am Aug 22, 2012 4:42am
Quoting juliuy
Hugh, no need to trouble FerruFx, unless he really has so much time. We already have four coders on the thread. If you have exact rules it would be our pleasure to program a code that would behave exactly on those rules.
Fair enough
  • #529
  • Aug 22, 2012 4:48am Aug 22, 2012 4:48am
Quoting juliuy
Hugh, actually a year ago I and my partner were running parallely the same EA both on TradeStation and MT4. I used the same quotes for both platforms, and it was possible to synchronize them that they would do almost exactly the same trades and give identical results. Although because of some idiosyncrasies, sometimes it was time consuming, and maybe not necessary. The good thing about Tradestation is that it runs optimization ten if non hudrend times faster than MT4.
I've never looked at tradestation, I will have to have a look.
  • #530
  • Aug 22, 2012 6:08am Aug 22, 2012 6:08am
Hi there

just finished reading the entire thread. Great stuff - with a fulltime job and a house full of kids - this style of trading is perfect for me

just wondering - theres has been some comments regarding correlated pairs - and Hugh has a good point about riskmanagement in this regard - however, if we were to compose a watchlist of pairs that does not pose a risk of being correlated - wich pairs would be a candidate for this?

Thanks for a great thread.
  • #531
  • Aug 22, 2012 9:26am Aug 22, 2012 9:26am
Quoting rschou
Hi there

just finished reading the entire thread. Great stuff - with a fulltime job and a house full of kids - this style of trading is perfect for me

just wondering - theres has been some comments regarding correlated pairs - and Hugh has a good point about riskmanagement in this regard - however, if we were to compose a watchlist of pairs that does not pose a risk of being correlated - wich pairs would be a candidate for this?

Thanks for a great thread.
Hello, glad you like the thread.

I think the best way to approach this is to watch all pairs but just be careful of overtrading. For instance it is a good idea to watch the euro against all other currencies but if they all set up at the same time be selective in the trades you take. If aud and nzd are the strongest currencies at that time then perhaps just trade those crosses. If it's cad and jpy strong then trade eurcad and eurjpy. I wouldn't want to trade more than about 2 crosses on the same currency and I factor that in to my risk management.
  • #532
  • Aug 28, 2012 3:38pm Aug 28, 2012 3:38pm
It's been a little bit quiet with the holidays and whatever so hopefully we'll be back on some trades soon.

I'm thinking of simplifying the system a little. I like the look of the monthly to give overall direction and/or the 100 and 200 sma's on the daily chart. Then I think a 10 ema would be enough. Add some horizontal s/r and trendlines and trade the trends that occur moving away from these levels. In effect what you are doing is not so much trading bounces from s/r but trading the subsequent minor trend when it moves in the direction of the overall trend. Combined with the Q+D entry method it should be a winner.

I'll post more on my ideas soon.
  • #533
  • Aug 28, 2012 5:29pm Aug 28, 2012 5:29pm
Quoting Hugh Briss
It's been a little bit quiet with the holidays and whatever so hopefully we'll be back on some trades soon.

I'm thinking of simplifying the system a little. I like the look of the monthly to give overall direction and/or the 100 and 200 sma's on the daily chart. Then I think a 10 ema would be enough. Add some horizontal s/r and trendlines and trade the trends that occur moving away from these levels. In effect what you are doing is not so much trading bounces from s/r but trading the subsequent minor trend when it moves in the direction of the...
Looking forward to reading your ideas...

Olu
If you think you can or cannot do it you are correct
  • #534
  • Edited 10:22pm Aug 28, 2012 9:10pm | Edited 10:22pm
Quoting Hugh Briss
It's been a little bit quiet with the holidays and whatever so hopefully we'll be back on some trades soon.

I'm thinking of simplifying the system a little. I like the look of the monthly to give overall direction and/or the 100 and 200 sma's on the daily chart. Then I think a 10 ema would be enough. Add some horizontal s/r and trendlines and trade the trends that occur moving away from these levels. In effect what you are doing is not so much trading bounces from s/r but trading the subsequent minor trend when it moves in the direction of the...
Man this is starting to sound pretty complicated....lol
How about this instead... nice and easy, ..... easy peazy.
http://screencast.com/t/b0w9ueuNV5R
The Weighted MA is faster than the Simple MA and as such is an indication of the trend.
Price is above the MAs and the WMA is above the SMA "trend is up" vice versa for down trend.
  • #535
  • Aug 29, 2012 12:44am Aug 29, 2012 12:44am
Quoting tradestar1
Man this is starting to sound pretty complicated....lol
How about this instead... nice and easy, ..... easy peazy.
http://screencast.com/t/b0w9ueuNV5R
The Weighted MA is faster than the Simple MA and as such is an indication of the trend.
Price is above the MAs and the WMA is above the SMA "trend is up" vice versa for down trend.
Lol, in my head it's getting simpler. I like the look of your idea too though. I'm sure there are lots of ways to do it. I like the look of the 10 ema on the chart. Once it starts moving one way above or below it, price can stay there for a long time moving one way or the other. The only thing we need to work out how to do is to avoid as much of the chop as possible at the start of one of these runs. Losses are going to be unavoidable while we are waiting for a move to get going but by waiting for the trend to be at least a little established rather than by trying to catch every move from the start I think we will miss quite a few losers.
  • #536
  • Aug 29, 2012 4:55am Aug 29, 2012 4:55am
nice one tradestar1
- seems interesting. Tell me, do you enter at the break of the retracement candle, or do open as soon as the daily candle closes? SL as I understand it is at the high or low of the retracement candle, depending on trend?

best regards
  • #537
  • Aug 29, 2012 8:11am Aug 29, 2012 8:11am
Quoting rschou
nice one tradestar1
- seems interesting. Tell me, do you enter at the break of the retracement candle, or do open as soon as the daily candle closes? SL as I understand it is at the high or low of the retracement candle, depending on trend?

best regards
Entry is at the breakout of the candle.
  • #538
  • Aug 29, 2012 11:21am Aug 29, 2012 11:21am
Yep, I'm seeing the idea now, forget the overall trend, just trade the direction it is moving in right now. You will take lots of small losses as the trends chop and change but if you keep stacking the trades when you get a good run you will clean up.
  • #539
  • Aug 29, 2012 5:24pm Aug 29, 2012 5:24pm
Quoting Hugh Briss
Yep, I'm seeing the idea now, forget the overall trend, just trade the direction it is moving in right now. You will take lots of small losses as the trends chop and change but if you keep stacking the trades when you get a good run you will clean up.
And by "overal trend" do you mean D1 trend or both D1 and H1?

Indeed, I also noticed when backtesting that "forgetting D1 trend" if you do it smartly doesn't make the system less profitable.

There are a lot of trades like when one I am attaching to this post. D1 was down, but looked like going to reverse, and H1 gave a good signal.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: gbpsgd d1 trade on h1.gif Size: 17 KB
Click to Enlarge Name: gbpsgd h1 trade.gif Size: 20 KB
  • #540
  • Aug 30, 2012 12:00am Aug 30, 2012 12:00am
Quoting juliuy
And by "overal trend" do you mean D1 trend or both D1 and H1?

Indeed, I also noticed when backtesting that "forgetting D1 trend" if you do it smartly doesn't make the system less profitable.

There are a lot of trades like when one I am attaching to this post. D1 was down, but looked like going to reverse, and H1 gave a good signal.
Yes, I mean forget the overall trend and just trade the direction that price is moving whichever side of the 10 ema/wma. That template is no good for this system though, you need normal candles to see the price levels.
1 trader viewing now