5 Min EUR/USD Strategy

Viewing 15 posts - 226 through 240 (of 251 total)

← 1 2 3 … 15 16 17 →

Author

Posts
03/03/2018 at 7:56 PM #64311 quote

Gertrade
Participant

Veteran

@ Grahal,

Indeed, v2.4.2 is the only strategy I developed on TF18s. Apparently, the TF18s and especially the TF28s are promising since we have several strategies that prove profitable over almost a month soon.

When to drawdowns, I have no information about this in ProOrder statistics, but from memory, I just know that these 2 strategies did not know like the drawdown v2 important and they never had a negative gain .

Thank you again for your interest in my work.

Cordially.
Gertrade

03/03/2018 at 8:11 PM #64313 quote

GraHal
Participant

Master

Gertrade the drawdown figure is on the Tab – Statistics of Closed Trades – see red arrows on attached

Thanks
GraHal

gert-39.jpg

03/03/2018 at 9:13 PM #64318 quote

Gertrade
Participant

Veteran

GraHal wrote:

the drawdown figure is on the Tab – Statistics of Closed Trades

Unfortunately the value of drawdowns is 0 (below the stats)

Gertrade

GraHal thanked this post

Boll-WR-28-sec-compare-V2.2.2-V2.4.2-07-02-03-03-Stats-ProOrder.jpg

03/03/2018 at 9:36 PM #64320 quote

Gertrade
Participant

Veteran

below the equity curves of the last week (no more history)

Gertrade

GraHal thanked this post

Boll-WR-28-sec-compare-V2.2.2-V2.4.2-07-02-03-03-Equity-Curve-ProOrder.jpg

03/04/2018 at 12:24 PM #64352 quote

Aloysius
Participant

Veteran

@Gertrade,

If you change the timeframe in this window (for 1 mn for ex), you can see (and show us) the equity curve since the begining of the strategy, even if the strategy is in a different TF.

Gertrade and manel thanked this post

03/04/2018 at 6:12 PM #64369 quote

Gertrade
Participant

Veteran

Aloysius wrote:

If you change the timeframe in this window (for 1 mn for ex), you can see (and show us) the equity curve since the begining of the strategy

Great! Awesome! Bravo! it’s a real pleasure this trick to see the equity curve from the beginning.

below the equity curve graphs.

GraHal and Aloysius thanked this post

Boll-WR-28-sec-compare-V2.2.2-V2.4.2-07-02-03-03-Equity-Curve-ProOrder-25500-units.jpg

03/06/2018 at 7:29 PM #64614 quote
manel
Participant

Veteran
Hi guys – Just an update from my end on testing of the original 5m versions in Live. I did test a few of the 28s versions initially but they were not very successful at all and the backtests and live were completely different so I gave up on that TF as I feel it’s a bit of a lottery and as a general rule I’m always a bit wary of running anything where there is such limited historic data. But I see that they have been successful for some of you so well done !

Anyway, with regards to the 5m versions I tested EUR/USD 1.1, 1.2.2 and 1.2.3 and GBP/USD 1.1. Following are key highlights :
- Live vs Backtest match perfectly and results have shown a net gain over last 3 weeks.
- As an observation, the strategies only have a high success rate in bullish price patterns – they tend to be much less successful in sideways or bearish price trends. This may create an opportunity to filter some of these losing trades out via a stochastic momentum, RSI indicator etc etc – For next stage of development.
- I changed the trailing stop to a fixed one as we are aware that the inbuilt PRT trailing code is not always accurate.
- I amended the stop losses to a much lower level than the original 90 odd pips as some of them I felt were too wide and caused too big a draw. I used 25-35 in most cases and that provides a more acceptable risk reward given the high 70%+ win rate. Each person can adjust to suit their risk profile.
- Short trades had a higher win rate than long trades historically in backtesting so I doubled the position size for shorts. Given that there are far fewer short trades than long this has not affected the results in a meaningful negative manner if there are losses on these types of trades.
- I have assumed an increased spread in my backtest of 0.8 whilst testing to allow for the occasional widening of market spreads.
- Attached are the backtest results of the original EUR/USD v1.1 vs 1.11 (amended) over 200k units with the slight adjustments mentioned above to provide an example of the tweaks made in each version.
Overall, I think this shows some good promise but there is more work that can be done to improve it, especially the risk reward and the average gain/trade. Thanks again to all those above who have been working on it. I will provide more updates in the future if anything interesting comes up.

GraHal, Aloysius and ltrade thanked this post

BolWR-1.1-Comp.png BollWR-5-min-EURUSD.v1.11.itf
03/08/2018 at 10:25 AM #64754 quote

Aloysius
Participant

Veteran

Thank you very much Manel, to improve this code in 5mn, which seems reliable. I will try to work on it too. The backtest works on my platform, But when I try live trading in demo, it is impossible and PRT tells me: “combined stops can not be used with Proorder”. But I see nothing like this in the code. Is someone know where is the problem? Thanks.

03/11/2018 at 6:42 PM #65011 quote

Cyril 21
Participant

Junior

Hi,

Here are some ProOrder results for several versions of the algo.

These versions run for 2 or 3 weeks now.

Unfortunately, all the 28-second versions have big losses. The best 30-second version that I kept (v5.12) is a little bit negative too.

But the 5-minute @Gertrade version (v 7.0 for EUR-USD) seems to have great performance on these 2 weeks. On the right of the picture are the details about it.

It seems that the 5-minute TF isn’t bad… this week I will try the @Gertrade v1.1, and the @manel v1.11 version, and will let you know after some days…
Note that I stopped the DAX30 v7.01 that had more than €1100 losses in 1 week.

Have a good week
Cyril

Gertrade, GraHal and ltrade thanked this post

ProOrder_Results.jpg

03/11/2018 at 7:12 PM #65013 quote

Cyril 21
Participant

Junior

@manel : I have the same problem than Aloysius with your v1.11 version : “combined stops can not be used with Proorder”. Is somebody able to explain that ? Thanks

03/11/2018 at 8:48 PM #65016 quote

manel
Participant

Veteran

Hi – Let me check again and see if it happens with me. It’s strange.

03/11/2018 at 9:10 PM #65017 quote
manel
Participant

Veteran
Well, this is strange, I’ve tried starting the code again in my platform and I do indeed get the same error message. Don’t see how this can be as the code I am running live is the same (bar my position size is different) and is working fine. And this code as you can see only has one coded stop function so I don’t understand why ProOrder seems to think that there is another conflicting stop in there. I’m perplexed as to why this is happening, hopefully someone with a bit more experience can help debug. I’ve attached the code below.
```
//100218
//Changes from original code - amended trailing stop to fixed stop with 28 optimised. Doubled short position size. Increased assumed market spread to 0.8

// Definition of code parameters
DEFPARAM CumulateOrders = False // Cumulating positions deactivated

// Prevents the system from placing new orders on specified days of the week
daysForbiddenEntry = OpenDayOfWeek = 6

// Conditions to enter long positions
indicator1 = BollingerDown[20](close)
c1 = (close <= indicator1)
indicator2 = ADX[14]
c2 = (indicator2 <= 17)
indicator3 = Williams[14](close)
c3 = (indicator3 <= -90)

IF (c1 AND c2 AND c3) AND not daysForbiddenEntry THEN
BUY 1 PERPOINT AT MARKET
ENDIF

// Conditions to exit long positions
indicator4 = Average[20](close)
c4 = (close CROSSES OVER indicator4)
c5 = (close CROSSES UNDER close[1])

IF c4 AND c5 THEN
SELL AT MARKET
ENDIF

// Conditions to enter short positions
indicator5 = BollingerUp[20](close)
c6 = (close CROSSES OVER indicator5)
indicator6 = ADX[14]
c7 = (indicator6 >= 49)

IF (c6 AND c7) AND not daysForbiddenEntry THEN
SELLSHORT 2 PERPOINT AT MARKET//1
ENDIF

// Conditions to exit short positions
indicator7 = Average[20](close)
c8 = (close CROSSES UNDER indicator7)
c9 = (close CROSSES OVER close[1])

IF c8 AND c9 THEN
EXITSHORT AT MARKET
ENDIF

// Stops and targets
SET STOP ploss 28//pTRAILING 30
SET TARGET pPROFIT 13
```
03/11/2018 at 9:28 PM #65018 quote

manel
Participant

Veteran

Solved it ! I deleted the stop after //”pTRAILING 30″ in line 51 above and its now fine ! Makes absolutely no sense as PRT is supposed to ignore anything that comes after //….. Maybe it’s a bug, will need to get them to look into it. In the meantime, I’ve reattached the file without this so it should be fine to run now.

Be aware, like I mentioned above, this needs a bit more work as I would ideally like the risk/reward to be much higher. Am working on some adjustments and one in particular looks promising and will post here after live testing if it’s looking good.

Cyril 21 and GraHal thanked this post

BollWR-5-min-EURUSD.v1.11-1.itf

03/12/2018 at 8:54 AM #65032 quote

Nicolas
Keymaster

Master

It is not possible to have at the same time SET STOP LOSS and a TRAILING instruction in a strategy running live through ProOrder.

03/12/2018 at 9:22 AM #65036 quote

Vonasi
Moderator

Master

Nicolas wrote:

It is not possible to have at the same time SET STOP LOSS and a TRAILING instruction in a strategy running live through ProOrder.

But surely if the pTRAILING stop is remarked out using // then it is not technically in the strategy?
Author

Posts