Monthly Profit - Daily Trading using SMA | Page 19 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Daily Profit - H1 Trading using SMA 751 replies

Jake Bernstein 10 SMA high 8 SMA low trading system 94 replies

SMA 200+SMA 55+SMA 5 is very profitable 49 replies

AUDUSD/AUDCHF/NZDJPY basket trading using 50 sma and 200 sma 16 replies

Programming of SMA and SMA cross over 20 replies

Attachments: Monthly Profit - Daily Trading using SMA
Exit Attachments

Monthly Profit - Daily Trading using SMA

  • #361
  • Aug 13, 2016 10:18am Aug 13, 2016 10:18am
Quoting cuchuflito
This system is crying for an accurate Back Test...from an EA wizard! (not me...I have no programming skylls) 1 Identify a basquet of, 10,15,20 pairs...hopefully non correlated. 1 Run the test. 3 Adjust parameters..? Shouldnt be that difficult to program, rules are simple.
You're right. Also thought the same already. However, I find EA programming in MT a pain in the ass, compared to NinjaTrader and C#.

So I took the opportunity today and spent some hours programming a strategy for NT7 using C#. I had it running over all the major pairs from 01/01/16 to 08/12/16 and the results are surprisingly good:
Attached Image (click to enlarge)
Click to Enlarge Name: sEbOqge.png Size: 228 KB


I attached the code for the strategy. Not sure if I implemented it 100% correctly. The script doesn't care for sideways or counter trend for instance. Just rename FX1DSMA5_3.txt to FX1DSMA5_3.cs. Couldn't upload .cs files - it said "File type not supported".
Attached File(s)
File Type: txt FX1DSMA5_3.txt   13 KB | 718 downloads
  • #362
  • Aug 13, 2016 2:54pm Aug 13, 2016 2:54pm
Quoting mw108
{quote} You're right. Also thought the same already. However, I find EA programming in MT a pain in the ass, compared to NinjaTrader and C#. So I took the opportunity today and spent some hours programming a strategy for NT7 using C#. I had it running over all the major pairs from 01/01/16 to 08/12/16 and the results are surprisingly good: {image} I attached the code for the strategy. Not sure if I implemented it 100% correctly. The script doesn't care for sideways or counter trend for instance. Just rename FX1DSMA5_3.txt to FX1DSMA5_3.cs. Couldn't...
AWSOME mw
didnīt expect soemone to come up with it so fast.
Great job, and yes, itīs good results...
Interesting to see a 52,94% win rate...sounds realistic..

THANKS for sharing.
  • #363
  • Aug 13, 2016 4:13pm Aug 13, 2016 4:13pm
Quoting cuchuflito
{quote} AWSOME mw didnt expect soemone to come up with it so fast. Great job, and yes, its good results... Interesting to see a 52,94% win rate...sounds realistic.. THANKS for sharing.
Thank you. I had that idea already in mind. Just needed some reason to "kick-start" it.

I updated the algorithm a bit. It has more clear entries now, tries to find "reverse signals" and exits the position if required. I also stopped using a fixed profit target, but implemented some sort of Trailing Stop instead. Also fixed some bugs, cleaned up the code a bit and added more comments.

EURJPY was butchering the script before. Now it comes out with good profit. I also disabled "Double on loss" for this run. Overall tho, the percentage went down from 52,9% to 46,3%, but the profit went up significantly, because the winners can run longer now. Also with "Double on loss" on the performance didn't change much, even went slightly down.

I think if you make the script more "intelligent", you can even more reduce the drawdowns and increase the profits.

Attached Image (click to enlarge)
Click to Enlarge Name: zDWb2JK.png Size: 236 KB


I attached the updated code for those interested. I think I let this run on a Sim account for a while and see how it performs on the real market.
Attached File(s)
File Type: txt FXD1SMA53.txt   18 KB | 700 downloads
  • #364
  • Aug 13, 2016 4:34pm Aug 13, 2016 4:34pm
An image from the USDJPY chart to see the implementation "at work":

Attached Image (click to enlarge)
Click to Enlarge Name: fpChMvZ.png Size: 150 KB


As you can see the script doesn't take all signals, because they don't match the criterias. You can fine tune it more, but then you also increase the chance of false entries. But overall I think it nicely represents the strategy as presented by Bk8.
  • #365
  • Aug 14, 2016 5:11am Aug 14, 2016 5:11am
Quoting mw108
{quote} I attached the updated code for those interested. I think I let this run on a Sim account for a while and see how it performs on the real market.
Thanks for the code mw108, i use it to upgrade my coding skills in C# for Ninja.
Right now i'm doing all my coding in AFL for amibroker which is much easier to learn then C#.
I like amibroker very much, and it get's the job done very well, but some coding skills in C# always come in handy.
Persistence will get it. Consistency will keep it.
  • #366
  • Aug 14, 2016 5:24am Aug 14, 2016 5:24am
Quoting mw108
An image from the USDJPY chart to see the implementation "at work": {image} As you can see the script doesn't take all signals, because they don't match the criterias. You can fine tune it more, but then you also increase the chance of false entries. But overall I think it nicely represents the strategy as presented by Bk8.
Nice work, can you tell us what exactly those criterias are ?
  • #367
  • Aug 14, 2016 8:51am Aug 14, 2016 8:51am
Quoting Hoffi4201
{quote} Nice work, can you tell us what exactly those criterias are ?
For Short entries:
- Current price must be below SMA
- Last bar must have been opened above and closed below SMA
- 2nd and 3rd last bars must have been closed above SMA

For Long entries just vice versa.
  • #368
  • Aug 14, 2016 11:19am Aug 14, 2016 11:19am
nice job!
Vucking good EA coder... https://t.ly/AZjRM
  • #369
  • Aug 14, 2016 11:20am Aug 14, 2016 11:20am
I may have a go at coding my implementation. Out of interest how do most of you identify a sideways market? or an indication of a strong trend?
So many in the Commercial section want to make me rich! :o
  • #370
  • Aug 14, 2016 11:59am Aug 14, 2016 11:59am
Hey mw,

why no long openings here:

Attached Image (click to enlarge)
Click to Enlarge Name: a.JPG Size: 71 KB
  • #371
  • Aug 14, 2016 12:17pm Aug 14, 2016 12:17pm
Here is EURUSD 1D from 2015 - Today
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Screen Shot 2016-08-14 at 18.16.09.png Size: 68 KB
Click to Enlarge Name: Screen Shot 2016-08-14 at 18.16.01.png Size: 106 KB
So many in the Commercial section want to make me rich! :o
  • #372
  • Aug 14, 2016 12:36pm Aug 14, 2016 12:36pm
Quoting Hoffi4201
Hey mw, why no long openings here: {image}
Indeed. Need to check this.
  • #373
  • Aug 14, 2016 5:02pm Aug 14, 2016 5:02pm
Quoting Hoffi4201
Hey mw, why no long openings here: {image}
I played around with it a little bit and added the entry conditions you mentioned. But this doesn't increase overall profitability because you will also get more false entries. Also keep in mind that using the new condition, the script will recognize a valid entry point only after the signal candle has closed.

I also added an attempt to detect "Sideways" markets, they are marked as gray areas on the chart.

USDJPY:
Attached Image (click to enlarge)
Click to Enlarge Name: J7jdYty.png Size: 311 KB


Combined results:
Attached Image (click to enlarge)
Click to Enlarge Name: m6oFNLq.png Size: 244 KB


Even tho there are huge numbers, I find this a bit worrying because all the other pairs are losing money, which isn't a good sign because we shouldn't rely on "carriers" like EURJPY or USDJPY. I will check later if I can tweak it a bit more without "overtuning" it.

I attached the updated code.
Attached File(s)
File Type: txt FXD1SMA53.txt   23 KB | 598 downloads
  • #374
  • Aug 14, 2016 5:04pm Aug 14, 2016 5:04pm
Quoting mw108
{quote} I played around with it a little bit and added the entry conditions you mentioned. But this doesn't increase overall profitability because you will also get more false entries. Also keep in mind that using the new condition, the script will recognize a valid entry point only after the signal candle has closed. I also added an attempt to detect "Sideways" markets. USDJPY: {image} Combined results: {image} Even tho there are huge numbers, I find this a bit worrying because all the other pairs are losing money, which isn't a good sign because...
Yeah im afraid you are going to "fix it" till its broke. Your second code posting seems to be the best and I would def keep that one as the live test. You can keep tweaking to try and see but never lose something that works.
  • #375
  • Aug 14, 2016 11:48pm Aug 14, 2016 11:48pm
Hi,
Anyone spot good setup for today?
  • #376
  • Aug 15, 2016 12:26am Aug 15, 2016 12:26am
Try gold, signal is old though, valid setup.
  • #378
  • Aug 15, 2016 2:44am Aug 15, 2016 2:44am
Quoting imperative
Try gold, signal is old though, valid setup.
Gold stochastic is at 20 range. It's a bit high risk.

i am in for usdjpy. Is it a right setup /valid signal?
Attached Image (click to enlarge)
Click to Enlarge Name: image.png Size: 235 KB
  • #379
  • Aug 15, 2016 3:18am Aug 15, 2016 3:18am
Quoting maniac
Hidden
I found it amusing thanks
So many in the Commercial section want to make me rich! :o
  • #380
  • Aug 15, 2016 3:50am Aug 15, 2016 3:50am
Quoting domdom1984
{quote} Gold stochastic is at 20 range. It's a bit high risk. i am in for usdjpy. Is it a right setup /valid signal? {image}
I wouldn't take that, unless you are looking for a short term move. Chances are high it bounces off at 100.820-100.557 area.
3 traders viewing now