Try My EA! Need Group Feedback... | Page 18 | Forex Factory

Options

Go to
Permalinks
Print Thread

Related Instruments

Gold/USD

Similar Threads

USD/TRY and EUR/TRY 12 replies

Fibo Group Feedback 23 replies

Need feedback on my trading breakouts system 8 replies

One World Capital... Need feedback 2 replies

Need feedback on FXCM Educational workshops 3 replies

Attachments: Try My EA! Need Group Feedback...
Exit Attachments

Try My EA! Need Group Feedback...

  • #341
  • Dec 5, 2014 3:56pm Dec 5, 2014 3:56pm
FYI, I have another one running with Armada but I forgot the password and hence not able to change the investor password and not able to create Trade Explorer. The account was created from MT4 itself, hence no email with passwords. Any suggestion for me to retrieve the investor password friends?
  • #342
  • Dec 5, 2014 4:01pm Dec 5, 2014 4:01pm
Thepipster/Kashtrader
Thanks for your response. I think I will shut it down, reopen it Sunday pm and await some new code update when available . I,m sure this wont be the last time for a reboot.
  • #343
  • Edited 5:29pm Dec 5, 2014 4:15pm | Edited 5:29pm
Quoting tradeforlife
FYI, I have another one running with Armada but I forgot the password and hence not able to change the investor password and not able to create Trade Explorer. The account was created from MT4 itself, hence no email with passwords. Any suggestion for me to retrieve the investor password friends?
You can try emailing the broker, but I've never had the password they responded with work.
Otherwise you can start over...

Here is a broker summary of this session's trades, I boxed in the times to be easier to read (why do they cram it so close?)


Attached Image (click to enlarge)
Click to Enlarge Name: times.png Size: 110 KB



It appears to me everything closed plenty fast, it's just the market must have been going up and down really fast...???
the last two losers at the bottom must have opened right before the EA closed other trades, their loss is not much more than spread. Everything happened within 4 seconds! at 11 the last 3 trades opened, at 12 half of the opened trades closed, at 13 the other half of opened trades closed, at 14 the last 3 trades closed, and then the next few seconds the pending orders were closed. market moved $0.82 pretty quick! 30 trades X 0.80 =24...makes sense on my negative account balance. I would like to see how this compares to kosmo's statement.

Too close for pips, switching to points.
  • #344
  • Dec 5, 2014 6:42pm Dec 5, 2014 6:42pm
Quoting Thepipster
{quote} Is the orphan trade still not able to be closed manually, like the last one? If so, the platform or broker has some weird problem. I think a closeable orphan trade could happen, like kashtrader said, if a pending order is opened while the opened orders are being closed, but the EA is supposed to go back through and close them all. It probably is worth posting an updated EA once we're sure of modifications. ~Kashtrader...your FXCM account has four sell trades at 1189.03-same price! that might have something to do with it. My trades currently...
Ignore my FXCM. I am going to be using Armada. Will update the explorer soon. Yes, We need to have a good discussion about the grid level distances, size of the box etc. I am trying to figure out and see if I can bring some figures to the table.

Also, I am further updating the EA to do what you suggested. Close opposing equivalent trades out early. This will help with the final real closing.

Once this is done I will add logic to open more pending order when box boundary reached. I will send you a version soon for testing.

Then the next step would be to modify TP's so that we have broker side closing.
  • #345
  • Dec 5, 2014 7:01pm Dec 5, 2014 7:01pm
Quoting tradeforlife
{quote} Currently the closing of all active and pending orders are handled by a single function. Maybe we can split that into two say, CloseAllActiveOrders() and CloseAllPendingOrders(). When TP is hit, we will first close all the active orders and then close all pending orders. Again we call the CloseAllActiveOrders() once more from inside the CloseAllPendingOrders() at the end which will close the active orders if any present. just my 2 cents.
Already done. But still there was a problem as the ea would only retry 3 times to close each set.
  • #346
  • Dec 5, 2014 7:49pm Dec 5, 2014 7:49pm
?...sorry.. am i right to say that basically this EA trades around a price point and just stacks buy above and sells below?... because from the trades you have in demo.. it seems to be that way....
  • #347
  • Dec 5, 2014 8:08pm Dec 5, 2014 8:08pm
Quoting kashTrader
{quote} Already done. But still there was a problem as the ea would only retry 3 times to close each set.
ok, thanks Kash.
  • #348
  • Dec 5, 2014 9:57pm Dec 5, 2014 9:57pm
shiva can you please provide the exact settings you are using for armada. I need to test the upgrades to the EA. Thanks.
  • #349
  • Dec 6, 2014 1:33am Dec 6, 2014 1:33am
Quoting kashTrader
shiva can you please provide the exact settings you are using for armada. I need to test the upgrades to the EA. Thanks.
Hi Kash, I am using the default settings OOBE, Out Of Box Experience and not Out Of Body Experience Timeframe is irrelevant really but for sake of clarity, I have it running on a 5M chart.
  • #350
  • Dec 6, 2014 3:16am Dec 6, 2014 3:16am
Quoting shiva
{quote} Hi Kash, I am using the default settings OOBE, Out Of Box Experience and not Out Of Body Experience Timeframe is irrelevant really but for sake of clarity, I have it running on a 5M chart.
Hi shiva,
Thanks for this. Yes I understand that the default settings should give you a high return on armada because of stacking but you should also get boxed in more quickly (at least what I think). I will try the new version with defaults and see what happens.

Thanks

@Pips
Since yesterday I have been toying with the idea of expanding the grid as necessary so that we don't get boxed in. So I was thinking why start with 60 trades? Why not just start with 4 and as soon as one side of th box gets hit we expand that side by another 2 and so on. That way we should only have open trades and at the max 4 pending orders at any time. This will reduce the closing times. I have also already done this last and it works.

I have also coded the cancellation of equivalent orders it have to test that when market opens. So now it's really time to think what settings to use for the backtest as we will never get boxed in but need to optimise the settings for broker use and profitability. TP is the next thing on the table if I am a bit busy and its complex so will have to wait a bit. Let me know if you need anything else.
  • #351
  • Dec 6, 2014 3:32am Dec 6, 2014 3:32am
Quoting shiva
{quote} Hi Kash, I am using the default settings OOBE, Out Of Box Experience and not Out Of Body Experience Timeframe is irrelevant really but for sake of clarity, I have it running on a 5M chart.
Shiva could you please do a favour and check what is the max number of trades you had open at any given time during your testing period?
  • #352
  • Dec 6, 2014 4:49am Dec 6, 2014 4:49am
Quoting kashTrader
{quote} Shiva could you please do a favour and check what is the max number of trades you had open at any given time during your testing period?
Hi Kash, how do I do this? I have no clue.
  • #353
  • Dec 6, 2014 5:04am Dec 6, 2014 5:04am
Quoting shiva
{quote} Hi Kash, how do I do this? I have no clue.


For this you need to look at your trade report. The EA should be closing all open orders at the same time. So when you see a set of orders closing, you need to count the number of trades that were closed at the same this. This should give you the number of trades that were open during the cycle but had to be closed when the Target Equity was achieved.

Shiva, also, can you please kindly do a screenshot of the EA settings. I am in no way able to replicate 4% equity rise over the 5 days. It takes me months to get that rise. This is important because you may have stumbled upon a set of optimised settings.
  • #354
  • Dec 6, 2014 5:05am Dec 6, 2014 5:05am
Quoting kashTrader
{quote} For this you need to look at your trade report. The EA should be closing all open orders at the same time. So when you see a set of orders closing, you need to count the number of trades that were closed at the same this. This should give you the number of trades that were open during the cycle but had to be closed when the Target Equity was achieved. Shiva, also, can you please kindly do a screenshot of the EA settings. I am in no way able to replicate 4% equity rise over the 5 days. It takes me months to get that rise. This is important...
Also, whats your leverage?
  • #355
  • Edited 5:21am Dec 6, 2014 5:10am | Edited 5:21am
Quoting kashTrader
{quote} For this you need to look at your trade report. The EA should be closing all open orders at the same time. So when you see a set of orders closing, you need to count the number of trades that were closed at the same this. This should give you the number of trades that were open during the cycle but had to be closed when the Target Equity was achieved. Shiva, also, can you please kindly do a screenshot of the EA settings. I am in no way able to replicate 4% equity rise over the 5 days. It takes me months to get that rise. This is important...
Sure, give me a few hours, I will post these. Leverage is 1:500
Attached Image (click to enlarge)
Click to Enlarge Name: steady gold settings.jpg Size: 227 KB


Here is the statement sorted by closing time, I can't see how to calculate simultaneous open position lol
Attached File(s)
File Type: xlsx steady_gold_armada_statement.xlsx   570 KB | 527 downloads
  • #356
  • Dec 6, 2014 5:57am Dec 6, 2014 5:57am
Damn it.... keep on blowing my account with the increasing grid size.
  • #357
  • Edited 8:29am Dec 6, 2014 7:58am | Edited 8:29am
Quoting kashTrader
**** it.... keep on blowing my account with the increasing grid size.
Please don't take this the wrong way, because I really appreciate your help and taking initiative to improve the system. I don't think you'll be able to use backtesting at all for this.
I do agree with your idea to start with a lot less trades, I would use 5 a side, and add extras if opposite orders are opened, maybe 5 more per other side opened? It shouldn't blow the account, unless the broker doesn't allow hedged orders to use no margin.
If you have the EA coded to close a hedged set of orders, just make sure it doesn't affect the target. When all the trades are closed, the EA has to make a net profit, preferably close to the target in settings.

It's kinda tough to figure out when trades closed on shiva's account, because some trades are new ones that hit target really fast with just 2-3 trades, and started over again, all in just 1-2 seconds, but may appear to have been part of an old set. That's one of the strong advantages of the EA. I've tried figuring out Kosmo's history, but it's ridiculous. He has mediocre latency. And he only increased the target to 2 (0.2%) and changing the buffer to 2 only adds $0.02, so that's inconsequential, yet his EA went to town...and drained the demo bank.!!!

While the best setting would be to take full advantage of 0.10 spread, I'll never be able to use it. I'm stuck with 0.5-0.65 spread, minimum, unless we change the coding to remove the "*spread" to use just pure settings for the gridlevel or distance. Having the smallest spread\gaps takes priority up to the point that margin becomes a problem. Their is a way to mathematically figure this out, I just have to ponder it. It is a matter of knowing how far price has to move and open new orders to hit target, for given losers.

Look at go4's Finfx explorer (#50)...It shows on Dec 5th at 4:50pm there was 4 loser buys overcome by 12 sells, so 16 total orders with same spread as on my account, about $0.66 between trades. So we know only 8 of those trades were using margin. That comes to 3 new trades needed per loser, so adding 5 should work. I'm pretty sure the gap between trades is irrelevant for the math. I think it takes the same number of trades to overcome losers, as long as the gap is consistent between orders. With 1:300 margin, a $1000 account can theoretically open 230 non-hedged 0.01 lot trades with the margin available. Margin is a far away problem, IMO.
Too close for pips, switching to points.
  • #358
  • Dec 6, 2014 9:31am Dec 6, 2014 9:31am
Quoting Thepipster
I would like to see how this compares to kosmo's statement.
Hey Pipster,

Please find attached my detailed statement. It looks like I've been too optimistic about the 10 sec cycle closure; from what I can see, it's more like 20-25 sec. I still don't mind as long as it's racking up profit. Surely, you can make more sense of the data, so I'd appreciate your comments.

By the way, Kosmo is a woman, lol.
Attached File(s)
File Type: pdf 0612armada.pdf   666 KB | 624 downloads
mind over matter
  • #359
  • Dec 6, 2014 9:37am Dec 6, 2014 9:37am
kashTrader,

I believe you missed my post (as quoted below). Would appreciate your reply... a simple 'yes' or 'no' will suffice as I know you're very busy with other improvements for the EA. Thank you.

Quoting kosmo
{quote} Hi kashTrader, If I may suggest and if Thepipster is ok with it, would you be able to add a function that will 'cut off' the bot once a set overall equity target is reached? Say, we have a daily target of 5%. A cycle closes => equity is up by min. 5% => the bot shuts itself down => we say 'Thank you, Mr Market'.
mind over matter
  • #360
  • Dec 6, 2014 9:58am Dec 6, 2014 9:58am
Quoting kosmo
{quote} Hey Pipster, Please find attached my detailed statement. It looks like I've been too optimistic about the 10 sec cycle closure; from what I can see, it's more like 20-25 sec. I still don't mind as long as it's racking up profit. Surely, you can make more sense of the data, so I'd appreciate your comments. By the way, Kosmo is a woman, lol. {image}
Your results looks impressive kosmo
0 traders viewing now