Gridtrading from daily open

[SteveHopwood]

I am putting a stop to some nonsense here tonight.

moebiusx, shut up and work things out for yourself or I will remove your ability to post at all. Dimwits are not welcome at SHF.

[Anima999]

Day one live results
GJ closed for +100pips at about 1300GMT
EJ + GN still open and about -200 in DD each.

I chose EJ and GN on the basis of their ATR (about 200ish) and I didn't want all 3 pairs with yen or all 3 with Gbp. Their charts had them reasonable trends so I think today was consolidation day after some hefty moves last week. I'm more than happy to sit on the DD as they will have to move eventually.

Note: I'm trading Griddy Whatsit live as that's what I demoed. I have Greedy Whatsit running on a demo as well so will upgrade next week.

[Gertje]

Tuesday 11h30 local time.

12 baskets closed for 690 pips. 15 pairs remaining.

GBPNZD is the main DD-cause, both last week thursday and friday the baskets didn't close.
Would I have opened new baskets yesterday and today, they would have closed for 100 pips each.

Greedy Wotsit.png

[Louisetano]

Great results Gertje

Yes GN is also the trouble maker at my end. Think it just moves too much up and down, whereas GA is a sure easy winner everyday.

[patman]

Hi,

First off thanks Steve for coding & Gertje for sharing this.

I put it on some charts last night. Most opened and shut the grids without problems.
GA had a hiccup
There were multiple

Greedy Wotsit GBPAUD,H1: GBPAUD Greedy Wotsit OP_SELLSTOP order send failed with error(130): invalid stops

I think this happened because the price moved down past where the sellstop wanted to be before the order could be placed
(I eventually realised this was my error - I was using ATR and put a 0.1 multiplier resulting in 2 pip spacing rather than 20pip)

In the meantime I found this on line 737

Code: Select all

   IsGlobalPrimeOrECNCriminal = false;

IsGlobalPrimeOrECNCriminal is not set back to true at any point.

Is this meant to be or a leftover from debugging?

Regards & Thanks
Pat

[BobbyT]

Hi,

First off thanks Steve for coding & Gertje for sharing this.

I put it on some charts last night. Most opened and shut the grids without problems.
GA had a hiccup
There were multiple

I think this happened because the price moved down past where the sellstop wanted to be before the order could be placed
(I eventually realised this was my error - I was using ATR and put a 0.1 multiplier resulting in 2 pip spacing rather than 20pip)

In the meantime I found this on line 737

Code: Select all

   IsGlobalPrimeOrECNCriminal = false;

IsGlobalPrimeOrECNCriminal is not set back to true at any point.

Is this meant to be or a leftover from debugging?

Regards & Thanks
Pat

Patman,

From what I recall, this function is set to false by default to protect some of us cretins. It will still check to see if your account is with GP anyway. This is in some documentation somewhere. Maybe check some of the earlier HGB bot manuals as I think that's where it popped up.

On a separate note, I also had a glitch with GA. Not so much an EA issue but more a Empty4 issue I think. When throwing ol' greedy up on some charts on Monday morning (upside down time) GA opened a huge grid all at the same price at the same sell and buy price.

After checking the helpful on screen feedback it seems for some reason or another the distance between trades didn't calculate. Anyway, a quick scipt-based order close and EA re-load all was back to normal though it put the account 230 pips in the red.

Not that it really matters. The EA is now 780 in the green Just thought I would put it out there for anyone else who may have noticed something similar.

This is by far the most insane thing I've ever seen. Lets see if it can keep it up

[passgo]

Not to say this EA should be changed, but on the subject of drawdowns in the Grid, here is a possible solution for minimizing the DD when ranging, while maintaining the upside once price decides to start trending out of the Grid.

I expect many of us have been watching the recovery from the DD after Sunday's opening. It is very heartening to watch the recovery.

One issue I had is that in comments alongside "GW" was added "(starting gap)"- this seemed to prevent Greedy managing that one position, so I closed manually and added the pips to the target. Is it possible to manage trades solely on Magic Number by leaving the Comment field blank?

I can see the logic of kayaktri's proposal to reduce DD. However, is there a problem if the reversal then reverses again?

As other people have mentioned, some of the large DD I experienced at the beginning of this week was down to trading the same currency across more than one pair. Is their potential to adopt a staged approach i.e. using GBPJPY as an example:

1. The first JPY basket opens on GBPJPY
2. Does not achieve breakeven so Greedy does not open any more JPY pairs
3. The first trade achieves breakeven + ? pips
4. Greedy now allows all other JPY trades to open

As usual with my suggestions, hope it is a constructive idea - if not consign to that proverbial waste paper basket!

[Gertje]

I expect many of us have been watching the recovery from the DD after Sunday's opening. It is very heartening to watch the recovery.

Recovery from what? The EA was distributed during the weekend?

One issue I had is that in comments alongside "GW" was added "(starting gap)"- this seemed to prevent Greedy managing that one position, so I closed manually and added the pips to the target. Is it possible to manage trades solely on Magic Number by leaving the Comment field blank?

To my knowledge, GW doesn't manage using the commentfield. It manages the pair it's put on, and if entered with the magic# in the EA.

I can see the logic of kayaktri's proposal to reduce DD. However, is there a problem if the reversal then reverses again?

I don't see the logic.
All it does is lock in the DD. Lower DD, but locked for sure.
The now HUGE loss of the single biggest loser has to be made up by less winners -> it takes a much bigger swing to close out.

As other people have mentioned, some of the large DD I experienced at the beginning of this week was down to trading the same currency across more than one pair. Is their potential to adopt a staged approach i.e. using GBPJPY as an example:

1. The first JPY basket opens on GBPJPY
2. Does not achieve breakeven so Greedy does not open any more JPY pairs
3. The first trade achieves breakeven + ? pips
4. Greedy now allows all other JPY trades to open

Either you trade multiple crosses or you don't. GW works with pendings so it's almost inevitable that more trades on the same currency have opened before one is at BE.

What I believe might relieve some of the DD is to take a look after a few weeks demo to see what pairs move the most during what time of day and base the openingtime on that.

JPY will move during Asian,
GBP will move during London
CAD will move during New York
Etc.

This way some pairs will enter after several baskets have closed already and reduce DD/ margin requirements.
.

[patman]

From what I recall, this function is set to false by default to protect some of us cretins. It will still check to see if your account is with GP anyway. This is in some documentation somewhere. Maybe check some of the earlier HGB bot manuals as I think that's where it popped up.

Hi BobbyT,

Thanks, for pointing me in the right direction
RTFM and study the code....

OK - I've read some more and figured it out (somewhat).

The auto detection you mention happens in the OnInit section.
Once an order is placed IsGlobalPrimeOrECNCriminal is set to false and stays that way until the bot is started again.

However I think it doesn't matter with GW as SL and TP are 0
So lets not worry about it - and move on

Pat

[kayaktri]

DD doesn't really bother me, margin does...

With 1 trade against you, you need 5 winners to close. There will be 4 trades to reach TP.
With 2 trades against you, you need 7 winners to close. There will be 5 trades to reach TP.
With 3 trades against you, you need 9 winners to close. There will be 6 trades to reach TP.
For every trade against you, you need 2 winners + the 3 original winners to reach BE and TP.

Gertje,

I may have not clearly stated my hypothesis in the post above, but my intention was to provide an example where one could have a mechanism to reduce the number of winning trades required to close out a basket in times where price was ranging. As with the current EA, no losses will be taken prior to the basket closing. I am sure there are much smarter people here that can come up with an even better solution than I in this endeavor.

Reading your quote above, I understand you are not bothered by DD, but are bothered with margin. If DD is kept small, margin issues are less likely.

In looking at your example #3 in the quote above, the reason why it requires 9 winners to close the basket is because you are 'dragging' 3 losers against you when you finally head in the direction of the trend. In my example, you would only be 'dragging' one loser against you, hence it would take fewer winners to close out your basket. The current EA holds onto the DD until the basket is closed, and my example does the same. The only difference would be how many losing trades you carry in the basket.

Thank you for your consideration, and I will let this go unless further interest arises by you. I appreciate your contribution.