Griddy Wotsit Mk 2

Post by **keydcuk** » Sat Mar 12, 2016 11:35 am

Sorry Steve, my comments where not meant to offend.

Post by **SteveHopwood** » Sat Mar 12, 2016 1:03 pm

I imagine most of you are also reading the HGBnG thread. Pop over there now if you have not already seen my weekly results post - it shows my live trading GW results.

I consider my live trading experiment on GJ to be a failure. Grid size: 10. TP 3: Distance between trades: 3. Hedge tp 2 pips per trade, unlimited.

Friday neatly sums up the problem. I returned home from teaching to about 450 open trades, skewed in favour of the buys - so looking for a basket tp of c. 900 pips.

I remarked earlier that the upl changes rapidly when there are umpteen trades open and I had forgotten to raise my cash tp input from $100. A half-way decent climb took the profit to that figure and all the trades closed at an overall loss of about $165. I will add a must-be-in-profit stipulation in time for next week.

450+ trades take a long time to close folks. My ea's close market trades first these days, but they must have taken about 3 minutes and the market was ticking all the time. A tick makes a hell of a difference when there are that many trades being closed.

All the time, two of my buy stops were showing green. I did not want to close them manually in case this upset the bot. They were amongst the last trades to be deleted, but they came close to being triggered. Doesn't matter a hoot trading 0.01 lots, but a bit of a bugger trading 1 lot sizes on a 100,000 account.

That doesn't work by the way - 1 lot sizes on a 100,000 account. That 104,000 balance shrunk to 97,000 after one of those umpteen trades events either caused a margin call or the trades were closed at the cash tp that was actually in loss.

I started an experiment on Thursday. Start balance $10,000. 0.01 lots on the four biggies + GJ.

GW.png

Fascinating to see how GJ outperformed the others by a long, long way.

Less encouraging is this little lot:

Less encouraging.png

Not entirely sure I would want to enter the weekend with that little lot still open. We shall see. The experiment continues.

Post by **TraderJoeForex** » Sat Mar 12, 2016 5:25 pm

SteveHopwood » Sat Mar 12, 2016 1:03 pm wrote:
450+ trades take a long time to close folks. My ea's close market trades first these days, but they must have taken about 3 minutes and the market was ticking all the time. A tick makes a hell of a difference when there are that many trades being closed.

Hi Steve, quick question. I am right in assuming that the EA will close the positions in profit before the positions in loss when a hedged basket closes or does it alternate one profit one loss etc.. (just thinking from a margin point of view as a big hedged basket is closed).

Post by **SteveHopwood** » Sat Mar 12, 2016 5:43 pm

TraderJoeForex » Sat Mar 12, 2016 5:25 pm wrote:
SteveHopwood » Sat Mar 12, 2016 1:03 pm wrote:
450+ trades take a long time to close folks. My ea's close market trades first these days, but they must have taken about 3 minutes and the market was ticking all the time. A tick makes a hell of a difference when there are that many trades being closed.

Hi Steve, quick question. I am right in assuming that the EA will close the positions in profit before the positions in loss when a hedged basket closes or does it alternate one profit one loss etc.. (just thinking from a margin point of view as a big hedged basket is closed).

It closes them in the order they were opened - FIFO.

Post by **TraderJoeForex** » Sat Mar 12, 2016 5:49 pm

SteveHopwood » Sat Mar 12, 2016 5:43 pm wrote: It closes them in the order they were opened - FIFO.

ahh yes of course. Thanks

Post by **keydcuk** » Mon Mar 14, 2016 11:16 am

Hi Steve,

At the risk of pissing you off, I am still going to post what my point was on GW from a previous post.

I am deeply sorry that you didn't understand the first time I tried explaining something earlier with the 10 pip grid example and maybe I was a little all over the place. I am not the best at explaining myself.

Here is a live example of a scenario that has happened overnight.

In my 10 pip grid, I have had 7 x 10 pip winners on the sell side without having to go in a hedge position.

The 8th sell has triggered. Because it has not been in a hedge, nothing of the grid has reset itself. This scenario means that should the PA reverse all the way up to the buy side, it will carry the sell position nearly 100 pips in loss before it hits the buy on the grid and starts a hedge.

I was only trying to suggest that if positions go to TP and do not go in a hedge that the grid resets itself.

PS: I am not trying to be like a dimwit, or a newbie that doesn't bother to learn. I am simply one of the traders that is genuinely trying to help contribute to all the hard efforts that is being done on this site.

Sorry in advance if I have pissed you off. I hope by the screenshot example that you can notice what I am trying to show in order to try to make improvements to GW.

Sincerely,

Daniel

Post by **Gertje** » Mon Mar 14, 2016 12:39 pm

Daniël,

There is an easy fix for this, make the grid smaller, or set the MaxDistanceBetweenGroups to a smaller value.
The downside will be that you would be in a hedge already.

Post by **TraderJoeForex** » Mon Mar 14, 2016 2:25 pm

Interestingly enough I have been playing with the "MaxDistanceBetweenGroups" today and it appears that keeping this figure the same as the "DistanceBetweenTradePips" lowers the draw down - net profit ratio to almost exactly 1-1 and any other settings go the wrong way.

BIG PIPS to the 1st person that posts a set offering lower draw down - higher net profit proven during high impact news this week (DOUBLE BIG PIPS if it's proven on a live account)

Post by **Gertje** » Mon Mar 14, 2016 3:02 pm

TraderJoeForex » Mon Mar 14, 2016 2:25 pm wrote:Interestingly enough I have been playing with the "MaxDistanceBetweenGroups" today and it appears that keeping this figure the same as the "DistanceBetweenTradePips" lowers the draw down - net profit ratio to almost exactly 1-1 and any other settings go the wrong way.

BIG PIPS to the 1st person that posts a set offering lower draw down - higher net profit proven during high impact news this week (DOUBLE BIG PIPS if it's proven on a live account)

Interesting, do you have an example?
I'm running 5pip grids at the moment, with 5 levels.
MaxDistance set to 50, so the first new grid is one way, and the second two ways as usuall.

Post by **TraderJoeForex** » Mon Mar 14, 2016 3:08 pm

Gertje » Mon Mar 14, 2016 3:02 pm wrote: Interesting, do you have an example?
I'm running 5pip grids at the moment, with 5 levels.
MaxDistance set to 50, so the first new grid is one way, and the second two ways as usuall.

yes...... in T - 12 hours

your GJ set but "MaxdistanceBetweenGroups" set to 5 instead of 30 and unlimited hedge profits pips/cash.

This week I am trading GW live to place new grids from 15mins immediately before high impact news .. and perhaps this figure needs to be less.. we will see. (Fixed spread account).