Griddy Wotsit Mk 2

[Gertje]

Hi,

I tried a lot of different things, but can't seem to get it to work.

Steve, when trade #4 triggers, a completely new grid with buys and sells is added.
Is it al lot of work to change this to;
- add only buys after buy #4 is triggered, and
- add only sells after sell #4 is triggered?

Don't waste any time over it, 'just trying to figure something out.

[SteveHopwood]

Hi,

I tried a lot of different things, but can't seem to get it to work.

Steve, when trade #4 triggers, a completely new grid with buys and sells is added.
Is it al lot of work to change this to;
- add only buys after buy #4 is triggered, and
- add only sells after sell #4 is triggered?

Don't waste any time over it, 'just trying to figure something out.

No time for details as I have to go out but the attached should do the trick. Set ReplaceOppositeGridAtL4 to false.

[SteveHopwood]

Gertje, did the update work as required?

The reason Gerteje asked for the update folks is the way the bot will behave in a bouncing market.

In a cooperative market, the trades might become hedged, but a half-way decent market movement will close the hedged position at a profit.

The problem comes when the market constantly bounces between the L4 trades - as GJ has done for a couple of days.

When this happens, then grids are sent that fill, then are resent at L4 and fill, then are resent and fill...............

We end up with umpteen trades open within a few pips of each other, over and over again.

I put GW on my live trading account, GJ. Frankly, there is only one way to test an EA or trading system, and people who think that is on demo are deluding themselves.

I ended up with about 150 trades split near enough 50/50 buy/sell. The DD was the commission, spread and swap. I closed the position tonight. No problem with the small loss but no need to continue to accrue it.

So, a waste of time and money to continue with the EA as originally coded. Forget it. It is bust. Dead. Only an idiot etc

Gertje's suggestion will do away with the bouncing market. Whether this will also do away with the possibility that hedged baskets will close quickly is something that will only show up with experimentation.

Oh, and for those of you with tiny accounts, I set up one to trade a $500 start balance. Current: 158 open trades; DD -$418: Margin level 172%. 'nough said?

[Gertje]

Hi Steve,

I was suprised by the speed of your update, and tried it immediately on EURUSD.

I was in time to catch some movement, enough to trigger a second grid, which was entered like I asked, only above the winning grid.

Time to try some settings next week.

Thanks a lot, and have a good weekend!

[SteveHopwood]

I have made the version I coded for Gertje the post 1 version. You do not need to re-download if you already have the one a few posts up from here.

[Secundus]

Greetingsl:

I have been demo-testing the amazing Griddy Wotsit, with results ranging from first class, to shock and horror. It looks as if Griddy has lots of promise when selectively applied to trending markets.

To test this , it would be helpful if the EA could start a grid optionally from either current iHigh(NULL,0,1) (Buy price) or current iLow(NULL,0,1) (SELL price), as appropriate. It might be even more helpful if the user could at his option, require a BUY grid start only if preceeded by an up (Green) Bar, with opposite (Red Bar)requirements for SELL Grids.

What do you think? Is it feasible tp add such options to the Wotsit , for test purposes?
Thanks for your consideration.

Chris.

[PhP71]

Hi Steve, Gertje,

thanks a lot for this Griddy Wotsit,
I tried it (on demo) and had the same issue of hedge grids filling up to 118 positions on GU leading to a DD of 180€ with a lot size of 0.01.

I decided so to test this modified version thinking it could avoid this kind of situation but can bring another one not more enviable where the price could be doing flip-flop in the hedge grid between buy and sell orders for a quite long time not going far enough to close the opened positions.

Anyway, I'm keeping on testing it with "ReplaceOppositeGridAtL4 = false" and I'm seeing this situation where the price goes mainly in one direction filling up only sells trades and opening a new grid of sell stop orders 3 times in a raw but without moving the buy stop orders which lead to have a quite big space betwwen buy and sell stop orders.

a picture is also good to understand.
May be I'm wrong and I missed something but it doesn't look fine to me.

Capture01.png

[SteveHopwood]

Gerteje's idea has been a success so far in that it has avoided those amazingly clogged positions we were arriving at with the previous version.

Headline figure was a stress-free 10% balance increase. The fly in the ointment is the left over position of 12 buys and 6 sells. A similar position sorted itself out last week, so it will be interesting to see what happens to this one.

[PhP71]

Here is what happen next:

Capture02a.png

Now it have entered into a hedge trading with 5 buys and 3 sells and a space zone of 140 pips between sells and buys, it will have to travel quite far to retrieve the actual 840 pips loss and add 30 pips profit (default value in the EA).

This is the situation I wanted to explain in my previous post and it happened sooner than I hoped.

Could we imagine a change in the trading mode in the way that before entering a hedge trading, the sells stop follow the buys trades as they close and reciproquely the buys stop follow the sells trades.

Lets take an example, we start with a fresh new grid, no trades yet:

- price move up and trigger a buy
- price continu to move up and trigger a second buy
- a spread later, the first buy is closed

At this point the space between the sells and buys lines is not 20 pips (with the default settings) anymore but 30 pips (we closed a buy trade) and this can continue to the situation we face actually.

My idea, in this example, is the sells stop should shift 10 pips up as soon as the first buy trade is closed.
and again 10 pips up after the second buy close and so on until the price travel back to trigger a sell trade and enter a hedge trading.

This is valid also on the opposite if the price goes down and trigger sells, the buys stop will follow down.

This way, the space between the sells and buys will never be more than twice 10 pips (default settings).

What's your mind on this idea?

Phil.

[SteveHopwood]

I have been contemplating the situation that Phil describes for some time. Your post finally booted me into life today Phil, so thanks.

I have added a new input to the grid inputs section - MaxDistanceBetweenGroupsPips.

What I intend to happen is this: if the distance between the lowest market buy price and highest market sell price exceeds MaxDistanceBetweenGroupsPips, GW should delete all outstanding pending trades and resend a new grid of buy and sell stops. This should have the effect of stopping too big a gap emerging between two groups of market trades.

This will not work ideally with a position where the market is in between two groups of market trades, so I have restarted the demo.

The code may need refining. It may even have a bug or two to squish. I have run a quick-and-dirty M1 open prices only craptest and the run stalls when the ea is trying to close an order that doesn't exist. No doubt problems will surface quickly.

Coders, you can have a look at what I have done by going to line 1224.

V 1b in post 1.