Gridtrading from daily open

[SteveHopwood]

Evening traders,

Thanks Steve for the updates.

An update on my live account:
GJ has hits its 100pip every day this week
EJ and GN trades are still alive from Monday and in a combined DD of ~700pips with a max of 1000 earlier today.

My demo on USDSEK exotic has smashed in +700 today alone but that might be expected as its so volatile. Don't worry, I won't be going live on this without demoing for a very long time first; just an interesting experiment at this stage.

A quick glance through my hindsight-ometre
Ranging currencies are a killer in terms of DD for this so it strikes me that a quick glance at a CSS chart might help in optimising performance. For those who don't know what it is have a look in the HGI forum.

As you can see from the beginning of the week it says EUR and JPY are both parallel heading south (weakening) = ranging
GBP and NZD are both heading up parallel (and at the 0 line) which means both are strengthening = ranging.
Therefore Grannies and eggs ;
GBP strengthening, JPY weakening... GJ has given me the pips.

I chose the currencies that I did by having a look at their trending and ATR over the previous week without looking at the bigger picture. If I had glanced at CSS, which is an amazing tool, I can't help but think I might have chosen different pairs.
So next week I will experiment with manually adding CSS into the GW mix next week.

Please do. It is not difficult to port the business parts of GW into my shell that includes CSS. I am not going to do so straight away, but will do with pleasure if you can make a case for it.

[Gertje]

Look what I woke up to.......

Greedy Wotsit.png

Doubled up on the pairs that stayed open over rollover, some closed already.
F***ed up on GBPNZD, is now perfectly hedged in huge DD.
If I take that DD of off the global DD, it seems reasonable for such a return.

Now I'm wondering if I should open pairs tonight or try to let them all close before the weekend.
On the other hand, then we now what a gap will do to a open basket.

[Louisetano]

Look what I woke up to.......


Doubled up on the pairs that stayed open over rollover, some closed already.
F***ed up on GBPNZD, is now perfectly hedged in huge DD.
If I take that DD of off the global DD, it seems reasonable for such a return.

Now I'm wondering if I should open pairs tonight or try to let them all close before the weekend.
On the other hand, then we now what a gap will do to a open basket.

Hi Gertje,

What did you do with GN? Did you allow it to open new grids everyday but with a different magic number? I was thinking we could learn a lot from how it behaved when it didn't close as we wanted it to.

I only had EJ that didn't close yesterday and I didn't allow it to open new trades today. I never had a very big DD, but I'm kind of hoping for a very bad day, to see how bad it gets . Maybe your GN trade could be an example of this??

So far, your idea is excellent, and I think it's all about not being greedy, stay with the best performing pairs, and keep trading small lots in case of those bad days, that I'm yet to experience

[Gertje]

Hi Gertje,

What did you do with GN? Did you allow it to open new grids everyday but with a different magic number? I was thinking we could learn a lot from how it behaved when it didn't close as we wanted it to.

Last thursday the basket didn't close, so I added an extra chart with a different magic# on friday.
This didn't close either, and I went through the weekend with 2 baskets open on the pair.

Mondaymorning I didn't add a 3th basket, it would have closed though.
Tuesdaymorning I didn't add a 3th basket, it would have closed.
Wednesmorning I didn't add a 3th basket, it would have closed.
Thursdaymorning I DID add a 3th basket, and it closed.

The 2 old ones I added a new grid on top of the existing ones, and a lot of them filled to hedge the existing ones. Doing so, locking in the DD in the hedge.
One basket is 15/14 and the other is 12/12.
If my math is right, I need ((losers x 2) +3) to close the basket so "just' 300-400 pips in one direction...

I learned doing this, to not tinkle with baskets and let them play out on themselves.
Maybe even adding a 2nd basket with a different magic is wrong, although it can pay to do so.
.

[Louisetano]

Last thursday the basket didn't close, so I added an extra chart with a different magic# on friday.
This didn't close either, and I went through the weekend with 2 baskets open on the pair.

Mondaymorning I didn't add a 3th basket, it would have closed though.
Tuesdaymorning I didn't add a 3th basket, it would have closed.
Wednesmorning I didn't add a 3th basket, it would have closed.
Thursdaymorning I DID add a 3th basket, and it closed.

The 2 old ones I added a new grid on top of the existing ones, and a lot of them filled to hedge the existing ones. Doing so, locking in the DD in the hedge.
One basket is 15/14 and the other is 12/12.
If my math is right, I need ((losers x 2) +3) to close the basket so "just' 300-400 pips in one direction...

I learned doing this, to not tinkle with baskets and let them play out on themselves.
Maybe even adding a 2nd basket with a different magic is wrong, although it can pay to do so.
.

That's valuable information though, thanks for that :hi:

[Gertje]

Scanning through the closed baskets from last night.
It appears that Greedy Wotsit didn't only have it's name wel choosen, but is fast on the trigger too!

In 2 seconds [let this sink in!] the Greedy Bugger;
- opened 6 pending trades
- closed 7 trades
- added over 300 pips

Greedy Wotsit.png

[Bruster400]

Hi Guys,

I've been playing with this on demo and trying to find the best sets of variables to produce the most likely chance of hitting TP with the least DD. (wouldn't we all like to know these!!) The number of variables is just too great to be able to test effectively in demo - it's just too slow - so I decided to try to do it mathematically.

I've haven't found the optimum but I'm a lot closer!

Take a look at the attached spreadsheet. It's a tool I've pulled together that allows you to look at the effect of changing the variables and how that would play out on the grid.

I've used the following user inputs: ADR, Distance between trades and profit target % of ADR. I haven't looked at gridsize because I'm assuming that is infinite as the bot adds more if it needs it.

Using these values the sheet then calculates the move needed to hit profit without hedged trades and the move needed to hit BE and profit if you're dragging some hedged trades. The key seems to be the distance between trades. Small changes to this make a massive difference in the structure of the trade. I've looked at the %ADR that price will need to move in order to hit profit. This allows you to check to make sure you're not looking for an unrealistic price move in order to hit profit. I've also looked at the BE point and the worst case BE.

NOTE: All calculations are done at grid levels only. So if it says you have to hit level 5 to hit profit it means that the profit will be greater than your target by level 5. That's why I've said "To Exceed Profit". I've also shown the actual profit at this level in pips too so you can gauge how far past your profit point price has gone. This is also true of the DD calc. This is taken from the worst DD at a grid level. It might be slightly higher on one side of that grid level but it will be lower at the two adjacent levels.

Hope this helps - please shout if you spot any errors!

EDIT: File updated to show grid levels where profit or BE is hit exactly on the level. (excluding spread and slippage!!). AND updated pip distance!!

Grid Trading Calculations.xlsx

[Gertje]

Hi Bruster,

I took a quick glance at the spreadsheet and noticed you didn't take into account the distance from the setting of the grid and the first trade for the unhedged part.

Not sure, but the same might be true with the hedged part, the distance between the 1st sell and the 1st buy is 2*griddistance.

Looks good, will play with it some more.

[tonyharmony]

Hi Guys,

I've been playing with this on demo and trying to find the best sets of variables to produce the most likely chance of hitting TP with the least DD. (wouldn't we all like to know these!!) The number of variables is just too great to be able to test effectively in demo - it's just too slow - so I decided to try to do it mathematically.

I've haven't found the optimum but I'm a lot closer!

Take a look at the attached spreadsheet. It's a tool I've pulled together that allows you to look at the effect of changing the variables and how that would play out on the grid.

I've used the following user inputs: ADR, Distance between trades and profit target % of ADR. I haven't looked at gridsize because I'm assuming that is infinite as the bot adds more if it needs it.

Using these values the sheet then calculates the move needed to hit profit without hedged trades and the move needed to hit BE and profit if you're dragging some hedged trades. The key seems to be the distance between trades. Small changes to this make a massive difference in the structure of the trade. I've looked at the %ADR that price will need to move in order to hit profit. This allows you to check to make sure you're not looking for an unrealistic price move in order to hit profit. I've also looked at the BE point and the worst case BE.

NOTE: All calculations are done at grid levels only. So if it says you have to hit level 5 to hit profit it means that the profit will be greater than your target by level 5. That's why I've said "To Exceed Profit". I've also shown the actual profit at this level in pips too so you can gauge how far past your profit point price has gone. This is also true of the DD calc. This is taken from the worst DD at a grid level. It might be slightly higher on one side of that grid level but it will be lower at the two adjacent levels.

Hope this helps - please shout if you spot any errors!

Thanks Bruster, that is very kind of you to take the time.. :hi:

Cheers

[SteveHopwood]

Wonderful stuff Bruster.