Slopey Peaky Bob

[SteveHopwood]

A week of mixed fortunes this week, albeit all good:

• My live trading account: +$200. This means a profit of $1313 since I went live on 10th April.
• Gareth: +$44. The balance is now $775, so the account is recovering well from its nadir at c. $226.
• PJ's Cambodia trip account: +$8. Not exactly a bonanza but the account has still made $476 profit in three months.

I speculated last week that my biggest losing pair might start to recover now that I have the correct number of bars loaded onto my charts - GJ. It looks as though I might be correct:

XAUUSDM30.png

I have CashTakeProfitPerLot set to 3,000 on all three accounts, giving a TP of $90 on my account and $30 on the other two. I am going to try increasing this to 4,000 on my account because of the coding mix-up that canceled my TP. I closed a position manually at +$156 when I noticed that the profit had shot beyond $90.

I am writing an illustrated "Forex Trading - Getting Started" guide. I have taken screen shots at every step; these included starting a new demo platform and opening with a start balance of $500. SPB sent a basket immediately I loaded him. I did not run the demo last week and the position is +$182 because the bot was not running to close the position.

I am starting to wonder whether CashTakeProfitPerLot should be a lot higher. A position can dawdle for days, then take off like an express train and reach TP in minutes. Time will tell.

[SteveHopwood]

There is a new version in post 1. I have adopted Paul's suggestion that we keep the proven V 1t as is and have created an experimental version "E 1u" so we can test new code. From post 1:

Experimental version

• Slopey Peaky Bob.mq4 is stable and its code proven.
• Slopey Peaky Bob Experimental.mq4 contains code that needs testing. These features are described in the user guide but the stable version does not include them. These new features are:
  • Shirt protection.
  • Baluda's request for a basket trailing stop by a percentage of the account balance.

"Shirt protection." was inspired by Gareth and PJ's accounts last week. Their margin level percent had fallen below 400% and 300% respectively. It is a loooooonnnnng way from there to a margin call but both positions were slightly profitable so I closed them manually rather than risk trouble. I probably sacrificed a basket TP by doing so, but us old stagers have learned the value of being conservative. From the updated user guide:

• Shirt Protection: we have a saying in the UK of gamblers who lose everything; we say, “He lost his shirt”. Our equivalent as traders is the dreaded “margin call”, where your broker closes your worst losing trade when ever your margin level drops below a certain point. This is 100% on a Global Prime account with 100:1 leverage. Xxx has two features to help avoid this:
  • DeletePendingsBelowThisMarginLevelPercent: this tells xxx to delete all pending orders once your margin level drops below this figure. This is not quite the same as the FkMinimumMarginPercent:
    • FkMinimumMarginPercent will only delete pending trades if there are no market trades. The rationale is that we want the grid trades to fill and take us to our basket take profit figure as quickly as possible.
    • DeletePendingsBelowThisMarginLevelPercent will delete all pending trades. The rationale is that our margin level is already at a dangerously low level, so we do not want more pending orders becoming market trades and depleting our margin level still more.
  • CloseDeleteAllBelowThisMarginLevelPercent: this protects at least some of the account when our trading has gone disastrously wrong – maybe enough to allow trading to resume. A “margin call” sounds innocent; this is why it is a total disaster:
    • Closing the biggest loosing trade(s) may well take our margin level above the minimum but:
      • Our trading has gone hideously wrong to reach this point. The margin level will usually drop below the minimum level again as the markets continue to go against our trades, so the broker again closes the biggest loosing trade(s).
      • This process will repeat until all trades have closed and there is nothing left in the account. It has “blown”.

It remains to be seen how helpful this will be. All we can do is experiment on demo and see.

Re Paul's request for a basket trail based on the account balance, from the updated UG:

• Basket trailing stop as a percentage of the account balance: note that enabling this automatically disables UseBasketTrailingStopCash:
  • UseBasketTrailingStopPercentage: turns this TS on/off.
  • BasketTrailingStopStartValuePercent: the percentage of the account balance at which to trigger the trail.
  • BasketTrailingStopGapValuePercent: the percentage of the account balance by which to trail the position profit.
  • For example:
    • Account balance = $10,000;
    • BasketTrailingStopStartValuePercent = 0.5%.
    • BasketTrailingStopGapValuePercent = 0.1%.
    • SPB will start trailing when the cash profit on the basket reaches $50.
    • SPB will trail the basket cash profit by $10.

I suggest not including the Experimental version in the Tests Database. Better to wait until the code is tested and then fire up a new database demo.

[fbs1979]

Week 2. Balance 5270. Open trades: None

Settings for the week:
- Size of dollop: 1700
- UseATRforgrid: true
- GridATRdivisor: 3.8
- CashTPperLot: 3334.0 (100 euros, since lotsize will be 0.03)
- Monday start hour: 2*
- Basketfridaycashtarget**: 10,90,12,70,14,50,16,20,18,1
- All the rest is OOTB

* This is something I dont understand. What is the reason to wait until 8am? I can unsderstand avoiding until 1 or 2 am, but after? Can someone enlight me on this?

** I have a doubt here. Targets are absolute or per 0.01lot? I suppose absolute (and so you see my settings), otherwise I need to change them

[pcorneli]

I am awere to take a serious risk with this, but maybe my age can be taken into account.

I have installed SPB and started to test on a test account at GP.

Since i like coding i made a kind of own version of the peaky indi.
Testing this i found that the number of bars as less than the lookback. (1682)
But that pair was also in use with SPB. AUDNZD
I looked up that code and i could not find a test.

The second part of my question is even more stupid (and risky).
Is there a call or function in Empty4 to force a download?

Kind regards and thank for this great environment.
Piet C.

[fbs1979]

Observing a good test for SPB....
With current settings it went (and stays) at almost 10% drawdown. Point is that now most D1SS have turned the opposite of the original direction so it will be interesting to see if (i hope) and how SPB will be able to manage them and get back into green.

Meawhile, I think that for GridATRdivisor 3.8 is still too much. Most grids came out very narrow and the only one that gapped 30 pips (the original setting) was GJ. Most came out at 12, 15, 18 gaps that I feel are too narrow. I will try 3 or little more asap.
I also believe GridATRdivisor should match TP. No sense in having wide grids and small TP, or huge TP and small grids.

[Radar]

Hey Piet,

I am awere to take a serious risk with this, but maybe my age can be taken into account.

I have installed SPB and started to test on a test account at GP.

Since i like coding i made a kind of own version of the peaky indi.
Testing this i found that the number of bars as less than the lookback. (1682)
But that pair was also in use with SPB. AUDNZD
I looked up that code and i could not find a test.

The second part of my question is even more stupid (and risky).
Is there a call or function in Empty4 to force a download?

Kind regards and thank for this great environment.
Piet C.

Age is no excuse for recklessness Enjoying being reckless for the sake of it, (as long as you can cope with the results), is, though
Still, if you're going to change the characteristics of the main indicator that SPB uses, do it on demo for a while first.
There are a few scripts and an EA that will force empty4 to download sufficient history... Check out this one.

Have fun!
Radar =8^)

[Gertje]

empty4

[SteveHopwood]

empty4

Yep. We are heretofore dumping "Empty4" in favour of "empty4".

So here is the burning debate. Should it be, "empty4" or "Empty4"?

I favour, "empty4" as it saves holding down a key for an upper case letter, but will go with the (rarely allowed) democratic consensus.

[tomele]

Germans learn in their English classes that names always start with a capital. Please don't confuse us. We are confused enough by the world already.

I suggest to automatically replace all instances of "empty4", "Empty4", "Empty4" and so on with "Empty4".



Cheers
Thomas

[SteveHopwood]

Germans learn in their English classes that names always start with a capital. Please don't confuse us. We are confused enough by the world already.

I suggest to automatically replace all instances of "empty4", "Empty4", "Empty4" and so on with "Empty4".



Cheers
Thomas

Instantly adopted.

From here on in, all references "Empty4" and "Empty4" MUST be made to, "Empty4".

Failure to accede to this rule will land the poster in hot water.

radar, whatever comes of your contributions to SHF in the future, this one earns you immortality,

I will do a forum-wide replacement of the relevant terms tomorrow, once I have worked out how to do so. Yes, I know I am the forum owner but there are thingies that are hard to work out and yes, I once worked out how to replace, "broker" with, "criminal" but that was a long time ago. You try managing this bloody software. We would not even exist without milanese.