Sixths trading - an EA by macman, Bob and Steve | Page 23 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Trading the Sixths with Snowballs 68 replies

Gann High-Low System with Laguerre/Fisher/VSA/Sixths and others 124 replies

My Sixths trading journal with TOB EA 50 replies

sixths indicator 1 reply

Steve Trading Journal 3 replies

Attachments: Sixths trading - an EA by macman, Bob and Steve
Exit Attachments

Sixths trading - an EA by macman, Bob and Steve

  • #441
  • Oct 23, 2010 12:14pm Oct 23, 2010 12:14pm
Quoting mbkennel
Old_dog:
The issue I have is that TB doesn't always fully take losses completely automatically. (I got a backtestable version going, and it works fine until it gets into a steady trend and then trades stay open for months with ever increasing losses, wiping out all profits and more).
This is my sticking point as well. On my demo account, I've been trading 10 pairs. The first week, when TB allowed for unlimited entries, I got way too many trades on USDJPY. That means for the last couple of weeks, I've only been trading nine pairs, and waiting for a retrace on USDJPY. Right now, the drawdown from that pair is about 23% (which is more than it would be under the current system). But even under the current default rules, it would be sitting on a drawdown for that pair of about 12%, which is about enough to cancel all profits so far.

Hedging might help for folks outside the US. But for US traders, there has to be a method for cutting losses. Otherwise, I'm afraid that this system will look good in the short term, but over time will almost inevitably blow up. The forward tests over a few weeks look great. But people who have been backtesting all seem to agree that it will hit patches that may be unescapable.
  • #442
  • Oct 23, 2010 12:16pm Oct 23, 2010 12:16pm
here is a volatility indicator. incase it is needed.

dont know if it fits the purpose. but i guess no harm no foul
Attached File(s)
File Type: mq4 volatility.mq4   2 KB | 284 downloads
  • #443
  • Oct 23, 2010 12:54pm Oct 23, 2010 12:54pm
duffyprat, Steve, Macman

if you got the time, please take a look at this indicator. this might help the people who dont have the posibility to hedge.

please add it on the 1 hour chart on your open trades that went bad.

for me the EURCHF went bad. it did 3 x a sell. and i am now like -500. since i didnt add the function hedge on true. however i have that possibility since i am from the netherlands.

but for people who live in the USA, this might be a possebility.

as you can see in the picture Laguerra is at top when primary sells happened.

now this indicator wont give you a edge. it only could help you to minimise your losses. so for example when you see a trade is done and languerra goes to bottom and all the way up again crossed the 0.85 line again. close the trade. that way you have a minimum loss.

any other ideas. ofcourse always welcome.
Attached Image (click to enlarge)
Click to Enlarge Name: beast-laguerre-acs1.gif Size: 30 KB
Attached File(s)
File Type: ex4 Laguerre-ACS1.ex4   4 KB | 245 downloads
  • #444
  • Oct 23, 2010 1:02pm Oct 23, 2010 1:02pm
Quoting hiredwhip
Hi Rob,
How long have you been using MTreport4? I picked up a copy of it about 3 weeks ago and I'm just getting the hang of it now....Looks like it could be a help....
Very flexible with the exception of the one computer only rule....I'd like to have it on all my demo units but that doesn't look to be possible.
Be interested to hear how you like it.......

whip
I've been using MTreport4 for about two years - really like it, too.

Rob
  • #445
  • Oct 23, 2010 1:28pm Oct 23, 2010 1:28pm
Quoting duffypratt
I wouldn't draw that sort of conclusion after eight days. Your pairs all have either USD or JPY. Limit your trading to these pairs, and you may have serious trouble when one of those currencies goes strong or weak for a sustained period.

As I understand it, one of the main points of this system is to do some automatic hedging of risk by diversifying the currencies over which it trades. So, the "drag" you noticed results from the design of the system, and over time it should be a good thing. (The same thing that is dragging your winners now,...
Point taken, but this is just a stepping stone for me. The overall mix is quite varied, and I will continue to monitor their performance. I just felt like identifying the strongest pairs during the first eight days. As mentioned in my post, eight of the pairs did well for me, and seven didn't. I understand that things can change.

I've read many times about the importance to "mix-up" the pairs. Based on all of the live accounts I've run, and hundreds of demos, I've often seen correlated DD from non-correlated pairs. For me, execution doesn't always follow theory.

Rob
  • #446
  • Oct 23, 2010 1:36pm Oct 23, 2010 1:36pm
Quoting duffypratt
The forward tests over a few weeks look great. But people who have been backtesting all seem to agree that it will hit patches that may be unescapable.
My experience, exactly. A lot depends on start time and which years are within the test period. I did run into trouble with some of my backtests, but have decided to forward test, anyway. The one thing that backtests can't do is deliver results for more than one pair at a time, so maybe there's hope with multiple pairs.

Rob
  • #447
  • Oct 23, 2010 1:43pm Oct 23, 2010 1:43pm
Quoting Old_Dog
Friends,

I have done a detailed retrace analysis over recent history and plotted the retrace probabilty density function (see pic). As expected, it is approximately Rayleigh distributed.
I'm very interested in this approach, and have a few questions:

1) What counts as recent history, and why does the analysis look back over that time frame and not some other time frame?

2) In producing this data, how did you define a retrace? What defines the starting point, and how do you determine when the retrace ends?

3) The first bar of your chart shows zero retrace. Is there no possibility in this analysis for a negative retrace?

4) Why would one expect a Rayleigh distribution? (I'm not sure that I need an answer to this. In my experience, when drawing curves, statisticians have a strong tendency to see what they expected to see in the first place. That's not necessarily a criticism; just an observation. What you have shown looks very convincing, and that's why I'm interested in learning more.)
  • #448
  • Oct 23, 2010 1:51pm Oct 23, 2010 1:51pm
Quoting ChicagoRob
My experience, exactly. A lot depends on start time and which years are within the test period. I did run into trouble with some of my backtests, but have decided to forward test, anyway. The one thing that backtests can't do is deliver results for more than one pair at a time, so maybe there's hope with multiple pairs.

Rob
Well, the MT4 strategy tester only allows for testing one pair at a time. If you set the same parameters for enough pairs, it would be possible to compile the data for a multiple pair test over a long period of time. I would think that weekly results for each pair would probably be sufficient for comparison, and it shouldn't be too hard to put all of that into a spreadsheet.

Unfortunately, my computer isn't fast enough to do extensive automated backtests on even a single pair without leaving me dead in the water for anything else. I hope to change this problem in the near future. (I'm assuming that even with MT4, using a new i7 computer with 8 gigs of RAM, one could do a five year test on a pair in a reasonable amount of time. Please correct me if I'm wrong about this.)
  • #449
  • Oct 23, 2010 2:13pm Oct 23, 2010 2:13pm
Quoting duffypratt
Well, the MT4 strategy tester only allows for testing one pair at a time. If you set the same parameters for enough pairs, it would be possible to compile the data for a multiple pair test over a long period of time. I would think that weekly results for each pair would probably be sufficient for comparison, and it shouldn't be too hard to put all of that into a spreadsheet.

Unfortunately, my computer isn't fast enough to do extensive automated backtests on even a single pair without leaving me dead in the water for anything else. I hope to change...
Yeah, the tester really chews up resources. I'm running a P4 3.6, top-of-the-line Seagate drives, 4-gigs of RAM, etc., and backtesting still manages to consume over 50% of my CPU. Fact of the matter is that current-day $500 laptops will smoke my rig. I have no doubt that an i7 would be a vast improvement. That's where I want to be for 2011.

I've also noticed huge variances in backtesting speed amongst different EAs. I had one EA that took over an hour to test three years' data. Later, the code was cleaned up and the EA ended up taking 10 minutes to complete the same test.

Rob
  • #450
  • Oct 23, 2010 2:14pm Oct 23, 2010 2:14pm
My results for last week were, opening balance £2808, closing balance £4654, closing equity £3597, criminal was Alpari UK demo trading 13 pairs taking primary trades only at 0.3 lot size, everything else default.
For those that may not already know this, if you are unhappy at downloading your platform from your criminals website, you can use the generic version of MT4 downloaded from metaquotes.net. You just have to copy the .srv files from the config folder on your criminals platform to the config folder on your generic MT4 platform.
  • #451
  • Oct 24, 2010 6:02am Oct 24, 2010 6:02am
The latest update is in post 1.


I have added CanucCT's take profit line. From the user guide:


Take profit line
When TB send a Primary pending trade, it adds a gray take profit line just a few pips away from the middle white line. Whilst this line is gray, TB will ignore it. If you select the line by double-clicking it, you will see it change colour to your choice in the inputs (default blue). TB will attempt to move the tp of all its open trades to the price of the tp line. It will ignore the line if in Recovery or Hedging is in operation. To deactivate the line, select and delete it.


For now, every use of the word 'will' ought to read 'should' as there is no quarantee the code is watertight.
  • #452
  • Oct 24, 2010 7:32am Oct 24, 2010 7:32am
Nice work...will give this a workout when markets open.
Cheers
CanuckCT

Quoting SteveHopwood
The latest update is in post 1.


I have added CanucCT's take profit line. From the user guide:


Take profit line...
  • #453
  • Edited 12:51pm Oct 24, 2010 11:55am | Edited 12:51pm
The name of the attached says it all. By default, trading is turned off so it is safe to put it in charts to play with.

I have started to think about how to make TB do some time-scale self-optimisation. A couple of contributors have sent pm's: rfking with an offer to help me work out how to create BarCount self-optimisation; Old_Dog with a potential methodology.

I will need to have a lot of time on my hands, and be feeling extremely strong before I embark on any of that,so i have come up with an interim measure for people to play with.

My premise is this; recent market movement is more relevant to us than distant market movement. I have added functionality that allows us to specify how far back in time TB is to go to find the hi-lo it is going to use for calculating the sixths lines.

You will see the inputs at the top of the list; their use should be obvious. Each one starts looking for the hi-lo starting with the previous candle.

For example, suppose I feel that the most recent 10 days are most relevant to my trading. I would set UseDailyCandles to true and NoOfDays to 10. This is not the same as using the weekly inputs, as these functions start by looking at the previous candle. On Wednesday, a 10 day daily candle search will start looking at the Tuesday market, whereas the weekly function would start a week ago Monday.

I am not sure if I have made this clear. Have a play and see what you think.

Attached File(s)
File Type: mq4 The Beast Experimental.mq4   75 KB | 253 downloads
  • #454
  • Oct 24, 2010 12:11pm Oct 24, 2010 12:11pm
I have followed TB a week on a RENKO setup as you suggested, but did not have any results with it.

Will try in a later stage.

For the time being, will stick to testing this on a TF based chart till I get a better understanding of it all.

Only SO much the old head can handle
  • #455
  • Oct 24, 2010 12:46pm Oct 24, 2010 12:46pm
Quoting Freddyfx
Only SO much the old head can handle
I know how you feel.

I just ran the experimental version through backtester and realised that I had forgotten to add a bit of code to tell it not to modify the tp when the tp line is grey.

If you downloaded the experimental version I posted a couple of posts ago, before reading this then you need to download.

The same applies to those of you who downloaded the version in post 1.

  • #456
  • Oct 24, 2010 1:14pm Oct 24, 2010 1:14pm
Will do, just came back from the pool, the sun even hurts on the eyes, but I rather experience that then rain all the time.
  • #457
  • Oct 24, 2010 3:42pm Oct 24, 2010 3:42pm
As this is my first post on the thread, I may have to apologise, in advance if I have failed to grasp what is necessary to run this EA.

I have read every post but, at my age, because I have read them all doesn't mean that I have remembered them all!!!

Am I correct in believing that the EA can be run OOTB, using the settings as laid out in .pdf document without the need to add MPTM or 6ths indicator as, they have been built into the latest version of the EA?
  • #458
  • Oct 24, 2010 4:00pm Oct 24, 2010 4:00pm
Quoting Alan5616
As this is my first post on the thread, I may have to apologise, in advance if I have failed to grasp what is necessary to run this EA.

I have read every post but, at my age, because I have read them all doesn't mean that I have remembered them all!!!

Am I correct in believing that the EA can be run OOTB, using the settings as laid out in .pdf document without the need to add MPTM or 6ths indicator as, they have been built into the latest version of the EA?
Correct, you will need to adjust the primary and secondary lot size per your trading acct size, also the the last version before the experimental version had hedging turned on by default (I believe), if you want to hedge leave at TRUE, is you are in US and hedging not possible, you will need to turn to FALSE. Also I belive it is on page 7 of this thread, you will find a list of trading pairs and what the bar count should be adjusted to on each trading pair. Cant think of anything else I changed from OTB.
  • #459
  • Oct 24, 2010 4:48pm Oct 24, 2010 4:48pm
In an astonishing piece of bloop-spotting, sq found a bug relating to the break even line when hedging has kicked in.

This is not a major issue until there are some hedged trades, so no desperate rush to download the fix. Just remember this is a prob when hedging kicks in.

The fix is in post 1. It is also applied to the attached experimental version.

Cheers sq. You are a star. I hope I mention this from time to time.

Attached File(s)
File Type: mq4 The Beast Experimental.mq4   76 KB | 310 downloads
  • #460
  • Oct 24, 2010 4:51pm Oct 24, 2010 4:51pm
Quoting pipstalker
Correct, you will need to adjust the primary and secondary lot size per your trading acct size, also the the last version before the experimental version had hedging turned on by default (I believe), if you want to hedge leave at TRUE, is you are in US and hedging not possible, you will need to turn to FALSE. Also I belive it is on page 7 of this thread, you will find a list of trading pairs and what the bar count should be adjusted to on each trading pair. Cant think of anything else I changed from OTB.
Thank you very much for your help, Pipstalker. I live and have my account in the UK so, fortunately, I'm not affected by the NFA's bizarre hedging ban.
0 traders viewing now