Journal related to: Spud's MTF Stochastics | Page 6 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Spud's MTF FIB Breakout System 383 replies

MTF Stochastics 1,485 replies

Spud's Predictive MTF Stochastics 1,696 replies

MTF Stochastics into MTF Stoch Histogram 8 replies

MTF Stochastics Trade Journal 12 replies

Attachments: Journal related to: Spud's MTF Stochastics
Exit Attachments

Journal related to: Spud's MTF Stochastics

  • #101
  • Feb 26, 2009 9:51pm Feb 26, 2009 9:51pm
New High: http://www.forexfactory.com/images/icons/icon14.gif

TOTAL TRADING PERIOD: START 28.January 2009
START Balance: 100000.00
END Balance: 105497.54


only small max DD till now < 2% -- never turned below initial deposit.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot.png Size: 10 KB
__Thanks__ MJ
  • #102
  • Feb 27, 2009 1:53am Feb 27, 2009 1:53am
Quoting M_j
Well, I had an idea how to improve some parts of my base automatic trading system.
I did implement that part already --hope not too many bugs

Only negative side effect: the configuration of each bot has become much more complicated (also more prone too errors) - which is not nice - but guess one can not have all at the same time.

Don't expect any changes on the trading results - as that was a more base automation part which is effected.
__Thanks__ MJ
  • #103
  • Feb 27, 2009 2:43am Feb 27, 2009 2:43am
For next week I will start looking at this 10 pairs:

"EUR/USD",
"USD/JPY",
"GBP/USD",
"EUR/GBP",
"AUD/USD",
"EUR/JPY",
"USD/CHF",
"GBP/JPY",
"GBP/CHF",
"AUD/JPY"

The Account I track here I will still leave only EUR/USD for now. But some slight changes to the TP and SL.
__Thanks__ MJ
  • #104
  • Feb 27, 2009 4:04am Feb 27, 2009 4:04am
Quoting M_j
PREVIOUS REPORT

TOTAL TRADING PERIOD: START 28.January 2009

START Balance:100000
END Balance: 105497.54

Trades: 23
Profit: 19 | Loss: 4
WIN/LOSS Percentage: WIN: 82.6086956522 | LOSS 17.3913043478


PS: because the transaction gets longer I attach it as a exported excel:
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot-4.png Size: 26 KB
Attached Image
Attached File(s)
File Type: xls TransactionLog.xls   22 KB | 311 downloads
__Thanks__ MJ
  • #105
  • Feb 27, 2009 12:48pm Feb 27, 2009 12:48pm
Hi MJ,

Just found your journals, very interesting as I just started forex trading with alpari demo account and trying to implement Spud's stochastic theory.

Have you tested these bots on real live account or just implementing it with demo accounts?
The results are great! I would like you to keep your journal as I would be following how things are going.

So, these tests you are running are metatrader EA?
Just to confirm, you said this is just based on price action and stochastics (MTF) yeah?
I am thinking about doing the same thing you are doing, hopefully I would have the time to do it, as I have to also work during the day.

Thanks,
Joan.
  • #106
  • Feb 27, 2009 12:51pm Feb 27, 2009 12:51pm
I forgot to ask you this question:

Which time frame do you use for all the trades, and if I am not mistaken, the screenshots you posted are using OANDA? If the EA is written in metatrader how do you implement it with OANDA?

Thanks again,
Joan.
  • #107
  • Edited 5:11am Feb 28, 2009 12:06am | Edited 5:11am
hi JoanArc,

As you are new you are forgiven!!

Properly you did not read everything here nor did read most of my posting in Spud's Predictive MTF Stochastics http://www.forexfactory.com/showthread.php?t=119483. Most of my Spud's related ideas as well as many technical approaches are described throughout the posting.


To your questions:

* Have you tested these bots on real live account or just implementing it with demo accounts?


again, it is written in the journal
Quoting M_j
If everything goes overall well - I properly will apply one version of it to a small live account in March.
Anyway, until now only Demo --- BUT it does not matter as Live is just the same. I did occasionally do some manual Live trading based on MTF stochs.



* The results are great! I would like you to keep your journal as I would be following how things are going.


I'm planing to keep the journal at least a little bit longer. Not sure for how long might not make much sense when I take it live.



* So, these tests you are running are metatrader EA?


again, it is written in the journal
Quoting M_j
hi fellow trader:

* for this joined project I will be using as broker: OANDA.

* Just to confirm, you said this is just based on price action and stochastics (MTF) yeah?


YES

* I am thinking about doing the same thing you are doing, hopefully I would have the time to do it, as I have to also work during the day.


I do not have a day job to bother. But I have to say it was a real lot of work. (just to give you a idea - the source code for only Spud's MTF stochs implementation has more than 25.000 lines. and the file is about 2.5 MB big.)

* Which time frame do you use for all the trades, and if I am not mistaken, the screenshots you posted are using OANDA? If the EA is written in metatrader how do you implement it with OANDA?


I have a very flexible bot system so I test with different time frames.
For each bot the maximum Time frames to include are 6 levels. But I could also include just one Time Frame.

Yes all the posts are OANDA's as I use them as stated above.

Quoting M_j
Thanks Skyline. Well EA's are not too useful for me as I'm using OANDA as my Broker/MM.

I'm using an own developed API/System - so I also do not have all the fancy indicators. (I have to program each one myself)

Anyway - I would be very interested implement Spudfyre's approach as a fully automatic system.


I hope this answers most of your questions. So if you are seriously want to follow my process you should really read most of my postings.

And study especially Spud's threads - the main once I did were:
http://www.forexfactory.com/showthread.php?t=119483

http://www.forexfactory.com/showthread.php?t=36574
http://www.forexfactory.com/showthread.php?t=37111

but also skimmed over some others.


All the best to you - and may I suggest something: if you really follow this through open your own relevant Journal - which could be of some value.
__Thanks__ MJ
  • #108
  • Feb 28, 2009 9:04am Feb 28, 2009 9:04am
I did some more programming on the MTF stochs system. Hope I did not get too many new bugs. It was easier than I thought - but I can not test it as there is no prober data feed on the week end -- so I don't know if it will work like that.

related to what I wrote here:

Quoting M_j

There is one thing I still might want to implement: At the moment the lowest defined time frame is used to decide the trades direction: which has advantages and disadvantages.

One of the disadvantages is that it does not nicely fit into Escalator to Pips. So I might still pull myself together to do a work around -- but this is quite some work.

Bottom line: it would not be possible to trade manually the current implementation. too complex.
__Thanks__ MJ
  • #109
  • Feb 28, 2009 9:39am Feb 28, 2009 9:39am
Hi MJ,

Thanks for forgiving me, as you could tell this is my 3rd post in this so I am really new on Forex.
I actually did read a some pages of Spud's MTF Sto thread, but as I went through the pages some people started intorducing their own methods and it gets confusing so I stopped reading. I haven't gone through all the links you provided.

I have done some testings on manual tradings using MTF Sto, it looks very promising, most of the losers that I got were because of my lack of discipline, entering trades before all the criterias were met.

I am still reading the basic of writing EA with MT4, so I know, it will take sometime for me to get this to where I want.

My background came from trading stocks 10 years ago, I have written many indicators and code for backtesting with Amibroker, now I am switching to Forex, so I need to convert some of my codes into mq4, so that's why I need to learn a lot about EA and may be turn stuff into auto trading at some point (as I already told you, I have a day job, it's hard for me to trade and work).

So from my understanding, you have written some APIs to be used with OANDA, I am thinking about opening account with OANDA myself, so I will probably need to investigate this dll integration as well, and I believe it's not free.
As a start I will start coding the EA with MT4 and as soon as I understood the layout and background I need to know, then I will start coding, this might take a while, metatrader doesn't really have a good object oriented development environment, you know, like Visual Studio 2008, well I will need to live with what is offered.

Thanks for your time answering my questions, I need to start asap as this project might take me years to complete.
My goal would probably start with writing functions that detect peak and bottom of a curve, ascending/descending line, these functions should take timeframe and periods as some of the params. etc.
Then codes will be added as I go when testing. I will try to write as object oriented as possible, this way modifications would be easy.

Thanks again, good luck on your journey writing working EA, I believe you with the lines of code around 25,000 may be more, as I would probably end-up doing the same thing.
Considering different conditions of market, we might need to consider the code to have the ability of 'learning' as new ticks are coming in, so then there will be no 'curve fitting' involved for certain time frame of market conditions. We might as well include some calculations to kind of anticipate an order to be executed when our 'goal' price is reached.
Oh well, so many things to do, I really need to start.


Joan.







Quoting M_j
hi JoanArc,

As you are new you are forgiven!!

Properly you did not read everything here nor did read most of my posting in Spud's Predictive MTF Stochastics http://www.forexfactory.com/showthread.php?t=119483. Most of my Spud's related ideas as well as many technical approaches are described throughout the posting.


To your questions:

* Have you tested these bots on real live account or just implementing it with demo accounts?



again, it is written in the journal...
  • #110
  • Feb 28, 2009 9:50am Feb 28, 2009 9:50am
Before Forex I also did stock trading - but manually.

Quoting JoanArc
Hi MJ,

I actually did read a some pages of Spud's MTF Sto thread, but as I went through the pages some people started intorducing their own methods and it gets confusing so I stopped reading.
I understand. Also Spud himself explained his ideas in many different ways which did confuse some people - for myself I liked it as it opened a lot of option to experiment with.
But there are a couple of pdf you can download
Attached File(s)
File Type: pdf MultiTimeFrameClosedCandleHelp.pdf   134 KB | 488 downloads
File Type: pdf SONIC PMTFS.pdf   529 KB | 561 downloads
File Type: pdf Spudfyers Predictive Stochs Pt. 1.pdf   1.2 MB | 1,839 downloads
File Type: pdf Spudfyers Predictive Stochs Pt. 2.pdf   911 KB | 3,095 downloads
File Type: pdf Spudfyers Predictive Stochs Pt. 3.pdf   780 KB | 669 downloads
File Type: pdf SpudfyerStochasticElasticity.pdf   288 KB | 1,536 downloads
File Type: pdf Spudfyre's Escalator to Pips.pdf   237 KB | 391 downloads
__Thanks__ MJ
  • #111
  • Feb 28, 2009 9:55am Feb 28, 2009 9:55am
more pdf some might be similar
Attached File(s)
File Type: pdf Spuds MTF Stochastic Combination.pdf   1.3 MB | 652 downloads
File Type: pdf Spuds Rope Method.pdf   1.3 MB | 866 downloads
File Type: pdf Spuds Stochastic Thread Theory.pdf   692 KB | 424 downloads
__Thanks__ MJ
  • #112
  • Feb 28, 2009 9:56am Feb 28, 2009 9:56am
Hi MJ,

I forgot to ask you some questions:

1. When deciding an entry point (long/short), I believe you are using the smallest timeframe and goes to the higher timeframe, when going from the smallest timeframe to higher, do you use standard 1M,5M,15M etc...?
The reason I ask you this is because there might be possibility of going to the higher timeframe not using those default numbers, for example:
we could write our own stochastic formula (the standard iStochastic function from MT4 takes only defined timeframed, I dont' believe you can enter 3 Minutes, or 6 Minutes for example, it only takes predefined timeframe enumeration) that takes whatever timeframe we enter, as I believe the timeframe would be some linear number that starts from 1 minute (since this is the smallest free data available for backtesting) going all the way to whatever timeframe you want.

2.Apart from the MM function that calculates lot size, SL, TP, Trailing stop, etc, does your API code call some function written in different language? (MatLab for example for complex calculation), I don't think it is, since you said you only use stochastics, I just wanted to make sure, because if you did, I would really be interested in knowing how MT4 interacts with other programming language.

Thanks again,
Joan
  • #113
  • Feb 28, 2009 10:00am Feb 28, 2009 10:00am
Thanks, for the PDFs MJ, I downloaded them and will go through them one by one.

Thanks for all your help!

Joan.
  • #114
  • Feb 28, 2009 10:07am Feb 28, 2009 10:07am
Quoting JoanArc
Hi MJ,

I forgot to ask you some questions:

1. When deciding an entry point (long/short), I believe you are using the smallest timeframe and goes to the higher timeframe, when going from the smallest timeframe to higher, do you use standard 1M,5M,15M etc...?
The reason I ask you this is because there might be possibility of going to the higher timeframe not using those default numbers, for example:
we could write our own stochastic formula (the standard iStochastic function from MT4 takes only defined timeframed, I dont' believe you can enter...
> 1. When deciding an entry point (long/short), I believe you are using the smallest timeframe and goes to the higher timeframe, when going from the smallest timeframe to higher, do you use standard 1M,5M,15M etc...?

well I used to have my lowest defined Time Frame as the main one to decide the direction: But if you read my new post from today
Quoting M_j
I did some more programming on the MTF stochs system. Hope I did not get too many new bugs. It was easier than I thought - but I can not test it as there is no prober data feed on the week end -- so I don't know if it will work like that.

related to what I wrote here:
I started to program so that I can configure any of the 6 TimeFrame Levels as direction giver.


> when going from the smallest timeframe to higher, do you use standard 1M,5M,15M etc...?

in praxis I use only the time frame for which OANDA also provides charts:
5sec, 10sec, 30sec, 1min, 5min, 15min, 30min, 1hour, 3hour, 1 day.

If needed I could do others

But which one I use depends on each bot - I can configure it completely freely.


> we could write our own stochastic formula

I wrote the one I use myself. so I could feed it any data.

> .Apart from the MM function that calculates lot size, SL, TP, Trailing stop, etc, does your API code call some function written in different language? (MatLab for example for complex calculation), I don't think it is, since you said you only use stochastics, I just wanted to make sure, because if you did, I would really be interested in knowing how MT4 interacts with other programming language

Always remember that I don't use MT4 to auto trade.
__Thanks__ MJ
  • #115
  • Feb 28, 2009 10:14am Feb 28, 2009 10:14am
Great thanks!
I think I have seen someone wrote a MTF stochastic indicator with some interpolation to smooth out the lines. I forgot where, I should look into that again.
I have always been a fan of Stochastic and WPR!

Thanks again MJ, have a great weekend!

Joan.

Quoting M_j
> 1. When deciding an entry point (long/short), I believe you are using the smallest timeframe and goes to the higher timeframe, when going from the smallest timeframe to higher, do you use standard 1M,5M,15M etc...?

well I used to have my lowest defined Time Frame as the main one to decide the direction: But if you read my new post from today


I started to program so that I can configure any of the 6 TimeFrame Levels as direction giver.


> when going from the smallest timeframe to higher, do you use standard 1M,5M,15M etc...?...
  • #116
  • Edited 10:54am Feb 28, 2009 10:15am | Edited 10:54am
> have done some testings on manual tradings using MTF Sto, it looks very promising, most of the losers that I got were because of my lack of discipline, entering trades before all the criterias were met.


It depends what criteria you use

Many others gave up.

Most of the time on the charts it looks easier than it is in praxis. Even with a lot of filtering I still get from time to time 'really bad entries - just wrong'.

So it's also the find the right balance of TP, SL - less filtering - more filtering. Each with it's on advantage and disadvantage.

From my 3 month of experimenting I still see a lot of possibilities.

Seems also it works quite the same across pairs. Soon I will know more about that as I start using 10 pairs this week.
__Thanks__ MJ
  • #117
  • Mar 2, 2009 1:26am Mar 2, 2009 1:26am
Big configuration mistake.

I checked this morning how things went overnight and was surprised that some trades were cancelled others had much much huger TP or SL settings. Took me a while to find the mistake. Actually I misconfiguration my price updater very badly.


The account I track here was only so much affected as that a valid trade was not executed.


On the positive side - it seems the changes (actually quite a number of them) seems to be working so far.
__Thanks__ MJ
  • #118
  • Mar 3, 2009 3:09am Mar 3, 2009 3:09am
Have got somewhere a bug - which wrongly executes sometimes more than one trade !! this is really annoying.

So till now I have to disregard all the new tests.


On the tracked account I had one loss.


I start to setup a small live account - should be ready in a couple of days.
__Thanks__ MJ
  • #119
  • Mar 3, 2009 4:45am Mar 3, 2009 4:45am
Think I found a nice solution (much better what I had used before to avoid immediately double entries). Changed the code have to test now.
__Thanks__ MJ
  • #120
  • Mar 4, 2009 12:49am Mar 4, 2009 12:49am
Since my last changes I had 40 trades --- none of them wrong doubles - so hope I finally solved it.

As this week was a messing around of changing - programming - bugs ... it was so far worthless in any statistical way.

I had over night a profitable trade on the tracked account -- still below last weeks high.
__Thanks__ MJ
0 traders viewing now