Trading the forex fully automated martingale

[Drip Torch]

And so our mad hero rights off into the sunset.

 

[Trombonista]

Me after reading and merging this guy's posts:

What? It's the face of the people who sank all their money into the FX.

 

[Syzyg]

while i was in indonesia, i downloaded the whole blockbrainchain and the command line interface.

On your old tradecode account on forexfactory you posted the following

asalamualaikum, saya pernah tinggal di indonesia sebelumnya .. saya tinggal di bogor .. saya dulu senang pergi ke kebun raya ..

Translation:

Asalamualaikum, I used to live in Indonesia before.. I lived in Bogor.. I used to love going to the botanical gardens..

How was your time in Indonesia? Do you still remember the language?

Good trading Edward.

 

[Devolve]

a successful backtest's equity curve will look so good, like this image, but it aint gonna be like that when trading live:

This is referred to as datamining bias

Basically the problem is, if you systematically search for solutions to solve for some dataset, theres an extremely high probability to solve it for that dataset and not for the general problem (IE: the future behavior of your set), in fact if you have an in-sample and out-sample that both happen to be a solution, theres still a high probability you simply got lucky. Theres far more sophisticated sampling methods to determine the probability that the solution is generalized rather than overfitted (IE: lucky), but that gets really complicated with some monte-carlo simulation p testing shit... I cant find a good resource dumped into a simple article, but I could link some literature that describes the problem and why the solution works when trying to find generalized solutions to forecast some timeseries.

This is half the problem since you still need to come up with some systemic way to generate forecasting, like you said, with moving averages or simple indicators or whatever you like, or more sophisticated methods like using stochastic models that can fit to the data youre interested in, but thats another rabbit hole (jump-diffusion stochastic methods).

 

[tradecode]

When I explained that back testing is not a valid process, await offers to show me articles showing me that it can work.

When I explain that the basic indicators do not work, await tells me that he read an article that the RSI is the most predictable of all the basic indicators.

If back testing in the basic indicators actually worked every automated trader with a math degree would be rich, but instead phds getting knocked out everyday by this business.

On your old tradecode account on forexfactory you posted the following

Translation:

How was your time in Indonesia? Do you still remember the language?

Good trading Edward.

Do you know why you researched my previous posting history?

And do you know why when I talked about forming a team you didn't say one word?

Because you're not a person that seeks out solutions but instead you want to prove something to be a flawed system.

I've been talking to my girlfriend tonight and I think I'm just going to scrap this project move on with my life.

Yep

 

[Devolve]

When I explained that back testing is not a valid process, await offers to show me articles showing me that it can work.

When I explain that the basic indicators do not work, await tells me that he read an article that the RSI is the most predictable of all the basic indicators.

If back testing in the basic indicators actually worked every automated trader with a math degree would be rich, but instead phds getting knocked out everyday by this business.

Just letting you know your critique of backtesting is well documented and addressed in peoples work; as you said, any idiot can fit their strategy to a day of data combed over a billion times, it takes real work to extract meaningful patterns from that data.

I feel like you really latched onto my RSI comment during our conversation and Ive kind of enjoyed you repeating some funny part of it, I feel like it should be pretty apparent that I dont subscribe to the same methodologies that day trading TT / Metatrader users do (although I know successful people with them); the tool is just crude.

 

[tradecode]

If I told you that car tires made out of bubble gum will not work, I think you would come back and tell me about all the articles you've read that tell me that if you freeze the bubble gum and then you bang it with a hammer that it might work.

And when I told you that the indicators don't work you came back and told me you'd read an article that RSI sort of works.

I am definitely going to scrap this project because everybody just argues with me online my partner's steal my code from me so I'm just going to accept it for what it is and I'm going to throw this code away and I'm going to move on with my life.

I have other things to do like taking a hot bath and jacking off.

Just letting you know your critique of backtesting is well documented and addressed in peoples work; as you said, any idiot can fit their strategy to a day of data combed over a billion times, it takes real work to extract meaningful patterns from that data.

I feel like you really latched onto my RSI comment during our conversation and Ive kind of enjoyed you repeating some funny part of it, I feel like it should be pretty apparent that I dont subscribe to the same methodologies that day trading TT / Metatrader users do (although I know successful people with them); the tool is just crude.

As I said I'm scrapping this project

 

[Devolve]

If I told you that car tires made out of bubble gum will not work, I think you would come back and tell me about all the articles you've read that tell me that if you freeze the bubble gum and then you bang it with a hammer that it might work.

Im showing you empirical evidence that addresses your critique on the methodology; Im giving you a real answer and showing real methods that step beyond the discrete, candle analysis driven methods from nearly 70 years ago and giving ideas about advancements somewhere in the 90s or 2000s regarding analytical methods for timeseries forecasting. they use these methods for weather, electric grid usage, airplane ticket sales, any random hard problem thats extremely high noise to signal ratio. Im not here to tell you your model is bad or wrong, but you should see things beyond what some dipshits using crap click trade software will sell you, it might even help you.

 

[tradecode]

I just had an idea.. before I scrap this project, I'm going to let my billionaire friend read this thread and I want to let him see the resistance that I'm up against.

I'm going to send him a link to this thread.

And I'm going to tell him that I am at the point where I am about to scrap this project but that I'm working on something way cool.

If nothing else he's going to get one hell of a good laugh.

Here's an interesting side effect of doing this..

I wonder if your behaviors will change any knowing that there is a monster angel investor in this thread?

Keep talking babies.

 

[Caesar]

Someone tell him he should unblock us all before linking his friend, so he can point out particularly stupid posts to him.

 

[Devolve]

You know Tradecode, reading this thread reminded me how much I used to really enjoy thinking about trading strategies; I remember first learning about candle patterns and indicators, eventually finding weird indicators from different backgrounds like the swami charts I linked or custom candles setup like ichimoku clouds... all kinds of silly stuff, for me it wasnt forex but crypto, I think around 2016. I wasnt satisfied with the answers I got then, since to me it seems like youre right, all the indicators dont mean shit unless you dial them in to fit your data, but if you fit it to a few days, of course the next few days completely destroy you. On top of that, wouldnt your own order execution alter the future projection of the market? wouldnt it react to your traffic and drift the trajectory in another direction? What about the assumption that you'll actually execute at the price you want? things like slippage or, if youre proactive you could place limit orders hoping the book will slide in your favor to enter your position, but now you need to manage that aswell and becomes a liability the longer it remains open... it was all a big mess to me, since the tools seemed like oversimplifications to a way bigger problem.

I remember going really far into the more analytical or statistical approaches, really trying to understand those for a few years, though I wasnt talented enough in any of the subjects needed to really execute at that level, in code or math or trading, and by the time I could really think about using any of these things my job sort of took over all my time, and my capacity to learn due to a demanding high pressure environment. The experience I gained showed me how hopeless it was to try to think about this stuff independently, since the actual execution at a competitive level is just millions of dollars in infrastructure and data and compliance and development to keep up with rotating exchange protocols and instrument updates and... the list goes on forever. I stopped thinking about it and started focusing on other tangent subjects that had similar problems, but no longer required me to have so much infrastructure, things I could accomplish with meager hardware from home.

It was nice chatting with you about this, since now I feel like I have new eyes and thoughts about approaching these problems again. Thanks for the fun and resparking some old interests I had; I hope you'll take some of my advise and knowledge to heart and in turn find news things to share with us since it seems to me, knowledge has the rare property of multiplying when it divides.

 

[tradecode]

Await he's a armchair warrior.. he's gathering up a folder full of theories and ideas, but he puts none of them into action.

You argue with me so that you will gather little pieces of information from me because you are thieves of my intellectual property.

Your ideas about back testing look good on paper but they fall apart in real practice.

Your ideas sound good when it's a dozen kids sitting around kicking around ideas, but go ahead and open up the live account and you will get your ass kicked.

It's like it's easy to sit around and talk about theories about how to be a boxer, but then you step in the ring.

It's like an old quote that I heard Mike Tyson say.. everybody has a plan until they get punched in the mouth.

 

[Asian tech support]

I wonder if your behaviors will change any knowing that there is a monster angel investor in this thread?

Call him a nigger for me, champ

 

[tradecode]

Do you honestly think that I would bring a serious contact of mine into a shit hole like this and allow you to fuck up my relationship with him?

You trannies are effing retarded

Only amateurs believe in backtesting.

Only amateurs believe in the basic indicators.

So many people fact-checked me by looking up my old posts on other forums..

Trading doesn't care what I said on any forum.

You people are bitches.

 

[Caesar]

I hope the gf being home brings us some fresh content

 

[tradecode]

I hope the gf being home brings us some fresh content

Note: nsfw audio content, lol

 

[Prokhor Zakharov]

This thread has been shrinking, which nigger is deleting their posts? Or is a janny mopping up the shit?

 

[Syzyg]

This thread has been shrinking, which nigger is deleting their posts? Or is a janny mopping up the shit?

A moderator has been merging Edward's posts together to assuage the concerns of another moderator, that is, of doubleposting.

 

[Prokhor Zakharov]

A moderator has been merging Edward's posts together to assuage the concerns of another moderator, that is, of doubleposting.

That's rather autistic.

 

[Free Dick Pills]

That's rather autistic.

taking the time out of your day to dumb-rate every single post by a particular user is about en par, fyi