3 Questions for Pros

[oleg 10]

Hi Pros!

 

I am designing automated trading software and I have some problem to understand Professional Traders’ mentality…

 

I publish my trades online in real-time. The results are fairly profitable (20-40%/week), but the first think I hear is “Do you have a back-testing result?”

 

The second question is always: “If the software is so good, why do you sell it?”

 

My answer to that is very simple: I trade, and I do what I do best – developing systems. Furthermore, I do not share the view that systems must be “top level” secret because the more traders use them the more reliable outcome will be (collective trading).

 

So, here are my questions:

1. Does it mean that hypothetical result of back-testing is more reliable and important than what you can see in real-time?

2. My tools is available for free, but download rate is about 0.5% of visitors, is “Free” scaring people?

3. What is your ideal of automated trading apps?

 

Thanks for your time.

 

Oleg Skoptsev

[Tams 20]

you can't win...

if you do backtest only, people are going to ask why not forward test...

 

LOL

[DugDug 10]

Hi Pros!

So, here are my questions:

1. Does it mean that hypothetical result of back-testing is more reliable and important than what you can see in real-time?

2. My tools is available for free, but download rate is about 0.5% of visitors, is “Free” scaring people?

3. What is your ideal of automated trading apps?

 

1 - no but it shows how the strategy may work over many different types of markets and instruments. Nothing works all the time - people want to at least gauge the drawdowns and volatility of returns.

2 - there is lots of free info on the internet yet still people look for more.

3 - one that makes money and I dont do anything......thats normally called a money manager.

[brownsfan019 17]

Only 20-40% per week?

 

That's probably why no one is downloading - who wants those horrendous returns?

 

 

Maybe people visiting your site realize that 20-40% per week is not sustainable, achievable, realistic, etc. etc.

[oleg 10]

Only 20-40% per week?

 

That's probably why no one is downloading - who wants those horrendous returns?

 

 

Maybe people visiting your site realize that 20-40% per week is not sustainable, achievable, realistic, etc. etc.

 

Have I mention that these are REAL-TIME result? As well as PL, Enter Time/Price.

And you can see PL update in REAL-TIME(5 sec delay).

 

[richd8980 10]

how long have yo been getting 20-40% per week?

[oleg 10]

how can we view the results you are talking about and the trading method?

 

I cannot publish links but you can shut me PM

[richd8980 10]

found your site nice profits however I cant afford the price. Great results have you had any loses since 11/17/09? it doesnt look like it from the spread sheet

[oleg 10]

found your site nice profits however I cant afford the price. Great results have you had any loses since 11/17/09? it doesnt look like it from the spread sheet

 

Perhaps you found my Pro site. I'll publish a link on my web to the new one there is a free stuff

[Tams 20]

.......... .......... ROTFLMAO

 

why is life so predictable?

[Kiwi 16]

.......... .......... ROTFLMAO

 

why is life so predictable?

 

 

And to our perpetual surprise we agree despite locations and nationalities.

 

As did Phineas T. Barnum.

[oleg 10]

found your site nice profits however I cant afford the price. Great results have you had any loses since 11/17/09? it doesnt look like it from the spread sheet

 

It's done.(link)

 

What you found that is exactly what most of you want - training, back and forward test results... And the reaction ... the same.

 

So, you have a back-test and ...

 

 

BTW: tere are results for 286 days and each trade by minute, you can check the prices.

 

In regard of spread sheet: do you really think that it is possible to put manually almost 146000 records with exact pricing?

 

Sincerely,

 

Oleg

[macdfx 10]

And to our perpetual surprise we agree despite locations and nationalities.

 

As did Phineas T. Barnum.

 

Watching this unfold is so much fun....:missy:

[macdfx 10]

It's done.(link)

 

What you found that is exactly what most of you want - training, back and forward test results... And the reaction ... the same.

 

So, you have a back-test and ...

 

 

BTW: tere are results for 286 days and each trade by minute, you can check the prices.

 

In regard of spread sheet: do you really think that it is possible to put manually almost 146000 records with exact pricing?

 

Sincerely,

 

Oleg

Of course....

 

If your trading is automated, so are the results. Enough of this "manually" nonsense. You ought to know how to take your data and format it for Excel...

 

That's if you have any real data, that is...:rofl:

[Tams 20]

It's done.(link)

 

What you found that is exactly what most of you want - training, back and forward test results... And the reaction ... the same.

 

So, you have a back-test and ...

 

 

BTW: tere are results for 286 days and each trade by minute, you can check the prices.

 

In regard of spread sheet: do you really think that it is possible to put manually almost 146000 records with exact pricing?

 

Sincerely,

 

Oleg

 

do I read it correctly? that's more than 500 trades per day?

 

what is your commission to net profit ratio?

[richd8980 10]

Is 500 trades a day a lot lol

[oleg 10]

do I read it correctly? that's more than 500 trades per day?

 

what is your commission to net profit ratio?

 

Yes, it is correct and speech goes about Professional systems based on biologically inspierd AI.

It uses up to 10 strategies that trade up to 10 contracts per strategy.

So it is not a lot ...

[Tams 20]

Yes, it is correct and speech goes about Professional systems based on biologically inspierd AI.

It uses up to 10 strategies that trade up to 10 contracts per strategy.

So it is not a lot ...

 

10 strategies?

 

that's more than 50 trades per strategy per day !

 

what is your commission to net profit ratio?

[sicktrader 10]

There are a few (just a few) gambling touts with winning systems that sell their picks because in the gambling arena there is a limit on how much cash/credit one person can obtain. In addition, if you constantly beat the book, then you ultimately get shut off and have to resort to using beards. Quite the pain in the ass.

 

There are no such limits in trading, and you can get down as much on a trade as you want. If it has been a successful system, then you should not have a cash issue.

 

I call bullshit, though I am hoping this thread will provide some entertainment.

[oleg 10]

10 strategies?

 

that's more than 50 trades per strategy per day !

 

what is your commission to net profit ratio?

 

In this case per/trade ratio is $143.64 and usually round treep cost is about $5.

 

 

This is an example of my training,back and forward testing reports.

 

And comming back to the original question what info you can get from this almost perfect result which is the purpose of the system training?

It was trained do not make mistakes?

Edited April 17, 2010 by oleg

[Tams 20]

In this case per/trade ratio is $143.64 and usually round treep cost is about $5.

 

 

 

This is an example of my training,back and forward testing reports.

 

And comming back to the original question what info you can get from this almost perfect result which is the purpose of the system training?

It was trained do not make mistakes?

 

thank you for your info.

 

I have many inquisitive questions, I hope you don't think I am trying to be sarcastic.

 

I don't understand your answer, I don't know if you are trying to beat around the bush, or don't think my question is important enough to give a straight answer.

 

As Judge Judy would say, "mmm" is not an answer.

 

What I am interested to know, and I think it is a fair question, is out of the trading sample you gave... how much net profits were there, and how much commission did you pay to obtain those profits. In simpler terms, what is the commission-to-profit ratio?

 

Most people don't worry much about commissions,

Maybe you don't either.

But in a system trading environment, the commission can make or break your deal.

Most high frequency systems pay as much in commission as their profits.

Some pay more to the broker then their banker.

 

This is not to insinuate a system is good or bad because of the commission they pay.

This is merely a question.

 

However, if you are not confident about your system, you might want to circumvent my innocent inquiry,

because I am sure you are aware, that your answer will lead to more questions...

questions which you might not be willing, or prepared to answer.

[oleg 10]

thank you for your info.

 

I don't understand your answer, I don't know if you are trying to beat around the bush, or don't think my question is important enough to give a straight answer.

 

Dear Tams,

 

I was under impression that the picture will speek for itself... and answer all your questions...

 

net profit per trade is : $143.60- $5 = $138.60 ...

[macdfx 10]

I doubt it.

 

He still hasn't addressed whether or not he could have his software output the data records into a format easily imported into Excel.

 

Since its his website and software, I should think he would be able to take care of this basic requirement as well.

 

Quite perfunctory for a software developer like Oleg, I should think.

[Tams 20]

In this case per/trade ratio is $143.64 and usually round treep cost is about $5.

 

 

This is an example of my training,back and forward testing reports.

 

And comming back to the original question what info you can get from this almost perfect result which is the purpose of the system training?

It was trained do not make mistakes?

 

if it was trained to not make mistakes,

and your commission-to-profit ratio is only 3.6% [edited for typo]

and you traded 510 contracts per day !!!

 

how come the profit is only 20-40% per WEEK ???

 

what is your MAE to profit ratio ?

Edited April 17, 2010 by Tams

[oleg 10]

I doubt it.

 

He still hasn't addressed whether or not he could have his software output the data records into a format easily imported into Excel.

 

That is exactly my point...

 

Why should I export it to Excel? To see the equity curve?

 

There is no drawdowns... System is trained do not have them...

So, the back-test does not answers your questions...

 

And, what are doubt? Result you see?

 

This is a reason why I am trying to disuss the solution.

Where is it?

Edited April 17, 2010 by oleg