A cautionary tale for system developers
- Thread starter Jerry030
- Start date
nitro Jerry,
I think if you can provide data between 2004-2007 it might be more convincing.
2003-2004 is the best time for long nasdaq 100, your spreadsheet shows 60% gain vs. 50% buy and hold.
People might jump on board with a system which can generate just 30% return last year with similar risk like yours.
I think if you can provide data between 2004-2007 it might be more convincing.
2003-2004 is the best time for long nasdaq 100, your spreadsheet shows 60% gain vs. 50% buy and hold.
People might jump on board with a system which can generate just 30% return last year with similar risk like yours.
codedeep007,
My historical data base ended in July 2005 in terms of the back data. The final year was held out as an additional Out of Sample test pending the integration of the signals I generated with the trading strategy and system/platform of the former potential client. The first 50,000 bars were used in training the model, so you never use that data for strategy development or performance analysis as the modeling application has "seen" it in training.
Normally I only trade futures as the leverage is much better but the request was for a stock trading system that traded NASDAQ stocks, so the NSADAQ 100 Index future was the closest I could come with using the data I had on hand to do an initial test. The idea was to either trade a basket of stocks that averaged close to the index or apply the same modeling process to individual stocks.
If you've got data for a period you're interested in seeing I'd be happy to run it as have the models saved or I can run it on the period of 4/2004 to 5/2005 if that represents different market conditions.
Trading stocks the annual return on account would have been less than 50% annually while trading the Index future would have returned several times the total account. Thatâs why Iâve never looked into stocks.
I'm siding with Jerry. He started the thread the right way then Crisis opened his mouth to remove all doubt.
I agree the thread needs some pillow biting references. That or animals.
No one has mentioned crack pipes.
I agree the thread needs some pillow biting references. That or animals.
No one has mentioned crack pipes.
I started the thread as a kind of public service message for those folks who spend a great deal of time and intellectual effort at what is probably one of the most challenging areas of intellectual and economic endeavor: trading system development.
That's not a put down of other disciplines like genetic engineering, nanotechnology or pharmaceutical research. However in those areas you can patent your idea and protect your intellectual capital. The results and products created are visible in the marketplace.
In contrast a trading system is not. If someone is using your system or ideas they extracted from it, there is no way to know that. Such usage is totally invisible. For every âgreatâ trading technique sold retail at investor seminars and trade shows there are probably 10 traders who end up using it as people share it with associates and friends or sell it on the Internet out of the reach of copyright law.
So it's not a bad idea if we all help watch each others back.
In some earlier posts there was discussion on the issue of extracting a trading system from a DLL or similar software package. There is no need to try and figure out the system from the code itself. Rather one could use the same methods that the CIA and NSA use to de-crypt certain kind of coded messages. It's called brute force assault.
In a nutshell, since the software app will generate an output when presented with the input of a price pattern, say Open, High, Low Close for the last 10 bars one could use that fact to extract the logic of the system with the following means. Create a massive data base of all possible 10 bar price patterns (There will be millions of possible combinations) and feed it to the software application and save the output.
By mapping the output of a signal to those price patterns that generated it one can create a "fingerprint" for the internal logic. This could then be used to reverse engineer the system. The result would not give you the actual rules of "if the RSI is below it's 5 bar average and the ATR is more than 1 stdev from it's mean" but would yield its functional equivalent in a kind of mathematical mapping. A somewhat similar approach is used to reverse engineer computer chips, if I recall, to avoid copyright infringement.
Unbelievable!
What?!
I'm stunned!
Let's do a reality check here. A couple of posters can see who the real BS artist is...
If Jerry really did have something that could return these kinds of results, why would he be trolling around here looking for "clients"? Why go through all of that trouble? Why not just trade it himself from the beach of his private island?
My guess for the real motive behind this thread is that Jerry wanted to point the finger at us before we pointed a finger at him. The system we have, that Jerry offered to help us with, is specifically designed for day trading. In the following post, Jerry finally revealed that his system is too slow for day trading!
Why would he have offered to help and waste all of our time if his system could not work for day trading? Maybe because he was trying to get us to fund a project to speed up his system so that it could be tested for day trading. In other words, he was trying to scam us!
If he has this great system, why would he be trying to get $10,000 from us? He should be fat with cash.
If he is such a great developer of such an awesome system, why would he have to hire a C++ programmer?
He talks a good game through a veil of pseudo-intellectualism, but too many things just don't add up. I also wonder why he can't master hypertext. It's been around for decades now. I'm sure there is still some room on the bandwagon for some late adopters. But really, who knew it would take off like it did.
What do I have to gain here other than try to clear my name? I'm not the one trying to take your money. But my words of warning go unheeded. The lure of easy money is very strong.
Happy trading. I see Jerry has found some lemmings. I can't bear to watch them led to slaughter.
As DeeDeeTwo said, post it on ET...they will believe anything.
How did these two ever hook-up in the first place? Who contacted who (and based on what)?
Looks like a match made in hell.
Edit: Never mind, I figured it out: http://www.elitetrader.com/vb/showthread.php?s=&postid=1775085#post1775085