Machine designed strategies. Do they work?

[luckyputanski]

Quote from jcl:

How does Price Action Lab avoid curve fitting? From their program description, it seems to be the textbook example of curve fitting, as it verifies its price patterns using the same data in which those patterns were found. This way it can generate any pattern, even from random data. Or does it use some OOS test?

According to their website, everyone does it wrong, they're the only ones that are right and they can sell you the software for few grand, so you can be right too.
PriceActionLab is a definition of curve fitting, it's surprising that such blatant advertising on ET goes unbanned.

 

[tim888]

Quote from jcl:

How does Price Action Lab avoid curve fitting? From their program description, it seems to be the textbook example of curve fitting - it apparently verifies the price patterns using the very same data in which those patterns were found. This way of course you can find any pattern, even in random data. Or does it use some OOS test?

It has a one-click OOS test, something that is very honest and straightforward. Some other programs I tried in the past may be using the OOS to improve IS results. I was attracted to this approach because it is outright honest. You can hide the OOS from the search and then apply it afterwards.

 

[tim888]

Quote from luckyputanski:

According to their website, everyone does it wrong, they're the only ones that are right and they can sell you the software for few grand, so you can be right too.
PriceActionLab is a definition of curve fitting, it's surprising that such blatant advertising on ET goes unbanned.

I think you have it wrong as far as curve-fitting. Price action cannot be curve-fitted because it is already the curve. It can be biased, I agree, but not curve-fitted. You must review your understanding of curve-fitting. I had the same misconception as you before. I had to go to my local college and talk to a math professor. Reality cannot be fitted to reality, this does not make sense. If you realize that you will start making money and be relieved from the various frustrations that cause you to declare posts discussing approaches to trading as blatant advertising, especially when they contain no links and such.

 

[jcl]

Quote from tim888:

It has a one-click OOS test, something that is very honest and straightforward. Some other programs I tried in the past may be using the OOS to improve IS results. I was attracted to this approach because it is outright honest. You can hide the OOS from the search and then apply it afterwards.

This is interesting, as I just described a test with another software for finding price patterns - you can read about it here -

http://www.elitetrader.com/vb/showthread.php?s=&threadid=239403

- and the test results strongly suggest that single 3-bar price patterns, such as generated by PAL, have no predictive value. They appear highly profitable with in-sample data, of course, but that's classical curve-fitting. They fail in OOS tests.

As you seem to have experience with Price Action Lab, maybe you can comment on that test, and on the difference of the test results and the claims by PAL? Surely there must be an explanation of the discrepancy?

 

[oilfxpro]

Quote from Humpy:

Are you saying tht the billions paid out to maths Phds by the big boys are probably money wasted ?

The money has become too efficient with these algorithms to leave any scraps for little guys.It is time money wasted to invest any more.

None of these outfits ramped up are worth bothering with ,out of their 100 strategies , the net effect is loss after spread and paying commissions to these scammers.

 

[oilfxpro]

Quote from v75z52:

I have looked at Trading System lab but they do not provide a demo. Some serious hype on the website but this is what a user of this program recently said:

"TSL doesn't work. Tried it personally. Was not able to develop a single working strategy."

A user of Adaptrade claims that "Strategies quickly fade"

Another user of Adaptrade claims that "strategy does not perform (equity curve goes down) with new data beyond the out of sample period."

Now it may be the case that these individuals failed to use the programs properly. What do you think?

I got 44 automated systems that worked in history ,I don't use them because they don't carry a guarantee of future performance.
The profitability has gone down due to the uncertainties of this financial crisis.

If the o/h/l/c scenario changed from extreme volatility to low volatility market conditions and noise volatility increased , with more of false market moves than profitable moves , would the systems be profitable?No.

 

[Rationalize]

Humpy

Are you saying tht the billions paid out to maths Phds by the big boys are probably money wasted ?

Quote from alexandermerwe:

The Phds work on pricing models for synthetics and on arbitrage. They rarely get involved with system trading. Some hedge funds use Phds for system development, like for models involving cointegration and related stuff. Big firms move size, they do not need trading systems. Those that try to predict what the big firms will do need the systems. I hope it is clear now.

+1

Right on the money there Alexander, but don't forget the big firm's execution algos. There's a bit if quant work there if you want to be moving big blocks quietly.

 

[oilfxpro]

Quote from Rationalize:

+1

Right on the money there Alexander, but don't forget the big firm's execution algos. There's a bit if quant work there if you want to be moving big blocks quietly.

So why does eur usd move in same direction as dax?

 

[Rationalize]

Quote from oilfxpro:

So why does eur usd move in same direction as dax?

I dunno. But if you're making money punting that, you're clearly smarter than me.

 

[oilfxpro]

Quote from Rationalize:

I dunno. But if you're making money punting that, you're clearly smarter than me.

Smarter isn't the correct word , experienced is more appropriate , or able to read context correctly.