You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Machine designed strategies. Do they work?
- Thread starter v75z52
- Start date
The goal is to fit a real underlying process and not fit random noise. And to detect when the underlying process changes.
"Curve fitting" and "optimization" are terms that have a bad public press but are used every day to find solutions to real problems in a large number of fields. Most people working in this space aren't that stupid. They know the issues.
Do the techniques need improvement? Of course. Are the techniques applied incorrectly by some people? Of course. Does that make the techniques invalid? Does that mean you stop trying to improve them?
"Curve fitting" and "optimization" are terms that have a bad public press but are used every day to find solutions to real problems in a large number of fields. Most people working in this space aren't that stupid. They know the issues.
Do the techniques need improvement? Of course. Are the techniques applied incorrectly by some people? Of course. Does that make the techniques invalid? Does that mean you stop trying to improve them?
Quote from ssrrkk:
I could be wrong but I think he means risk management in longer-term asset allocation strategies. I have a neighbor who does this type of work in NYC. Very dry and slow-paced, but lots of funds are doing this type of work I think.
This is middle office mumbo jumbo & buy side voodoo. Not trading strategy development.
Risk managers use these models to police the front office.
Risk to Trader: "You need to unwind that, because the model says your risk is too concentrated".
The middle office is another career path for failed traders.
Quote from Rationalize:
This is middle office mumbo jumbo & buy side voodoo. Not trading strategy development.
Risk managers use these models to police the front office.
Risk to Trader: "You need to unwind that, because the model says your risk is too concentrated".
The middle office is another career path for failed traders.
Actually that is sort of what I meant -- not so much for trading (day or swing) but more for balancing risk in a long-term portfolio of assets. My neighbor does not describe himself to be a trader but is a principal in a small financial firm that specializes in asset allocation. My guess is that longer term, there is bound to be more signal as you can model things like interest rates, inflation, GDP, real estate prices, corporate profits, etc etc...
Quote from luckyputanski:
Can we stop blatant PAL ads? According to them, everyone else is doing everything wrong, they're the only ones who know how to create systems. Stating that indicator based system fail by nature is simply not true. Systems from PAL have the same (if not bigger) chance of failing, and no number of posts on ET and their blog can change it.
I did find the authors' claim that certain system types had higher failure rates weak, since he didn't provide any empirical substantiation.
Laughable.
Pay any commissions?
Pay any commissions?
Quote from fx666:
Hi,
I have been using both APS (the predecessor of PAL) and Adaptrade Builder.
They do work and you can produce great systems with them but you need to do following:
1) Know very good the instrument which you are trading and its basic characteristics.
2) Provide such input to the program which exactly fits to (1).
3) Carefully choose the in sampe data.
4) Verify the boundary conditions under which your system remains valid.
5) Analyze carefully the results and if necessary apply changes manually.
Bottom line: you need to have really good knowledge about the instrument you trade in order to be able to take advantage of the programs. If you want to end up with a profitable strategy after a couple of minutes without knowing what you do, then you will fail even with the best program. Remember, these are just tools which take some work load from you but you need to provide them with the right input.
I have attached a screenshot from a system for EURUSD which I trade live. I have used APS to find the patterns and have combined them in the EasyLanguage code. The equity curve which you see contains 2 months live trading.
If you have questions you can contact me.
Many of the losers here may think you advertise the programs because they think like losers and will stay losers all their life. My main problem is that I cannot afford to purchase a lot of software at this point. But I believe some people have success using them. Is the APS the same with Price Action Lab?
The equity curve looked like commissions were missing.
Looks like the entire post has disappeared now.
Looks like the entire post has disappeared now.