Quote:
Originally Posted by darthtrader3.1
This is a complete waste of time besides for getting experience programming. For algo systems you really probly want to start with Acrary's threads on elitetrader.com and the book Evidence Based Technical Analysis..You can at least get from that the idea of backtesting against a random variable instead of backtesting on what will end up as nothing more than data snooping and over optimization..Even then though, retail algo software is a complete joke that borders on a marketting scam.
Seykota has a website because Seykota was left in the dust 20 years ago.
Jim Simmons doesn't have a website that tells you what to do and you have to trade against him.
Really, your best bet is to forget algo trading unless you want to under take a decade long project and just forward test your brain on charts..
Sorry, I think my post was a bit confusing. I'm not trying to program trade, although I do want some sort of "system" in place for discipline etc.
I'm quite aware of the dangers of curve-fitting, having spent my previous job trying to convince the field of psychiatric genetics that many of the findings were of a similar problem (publication bias when analysing huuuge datasets). I failed, sigh. Hopefully I can do better with trading
Basically I'm trying to ask this: For a newbie, how much should you just stick with a basic plan (think basic trend following with fundemental filters), and how much should you trust your judgement?
I know its an open ended question, just looking for some general guidance I guess....