Well I got a little more data and that magic 15 lines of code weren't so magical anymore. Heck, I think Something I fail to do is to look at these failure signatures and do that one last, deep analysis that might expose something more interesting. I am due to type up here something I did with neural networks a long time ago to illustrate my point, but I don't have time for all the paragraphs I would need. I had some forex questions: 1. I am still getting historical data from Dukas. When I get to select the data specifications, it asks me if I want to know the bids or the asks. I have been downloading the bids so far. What trouble am I getting into just using the bids? It's my understanding that I should have both to determine the real price I'd get. Or is it really dependent on the broker? Does it not matter at the 1-minute window? 2. Does anybody have some tips on working with Forex in AmiBroker? Their own site shows how to change how a ticker is represented so it properly trades on margin, but it generally looks like forex is pushing it in Amibroker. 3. If I only get EURUSD can I safely assume taking a short position on it is 99.9999% equal to going long USDEUR? At the 1-minute window, I assume all arbitrage has happened already. I was working on a script to invert the information, but it looks like I screwed up something. 4. Is 100:1 margin the common way to trade forex spots?
no, 50:1 at most usa market makers and 40:1 where I trade at IB you'd be safe assuming 1 pip spread on eur/usd only USD/EUR I know of is the futures at the merc symbol 6e
If that works for you and you have a profitable trading system, then keep on doing what you're doing I personally have found that past performance has no bearing whatsoever on future performance. Choppiness may continue or may abruptly turn into a trend at any moment. And a trend may go and go and go, or it may randomly end and reverse, or end and turn into chop. And there's no way of knowing when or for how long. At least that's my experience. YMMV. And if it's working for you then, like I said, keep it up!
no, it's not working for me and I don't do it that way, I was just commenting that years and years of data and testing a system across all markets which is the usual way of declaring a system "robust" may not be any better than just analyzing the last 15 minutes of data as for me, I never predict I just react, but I always have an opinon. If I'm flat, I enter based on my opinion. After that, I just react until my opinionj tells me it is probably a good time to get flat again. But I think the whole thing could be automated, for sure the reaction part, it is just mechanical, not sure about the entry opinion, but the exit opinion could be.
I got this. Just a few lines of code; LA = Los Angeles or Louisiana; if (Home= LA) { Light 1Up=1; Exit = CloseAllPositions;
There is a pretty fundamental reason that no matter how fantastic you can make something perform on past data, it will eventually 'break.' What happens? Markets change... The best methods I've seen traders utilizing over the past few years have common themes: 1. multiple time frames 2. simple strategies - tailored to the trader's liking that basically get you in the market at the best possible time (new trend). 3. low risk - out quickly without hesitation
I am the trading God. If you need help, just ask. Been running a serious number of tested programs. But, my main GoTo program is called Doomsday. You are gonna be amazed at the number of Buy orders when prices go to $0. I won't be here but the program is on auto. Just browsing looking for a winter hobby/project, getting kinda bored. Hope I didn't honk-up the thread. <g>