I'm hoping to be a zerfetzenTOR Sounds good, I'll stick to stocks for some time as I learn my way around. About longer term vs. shorter term, I'm playing around with a custom method that yields the estimated probability of an increase or decrease on the following day, or as we all know, the next point in the series, of course. I've been testing it on 5 stocks, each with 5 series (Open, High, Low, Close, and Volume). So far, the testing is going well. Of course, it's just as important to estimate the size of an increase or decrease. At one position in my past, I was a statistical forecaster, but not with stocks. So of course I'm trying a variety of other good ole' statistical methods as well. At any rate, it seems I can nail the next day as an increase or decrease to my satisfaction so far (we'll see after I have a large enough sample of results). What are different ways to approach a 1-day-ish turn-around? Assuming you work all day, come home, run your forecasts and if comfortable with a forecast, trade after hours on an expected increase in Open? Put in the trade, hope it opens higher tomorrow, put in a sell after hours on tomorrow, then hope it opens yet higher the following day? That would imply that whatever it opens at, the order sells virtually immediately? I'm just guessing that's a possible scenario. What kinds of general strategies do swing traders use when they work all day? Thanks.
The strategy for swing trading is simple, it's best to search around for patterns recognition and info; what you typically do is to buy on a confirmed reversal and hold for couple of days until it reaches the normal range and then exit the trade, that's it. What you described as overnight trading is not necessary for swing trading because it's typically move within the prescribed range anyway, just buy and hold. sg20
the most reliable patterns i've found have been continuation type patterns like pennants and flags, if you aren't around to trade you can use buy stops slightly above consolidation areas for entry. I traded this way for several years before trading professionally, it will work great for you in the next bull market. Now it's better to just trade the noise.
You could be right of course, but I beg to differ. The only way a stock is unpredictable is if it is truly random. And it is a valid position in philosophy of science that nothing is truly random, depending exactly on its definition. Any exception to randomness is a pattern that can be discovered. But that's just my two cents.
I trade randomness everyday, and it is far easier to accept it than to try to predict it. Learn how to capture it and you will be able to make money everyday, that is the key to trading for a living.
sg20, it's funny you say that, because that's what I'm using, my own implementation of a nonparametric pattern matching algorithm. And it's showing promise.
I can accept an amount of uncertainty, but not true, pure randomness. Sounds good, we're saying the same thing, find patterns that overcome enough seeming-randomness for each of us.
Last question in this thread then, I promise. If you were to try swing trading after hours, what do you recommend for software, online brokerage, or what not?
yes, the key i believe is that your management of risk when you are wrong is strongly set in stone. too many traders look at how much they stand to make and not how much they can lose. I like to trade relationships (arbitrage) rather than predict where each instrument will go.