like if reddit is buzzing about certain stock you can predict with rather high probability that it will spike when forums are chatting up and the stock ticker gets mentioned more and more often
or if you see certain pattern develop you can predict with certain probability that the stock will follow through with the pattern because most people read same books and recognise same patterns. what you cant predict is people who will purposely break the pattern because they also know that many people are about to make a certain move since thats what the books seminars and web gurus taught them to do with this pattern.
Oh, well in that case, we're all a bunch of Kreskin's up in here. Edit: sea9003, here's your prediction machine over here ... click on that little red arrow, Rickshaw Man has quite a streak going with his buy close, sell open operation. Is he predicting? You'll be rich in no time, your mileage may vary.
But can you? Have you tried to collect the data and run a regression against price performance and volume?
im not predicting im asking a question lol. people need to chill. im trying to learn and understand not to flex
well thats the problem we cant collect data on a system unless we have access to every single point of that system... which is every active trader adn trading machine out there. but cant we still predict some things (limited things) based on how much data we have access to?
Yes, you're very wise to have begun with this train of thought. E.g. we know that certain funds who have position adjustment plans, usually would want to complete planned adjustments by end of day, therefore you can "predict" that increased realized vol is highly likely near the end of certain days, especially Fridays, since nobody wants weekend risk if it can be minimized. If you're trading commodities, PM me I'll give you some help.
What timeframe charts would you like to trade? What's easier to predict, the weather Tomorrow or the weather 14 days from now? How specific do you want the 'predictions' to be? With what accuracy would you like your 'predictions'?
First, since many here have issues with semantics, and you used the vile word "predict," and you'll be (have been) told it's impossible to do what traders and weathermen have been doing for centuries--make predictions, you should read about how people can't understand that reacting a particular way (even if at the last possible minute), in a particular situation, is essentially making a prediction, which is also essentially, making a forecast. They will conflate the act of making a prediction, with the act of making a correct prediction. https://www.elitetrader.com/et/thre...dict-market-moves.357176/page-38#post-5358618 To your question: Anyone can predict anything. As to accuracy, it depends on the skill of the trader, and/or the quality of the algorithm. There are at least two things, price and time. Both can be predicted, as I indicate above. If I told you, hypothetically speaking, that price and time (time of swing high/low) can be predicted with 100% accuracy, what would you do? You'd probably say, essentially, but with a bunch more words, and less obviously: "Show me how!" And suppose I said, "No, it's proprietary, duh!" Then what would you do? Hopefully, you'd finally start doing what you should be doing now: trying to develop/improve a system. Obviously, traders exists and are real, and profit from the markets. So obviously, stuff is being predicted ... even if they want to call it, 'being reacted to' instead of 'predicted.' Good luck, and keep us posted!