Created by Steliyan_Petkov_Georgiev on Dec 3, 2010 in  Business ->  Markets

Is it possible for a computer program to predict stock market movements accurately enough so it can give its users real profit?

Yes
No
50.0% 50.0%

Log in to debate or

@kroko7
0

@z_free, with pleasure on the PD party - Stelio will organize it :) So I will start training - let's go party!!!

on Dec 10, 2010
@z_free
0

@kroko7 I hope you won't be reading this today 'cause you know, it's a hard day, and you don't wanna argue...

:) but if you say it is possible to build such a system it means that you have either seen one or you can envision the structure and workings of one. If it's the latter, I would very much like to have a beer or four with you :)

on Dec 9, 2010
@kroko7
0

OK, I agree. The question was is it possible, I think it is. I don't want to argue any more for today in PD, it was hard day :)

on Dec 9, 2010
@z_free
0

@kroko7 Like I said, there are such systems currently in operation. They try to exploit certain market behavioral patterns, which only manifest themselves for short periods of time. The problem is that the market "does not like" people to make money off it, and so it adapts to make such systems inadequate. So while *yours* :) may work today, it could very well make you go bankrupt tomorrow.

on Dec 9, 2010
@kroko7
0

I believe this is possible only in theory. Nowadays there are a lot of mathematics methods for predicting the future. They are specialized and are based on collected previous information, mathematical models and analysis tools. One example is the what-if analyse where you can play scenarios for different cases and calculates different results. So if we have enough big space to put all past data for the market and after that we paly all possible cases it could be calculated the most possible. But if there are such program no one will want to play without using it :) So in practise I believe it is not possible, but in theory it is.

on Dec 9, 2010
@z_free
0

@Steliyan_Petkov_Georgiev Really, it's a hard task for people to take news into account as well. The problem is that you never know if the event being reported has already influenced the price at the time the news is published, or not. So even if you can tell the news is good, you can't be sure it's not too late to buy. You might as well ignore news.

on Dec 7, 2010
@Steliyan_Petkov_Georgiev
0

Another thing is that the markets are influenced by the news. For example if a tanker is hijacked by Somali pirates or bad data is reported for a major economy - this all influences the markets and it is a hard task for an automated system to take these into account. You sort of need an artificial intelligence to do so.
What an automated system currently can do is to measure the early effects of such an event on the markers, but still this is too late and it misses crucial data which influence the system.

on Dec 5, 2010
@z_free
0

Of course, buy mine!
-
Well it's a tough one. My answer is "I don't know".

This is like trying to answer whether or not it's possible to teleport a person. I can't say "yes" unless I know how. Still, me not knowing how to do it doesn't make it impossible.

The fact is, there are such systems currently in operation. "But will *yours* still work tomorrow?" :)

on Dec 3, 2010
@Steliyan_Petkov_Georgiev
0

In my opinion this is not possible. Market swings are too complicated to be predicted even by the most powerful computers. Also I believe there is a random factor which cannot be calculated in any way. Also if such a program and method of any kind exists it will quickly lose its advantage when a greater number of market players are aware of it or use it.
So to try to come up with such a method/computer software is a waste of time as it is a folly if you consider buying one.

on Dec 3, 2010
1 - 9 out of 9