Author: Graham Laight
Date: 14:50:55 06/06/01
Go up one level in this thread
On June 06, 2001 at 13:56:19, Bruce Moreland wrote: >An appropriate discussion of this is probably not possible here. Technical >stock market analysis is a topic that has been the subject of many extremely >dense books. > >I am not someone who pretends to understand this topic, seeing as I am more >interested in computer chess, but as I barely understand it, a lot of the idea >is that you can predict future stock fluctuations by examining the patterns that >appear to exist in past fluctuations. Phew! For a moment there I thought you were about to moderate this thread out of existence. >I think I am making a safe guess when I say that your contention that more use >of technical analysis software will become extremely widespead and will lead to >flat stock graphs is complete nonsense. > >bruce I hope this doesn't lead to a protracted discussion of what "Technical Analysis" of markets is, because I'm sure that neither of us wants that - but basically it's looking for patterns in graphs that tend to repeatedly lead to certain consequences. AI can be more than that. It can borrow from data mining, neural networks etc. There is also the opportunity to make use of external factors other than the particular stock or index itself. Even the simulation of market participants has proved to be successful (at predicting market crashes) - see http://www.economist.com/science/displayStory.cfm?Story_ID=638530&CFID=107777&CFTOKEN=24308342 The following statements are, IMO, beyond reasonable doubt: * if AI software could make more money at lower cost than human traders, people will switch to using it * when large numbers of traders have this level of expertise, traders who make their own stock selections unaided will tend to lose money * people will tend to pick the software that produces the best results * as has happened (to a significant degree - though obviously not 100%) in computer chess move selection, eventually all the software will find itself recommending the same stocks at the same prices When all this has happened (say 10 - 20 years), if anyone with inside knowledge of a stock tries to trade on that knowledge, the signal will be picked up by everyone almost instantaneously, and the price will adjust to the new appropriate level. The rest of the time, the stock will remain flat. -g >On June 06, 2001 at 13:01:56, Graham Laight wrote: > >>Once upon a time, chess computers were expensive, and not especially good. >>Gradually, they became cheaper and better. Now anyone (who already has a >>computer) has the ability to sit at home and generate excellent chess moves very >>cheaply. >> >>Right now, if you want to buy the stock selection program "Tradingsolutions", it >>will cost you $1,000. Presumably, it will allow you to select stocks and time >>your purchases to improve your chances of making money. >> >>In the fullness of time, one would expect the number of people using such >>programs to increase, and therefore for stock market timing and selection in >>general to improve. >> >>However - if most market participants become cleverer at stock selection, nobody >>will ever buy a stock if it is significantly above its "true" (by consensus of >>trading programs) price. Equally, people will be unwilling to sell below the >>"true" price. >> >>The consequence of this will be that stock market graphs, which are currently >>bouncy and full of life, will become a deathly dull flat line. No more >>opportunities to make "a killing" for anyone. >> >>Right now, I believe that this, fantastic as it sounds, will actually happen. >> >>AI could have many strange, unexpected consequences! >> >>n.b. - website for AI trading - http://www.neural101.com >> >>-g
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.