and if they did, what would happen?! wouldn't everything then just be priced efficiently, people wouldn't do anything, and us traders won't exist?? p.s. i don't mean something like in "the matrix" or anything....i just mean, will trading to capitalize on inefficiencies in the markets become increasingly harder and harder..and eventually not even be something doable?
you are assuming that all of the computers are going to be acting in a similar manner, all executing on the same fundamental technique...But I dont believe that is the case or will ever be the case, since they simply reflect whomever designed the automated strategy in the first place... Something to think about is going back 10-20 years when breakouts were all the rage and many people relate how simple,yet effective, breakout strategies worked in a variety of different markets...Then one day, someone must have started fading those breakouts, and posted some superior returns as a result of capturing all the instances where there was a failed breakout...Then more people jumped on this same technique, executing the same strategy...So more money chased this technique, thereby offsetting the effectiveness of trading breakouts... You could also argue that the s&p futures are already a good example of a market where almost half the trading is done via computerized trading strategies, with little regard for direction, at times, and just with capturing a short term inefficiency...
I don't think there are that many inefficiencies left lying around to be exploited.. One would still be able to make directional trades
Most markets trade differently than they did 20 years ago: When I started in the markets this same argument was being talked..... Traders and analysts will be replaced - same argument was being made for software engineers and programmers by the way .... IMO automation has led to different behaviours in the market: people still trade them and make money. Things still are not completely "automated" and they wont be for quite some time. Humans are still at the controls and they still program their own biases into the systems. In the end there is a herd behaviour that applies, not some completely rational non-emotional based algorithm. In fact, I doubt that big fund managers and institutional players are going to give up their jobs to a piece of software any time soon .... If this ever happens then there wont be much need for traders. I'm not changing my activities based upon this assumption.
Been there, done that. A few years back a serious effort was made to apply neural networks to the market. It didn't work very well. Computers will always play a part, but they will never replace human interaction.
There are a lot of reasons for that.. It's not really a NN problem, its how they're used, what data they're fed, avoiding overoptimization, etc, ensuring that your results 'make sense' (with walkforwards, etc) In my own NN research, I get much better results trying to make my nets as stupid as possible. You have to learn to live with a low r^2 value, etc. Anyway, you can't just feed an NN every possible indicator, let it grind over the last XX years of data, and expect to come up with a model that works. As the Ward people say, "NN's see farther into the future when standing on the shoulders of giants". A lot of my work now is directed at taking working systems (ORB, etc) and apply NN's for extra accuracy.. Regards, Laz
IMHO: My guess is "no". Trading is an "Art". People's psychology changes all the time. New instruments, which may lead to new correlations, are added to the financial market regularly. A lot of trading strategies that are not feasible now, may become profitable in the future because of the decrease in commissions, bid-ask spread, ect.
in one of his clssic movies postulated that computers would take over providing sexual pleasure and orgasm... at the touch of a button, like a microwave... anyone ever find an orgasmitron being mass produced?!