in my experience if you want to work for a real trading firm/hedge fund/bank in a trading capacity then study a hard science, learn to program, and know statistics.
in my experience if you want to work for a real trading firm/hedge fund/bank in a trading capacity then study a hard science, learn to program, and know statistics.
As an individual trader, I forsee that increased automation will make it easier to "read the tape". Algos can pick up patterns, but not nuances ( think about when central bankers speak). Humans pick up nuances allowing them to possibly fade a move. Algos will keep on going with a trend, as in "if x, then y". A student of the market will be able to see more convincing patterns, even more so then today.
These threads about sophisticated stuff make me laugh... Completely full of nonsensical opinions by people unqualified to have one. The algorithmic systems out there out very basic... And probably only marginally profitable... Fast yes, but not very "intelligent". To view them as omnipotent... As anything more than order fillers or low level profit scalpers is delusional. As both a pro trader and a pro software engineer... Can at least one person out there with a brain get this simple point: For me to eliminate myself as a trader... I would have to develop a software system that trades ** better than I do **. REPEAT: For me to eliminate myself as a trader... I would have to develop a software system that trades ** better than I do **. I build and enhance and optimize my ATS systems every day... But to build an "expert system" that trades as well as I do... Is ** totally beyond the state ** of Artificial Intelligence technology as it is today. The only present advantage of ATS is speed and wide market coverage. I'll bet the next 5 posts are by people who are neither successful pro traders or pro software engineers... Who will then proceed to completely miss the point. Why not just give us your opinions on neurosurgery instead?
I think the reason for the IBM white paper is that there is huge skepticism of "algorithmics" (AI) in England going back to the Lighthill report produced for the British government in the 1970s. The Lighthill report is available on my web page (www.tdl.com/~smeyer). I wouldn't worry to much about losing your trading job since AI is pure hype. If you don't believe me, try submitting a paper criticizing algorithmic trading to a refereed journal. It will not be accepted.
Traders will turn into systems developers, programers will write the programs. Programing have little value, compare to the knowledge behind the programs.