Hacker News new | ask | show | jobs
by SEJeff 3878 days ago
Well for starters, someone has to write the algorithms. Not everyone has vast amounts of computing power to find the right inputs for the right algorithms or genius level quants to write the math which coders turn into strategies. There are so many small niches in the market that a human can still make a reasonable living if they find an area not traded by others aka "making a market". Now when it comes to competing with the machines, you pretty much nailed it.

Source: I work in electronic trading and have the past ~7+ years.
1 comments
Coding_Cat 3878 days ago
Would you happen to know some good reads (either on- or off-line) for someone interested in computer trading with a technical background (currently doing a Msc. In Computational Science != Computer Science)?

I have been interested in doing a bit of electronic trading as a hobby (and hopefully a small profit, but I know better than to hope for now) and possibly a back-up career path. But while I've found enough information about trading in general I found it rather hard to find anything dealing with automated trading specifically beyond very simple models.

link
walterbell 3878 days ago
Try this blog and book, a rare mix of practical code and market analysis, http://qoppac.blogspot.com/p/systematic-trading-start-here.h...
link
Coding_Cat 3877 days ago
Thanks, seems to be some useful stuff in there.
link
SEJeff 3877 days ago
Unfortunately I'm not a great person to ask about that. It is a bit ironic, but I keep my head out of finance and don't much care for it. Now the technical problems presented in a low latency financial environment... Absolutely fascinating to me.
link
Coding_Cat 3877 days ago
Ha, yeah. My own approach has also been to mostly keep out of finance myself, but lacking a team of geniusses I have to do the whole 'stack' myself. I tried some demo-forex trading manually, just to test the waters, and while it worked quite frankly after just a week or two it just became busywork.

But now the translation from data->model->strategy that's something else, and highly relevant to my interests (the quant+technical side). But it seems like most traders don't much care for the technical side (nor brokers with their lack of api). Either that or the smart ones know to not make their fancy modeling and tricks public :P.

link