[arthurdent]

I know trading isn't a popular topic, but I consider quantitative finance my one man startup.

I have to wear a lot of hats, am constantly learning new things, get frequent feedback that I have to use to refine my product, can't do everything so I have to choose what things I'm going to spend effort on and what things can be suboptimal. Release early and release often and I'm playing against better funded often slower moving incumbents.

[roel_v]

Do you mean you have your own funds and trading algos? What kind of strategies do you follow? I assume 'traditional' HFT is out of reach for a one-man show, no? I've thought for a while now that there must be opportunities in niche markets that aren't profitable enough for big players to get into, but that someone with knowledge of the industry can exploit. I'd love to hear some details on what you do, and what kind of money you make!

[arthurdent]

I have my own trading algos, but I do not trade my own funds.

I have a single investor, so you could imagine that he's the web startup equivalent of an angel or venture capitalist. He's extremely knowledgeable about markets, and we keep open lines of communication about what I'm doing. He trusts that I'm a good founder and that I'm doing my best to capitalize on market opportunities, but he never tells me what to do. However, if I want his advice or access to his resources, he's happy to share.

It isn't HFT in the traditional sense (or maybe not any sense), but I do tend to prefer opportunities that are quantitative in nature and spend a lot of time programming.

Right now I'm making laughably small amounts of money (non-negative, but insufficient to support me), but I'm hoping that that will change soon. Again, very much like a startup -- I take a hit on the salary and stability because I want increased exposure to the success upside. And like a startup, my investor(s) win if I win.

However, I've recently seen an opportunity to join another such "startup" (currently 2 guys) and I'm considering that, which might be the death of my own little trading startup.

[roel_v]

Thanks. I have a few more questions if you don't mind :)

"Right now I'm making laughably small amounts of money (non-negative, but insufficient to support me)"

Is this because of lack of volume (i.e., fund too small), or because of algo quality issues? Do you try to compete with the big boys heads-on (in generic stock trading) or do you specialize, trying to add market knowledge to your model?

I guess my question is, do you have purely statistical models or do you try to model process-based, with information from outside the exchange feed?

How do you handle this technically? Do you have direct access to a broker API? Don't the fees kill you on that?