[neuronsguy]

Depends what your goals are.

If you want to get a job at an HFT e.g. Jump: as a student you're not expected to know much about finance or trading, the prerequisite knowledge is similar to getting hired at eg Google. I work at one of these firms, when we hire people we have them come in and code in an IDE of their choice on a problem of our choice for about 2 hours, and we watch them do it and discuss it after. We also do algorithm interviews, and try to find people who are demonstrably smart and also excited to work with us (note this is for dev roles. If you want to work on trading roles you need to have a strong intuitive grasp of probability, games and asymmetric payoff situations, these will come up in interviews).

If you want to get a job as a quant: other comments here have addressed this.

If you want to learn about algorithmic trading from a tech perspective: go read some exchange specs (BATS, CME, Eurex tech specs and market model). That's the nitty gritty and you'll learn more about trading from that than anything else you can do if you're not employed in trading.

If you want to learn about machine learning in the context of finance: get a job at one of the quant hedge funds like Two Sigma. You do not have access to the data you would need to learn on your own, and you cannot afford to get it yourself.

In general to learn about modern algorithmic trading you have to work in the industry, there is almost no public information of any value (maybe read the Sniper in Mahwah blog if you haven't, he's pretty smart).

If you want to get a job you do not need to learn about this, you just need to be worth teaching it to.

[jackbrian]

This is the best answer if you're looking into the field as a career. Within any subset you would be at a competitive disadvantage trading on your own whether its due to speed (HFT), execution costs (market making), or very expensive/exclusive datasets (machine learning). There's learning benefits to trading on your own, but you will most likely lose money and would have to keep it within certain bounds (i.e. daily position updates). As neuronsguy said, the best strategy is to be demonstrably smart/interested.

[ryanx435]

is this the right link to the blog? I can't imagine more than one blogger using that name:

https://sniperinmahwah.wordpress.com/

edit: this blog needs to be organized: best post, recommended posts, etc for new readers. What a waste of writing if your new readers can't find good articles.

double edit: interesting investigation into HFT companies building microwave towers to cut transmission times to milliseconds from chicago to japan https://sniperinmahwah.wordpress.com/2016/09/23/once-upon-a-...

[jackbrian]

Yeah, that's the right one. He talks about HFT microwave towers and things

[fspear]

Might be a long shot but do you have any advice on how to get a "strong intuitive grasp" of probability, games (I'm assuming you mean game theory?) and asymmetric payoff situations.

Any recommended study path?.

Thanks in advance.

[neuronsguy]

That is specifically if you want to get a job as a trader.

The best way is by spending time playing poker, bridge, etc. Sports betting is also good. Any game where you are making decisions facing asymmetric payoffs under uncertainty.

Trader interviews for students tend to focus on whether the candidate can be rational and self-consistent. "How much would I have to pay you to not wear a seat belt on the ride home?" "You roll a 100,000-sided, and if it's a 6, you die. How much would you pay not to play that game?" Your answers to both questions should be self-consistent.

Asymmetric payoffs: tell me an interval around the number of bus stops in Taipei, where you'd be willing to bet 20:1 your interval contains the true answer. People make the interval way too tight. The interviewer chose this question to ask you. He knows the answer, and since he asked you this question, he probably also knows it's something surprising that people get wrong. 20:1 is a huge asymmetry. You should be afraid of making this bet.

That said I'm not the best person to ask about this, it's not my strong suit.

[fspear]

Thanks appreciate it.