The ‘Final’ Frontier of Algorithmic Trading | Hacker News Reader

The ‘Final’ Frontier of Algorithmic Trading

vccafe.com

15 pointsyannis15 comments

20% of the DAX, and only 100M in profits? I call total bullshit.

"Financial Algorithms that aparently are able to ‘beat the market’." Right. Until it blows up.

To use my English vernacular, what a load of bollocks.

There's many firms with similar stories and profits. There's a few with far more. They don't have a nearly perfect solution, nor is there one.

This would be a much less interesting story if it began, "Last year 1000 new algorithmic trading firms began trading; the statistical distribution of their profits follows what you'd expect from a random process. We're picking one of those with the highest profit and profiling it in a story."

It doesn't really say much, does it.

Unfortunately it does not. It does leave a couple of hints but that's about it. Perhaps someone speaking Hebrew can unearth a paper or two from the time these Maths graduates were doing research or even a thesis :)

Where can I learn more about the mathematics behind this and similar companies?

Start with Stochastic Calculus for Finance I & II by Shreve. It starts of with a binomial, discrete time, model and builds up to the BSM model in continuous time and more, but really what you need is the math it teaches, not necessarily any specific model. This is the first book you would study in mathematical finance. Here's a free textbook pdf on stochastic calculus/ SDEs, just to give you an example of what you're looking at http://math.berkeley.edu/~evans/SDE.course.pdf . I can give more recs if interested as I'm currently getting a masters in math, concentrating on mathematical finance.

From Buffon's needles to $1 million per day. Fascinating thought! Thanks for the link. Can we take you up on your offer for a few more recommendations?

Sure thing! Here are some I've used in addition to the above:

Lecture notes from Dr Kohn at NYU: http://www.math.nyu.edu/faculty/kohn/math-finance-lecture-no...

Options, Futures, and Other Derivatives by Hull (considered one of the best books in this field, but I like the next one better)

An Introduction to the Mathematics of Financial Derivatives by Neftci

Numerical Methods in Finance and Economics: A MATLAB-Based Introduction (Statistics in Practice) by Brandimarte

Also, we take a general C++ course. Coming into the program they wanted us to have completed math through calc 3, differential equations (first course), and linear algebra (first course). I was a finance undergrad not math, so from experience I can say I wish I would have also studied some real analysis (just the basics), and a high level undergrad course in probability (although we took a grad level intro course first semester). Also, the better you are at proofs, the easier time you will have (I had no training in this and have picked it up as we go along).

Also, some schools make a distinction between computational finance and mathematical finance, so that's something to think about too.

Thanks.

http://www.quantlib.org/ is a good place to start

so is finance still really going to be the place to be if you're good at statistical analysis? or is there a selection effect of not seeing all the losers?

I would also look at natural language processing (better domain-specific and realtime search engines being a prominent subset) and computer vision/image processing as areas with huge upside.

Final makes $1M a day, and are extremely secretive about their investments or technology. CEO is Gideon Bar Sinai and they have 80 employees...

Just posted on Twitter: Final Algorithms is also operating Quantis, Stock Tech and medical devices company BSP

Reply on news.ycombinator.com