https://www.ted.com/talks/jim_simons_a_rare_interview_with_the_mathematician_who_cracked_wall_street
What sort of math would I need to study to help me predict the stock market?
>>8008882
computer science , AI
womens's studies, art, humanities, literature, history, everything super important and relevant in 2016
>>8008882
>predict the stock market
They don't predict the stock market
They take advantage of inefficiencies resulting from lag and (often consequentially) suboptimal traders
>>8008882
>What sort of math would I need to study to help me predict the stock market?
The math that he used, dumbo.
He mentioned all of it in his interview with numberphile. First that it wasn't any kind of hardcore graduate level pure math, but instead normal statistics and some cleverness that I suppose can only be obtained by being an actual mathematician for so long.
However, this would all be for not because his company is already doing it, and they are doing it way better than you will ever be capable of. Furthermore, capitalism pretty much made it obvious that like 50 other guys would start the same kind of company back in the day.
All you can do now is study pure mathematics in university and then hope you will get hired at his company... or one of the other 50 companies doing the exact same thing.
>>8008882
I met Jim Simons once at a party. He was with his retarded screaming daughter. Eugenics BTFO.
wall st quants
>>8008882
get terabytes of data
train a neural network with said data
see if it predicts trends
let it start trading
Mathematics may appear to apply to many things, via the logic of our thinking, but how much gets excluded for the same reason?
If there is ANY “tortuous” link from geology to biology, then the crucial impetus must be one science doesn’t consider at all – intent.
>>8008882
>2016 and still being hedonistic to the point of willing to be a millionaire or billionaire
clearly mathematics does not make you smarter.
>>8008882
>smartest man ever posted on sci
>>8010753
Is he also working on Inter-universal Teichmuller theory?
>>8012143
If 90% of US trades are HFTs, shouldn't most of them be getting latency fucked? How do they all make money?
Statistics.
You can model the returns in the market with a basic arima 2,2,2 model and then model the volatility with an arch m. I've turned four thousand dollars into 800,000 dollars in three years with this. I don't understand why more people don't do this.
>>8008882
you like maths and you want to make money? stats, you can do accounting and shit makes hella monei, you just have to give away half of your 20s away to math god to get diploma
>>8015511
Trading what? No significant autocorrelation in the market.
This guy is Max Cohen from Pi
>>8008882
>What sort of math
Economics. You're probably better off asking /biz/.
Note that Bayesian semantics are not redundant.
>>8016737
That's what they want you to think. Just run LM tests on a couple thousand log differentiated series and you'll see what I'm talking about.
>>8015511
Yes, I'm sure you've built a trading model that returns 500% annually
