Serving the Quantitative Finance Community

 
User avatar
Cuchulainn
Posts: 64433
Joined: July 16th, 2004, 7:38 am
Location: Drosophila melanogaster
Contact:

Re: thinking about quant work in the future, how should I approach programming

March 26th, 2021, 10:06 am

Caveat: the article is anecdotal, so be careful.

Some laws of '"programming gravity"

1. Critical legacy systems
2. Performance
3. Reliability and robustness.

This is a KO for Python. Reality is setting in.

Example: Baruch college (and other top 5) have C++ as prerequisite. Baruch is consistently top 1-2 in the Rotman high speed competition.

https://quantnet.com/courses/
Last edited by Cuchulainn on March 26th, 2021, 10:12 am, edited 3 times in total.
"Compatibility means deliberately repeating other people's mistakes."
David Wheeler

http://www.datasimfinancial.com
http://www.datasim.nl
 
User avatar
Cuchulainn
Posts: 64433
Joined: July 16th, 2004, 7:38 am
Location: Drosophila melanogaster
Contact:

Re: thinking about quant work in the future, how should I approach programming

March 26th, 2021, 10:10 am

Also... Bjarne Stroustuo is taken .. he is MD at Morgan Stanley.
"Compatibility means deliberately repeating other people's mistakes."
David Wheeler

http://www.datasimfinancial.com
http://www.datasim.nl
 
User avatar
Cuchulainn
Posts: 64433
Joined: July 16th, 2004, 7:38 am
Location: Drosophila melanogaster
Contact:

Re: thinking about quant work in the future, how should I approach programming

March 26th, 2021, 10:28 am

And none of us at getting any younger; many of the C++ pricing libraries were written in the 1990s.. (yours truly worked on RiskWatch code 1992).
"Compatibility means deliberately repeating other people's mistakes."
David Wheeler

http://www.datasimfinancial.com
http://www.datasim.nl
 
User avatar
bearish
Posts: 6448
Joined: February 3rd, 2011, 2:19 pm

Re: thinking about quant work in the future, how should I approach programming

March 28th, 2021, 1:42 am

C++ is in this funny intermediate position these days. If you really want high performance at execution time, you go for ASICS or FPGAs. If you really want high performance from your quant developers, you go for Python or Julia, or something at that level of human comprehension. If you need a bit of both and have a huge code base from the 90’s and the naughties, then C++ is probably it.
 
User avatar
Cuchulainn
Posts: 64433
Joined: July 16th, 2004, 7:38 am
Location: Drosophila melanogaster
Contact:

Re: thinking about quant work in the future, how should I approach programming

March 28th, 2021, 4:34 pm

Julia is a bad name for a language. e.g. a googl on Heston gives

Weiters m ̈ochte ich mich noch bei Anita, Can, Christian und Julia bedanken,
"Compatibility means deliberately repeating other people's mistakes."
David Wheeler

http://www.datasimfinancial.com
http://www.datasim.nl
 
User avatar
Cuchulainn
Posts: 64433
Joined: July 16th, 2004, 7:38 am
Location: Drosophila melanogaster
Contact:

Re: thinking about quant work in the future, how should I approach programming

April 11th, 2021, 3:16 pm

What about Kung Fu and Pizzas.

Après la finance et le Kung fu, il devient une star de la pizza à New York

https://frenchmorning.com/apres-la-fina ... xDIxX6zjek
"Compatibility means deliberately repeating other people's mistakes."
David Wheeler

http://www.datasimfinancial.com
http://www.datasim.nl