Use Case
Kevin Atteson has a PhD in computer science from the University of Pennsylvania and more than 15 years’ experience in quantitative finance, including Goldman Sachs, UBS, Morgan Stanley and Summer Road.
What language does he prefer for quantitative finance modeling?
Julia.
Kevin explains why:
"I have worked in industry as a data scientist for more than 15 years, from before it was called data science. During this time, I have explored and built models on moderate-sized data, typically having billions of rows with a total data size in the terabytes. I've used most of the common platforms out there including Mathematica, Matlab, SAS, R, KDB and Python.
Julia is the only language I’ve found that meets the following desiderata:
Performance: Fast enough to manipulate terabytes of data in a time useful for exploration
Flexibility: No limitation to the algorithms the user can bring to bear on the data
Eliminates the two language problem: Easy to build production code without reimplementation in another language
About six years ago, I started a new team, and, because we were starting from scratch, I had the opportunity to choose any platform. I explored the available platforms at that time, and found Julia to be the fastest, easiest way to maximize speed and performance and get from concept to production in a single language.
I’ve found that Julia is about 50-100x faster than other systems I’ve used that have similar flexibility – such as other dynamically typed programming languages and SQL databases."
