4 ms·
Having built systems that try to do this in the past, I encourage you to question a few assumptions that many people have about the inefficiencies of this silo'
by kasey_junk 13y ago
Having built systems that try to do this in the past, I encourage you to question a few assumptions that many people have about the inefficiencies of this silo'd model.
(1) That quants want to program (or that the business should want them to) in something other than R, Matlap, NumPy, Julia, etc.
(2) That the hard part of moving a model into production is translating the model.
(3) That the model is the part of the system that will be the main driver of success.
Good Luck!
- Estragon 13y agoThis is very interesting. What actually is the hard part and what drives success, in your experience?
- kasey_junk 13y agoI've found that quants do not follow software best practices. They do not produce production ready code and it isn't because they aren't using the right language. It's because they aren't production software developers (and I don't want them to be, I want them out there doing analysis). Further, I've never encountered a model that didn't over simplify operations/execution requirements. They do this so that they don't get bogged down in details when they haven't figured out the big picture yet, and that's a good thing. What drives success is having a good set of quants who can communicate with production capable software engineers. At the end of the day, the hard part of a quants job is usually finding concepts. Once they are found translating them into something actionable is usually reasonably easy.