It's always good to see more options when it comes to quant trading libraries.
I was going to make a snide comment, based on the number of broken internal links and examples using 2019 as "today", that I'm not sure Goldmans is serious about this. But then I went to the GH repo and found they have been doing releases every 1-2 weeks and have 903 stars. So that's a decent start.
As an ex-JPMorgan front office developer, the market and pricing context conventions are reassuringly familiar, indicating the lineage of JPM's Athena, BAML's Quartz etc back to GS's SecDB.
And for people who want to read more about "bank Python", this HN story from 6 weeks ago got 325 comments:
I always seem to hear from finance people that asset modeling is done in excel. Or are we talking modeling statistical relationships between assets?
The fundamentals-based valuation and statistical arbitrage approaches are so different, I wonder how they get reconciled into an overall profitable strategy.
Or are fundamental valuations strictly used for the department and statistical relationships for the active trading department?
Trading on the stock market is a negative sum game (you pay fees to play, so the sum of all outcomes is negative).
Tooling is an edge, giving it away is not nothing. Typically data is also an edge. Realistically though, they're small edges and I guess GS believes the upside is worth it
What's the upside? I can imagine:
- easier to hire/train, candidates will compete by learning the internal library before joining
- community bugfixes/feature contributions, although I can't imagine there's anyone close to being as on top of this as the GS dev team, given the GS work-life-balance
This is a great library. I’ve been using it on and off for a few years since it was released and it’s fairly well written, and the developers provide great support on github.
So Goldman Sachs has finally started using Python? I think they used to have an in-house developed scripting language called Slang (Securities Language) and all quants (so-called “Strats”) had to use it. Sounded like a great way to lock in people.
11 comments
[ 3.5 ms ] story [ 38.8 ms ] threadI was going to make a snide comment, based on the number of broken internal links and examples using 2019 as "today", that I'm not sure Goldmans is serious about this. But then I went to the GH repo and found they have been doing releases every 1-2 weeks and have 903 stars. So that's a decent start.
As an ex-JPMorgan front office developer, the market and pricing context conventions are reassuringly familiar, indicating the lineage of JPM's Athena, BAML's Quartz etc back to GS's SecDB.
And for people who want to read more about "bank Python", this HN story from 6 weeks ago got 325 comments:
https://news.ycombinator.com/item?id=29104047
and no.
The fundamentals-based valuation and statistical arbitrage approaches are so different, I wonder how they get reconciled into an overall profitable strategy.
Or are fundamental valuations strictly used for the department and statistical relationships for the active trading department?
Tooling is an edge, giving it away is not nothing. Typically data is also an edge. Realistically though, they're small edges and I guess GS believes the upside is worth it
What's the upside? I can imagine: - easier to hire/train, candidates will compete by learning the internal library before joining - community bugfixes/feature contributions, although I can't imagine there's anyone close to being as on top of this as the GS dev team, given the GS work-life-balance