tf-quant-finance
Finance and crypto
quantitative finance library
SHOW HN2021
How HN took it
4 positive, 12 neutral or asking, 0 mixed, 4 critical, out of 20 comments about the project. Thread score -0.05.
Idea: 3 for, 2 againstPerformance: 1 for, 1 againstExecution: 0 for, 1 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Large market makers use similar abstractions, so GPU-accelerated quant tooling has real value
From 3 comments
Agreed this is less convenient than pandas for many tasks, however large market makers (citadel securities, Susquehanna, Jane st, Hudson River trading) employ abstractions like…
cliftonk on HN> but does anyone actually do that? Backtesting a regular single symbol model on CPU is often not cheap, especially at higher frequencies. Now multiply that by 50 correlated…
jabberwcky on HN> I get that calculating a zillion option prices on a GPU is faster, but does anyone actually do that? Yes. Counterparty credit risk calculations are done as Monte Carlo…
mynegation on HN
Deep learning libraries could match C++ performance with less code
From 1 comment
If you are pricing and risk-managing complex derivatives books then Pandas is not the right tool for the job. Pandas does not include derivatives pricing models, daycount…
adament on HN
Objections
Pandas is simpler and more mature for most finance data work
From 1 comment
Is there any actual value in doing quantitative finance via TF as opposed to just Pandas? I get that calculating a zillion option prices on a GPU is faster, but does anyone…
dvt on HN
Floating point precision may be unsuitable for finance
From 2 comments
Is this using floating point numbers under the hood? There is a lot you can do with floating point, but for some of what I've worked on you really do want fixed precision types.
proverbialbunny on HN>Always ask yourself - do you need to know this number to better than 1 in 10^14? Yes, because over time thousands of multiplies and divides means you will get more than a penny…
proverbialbunny on HN
Project appears to be by people without market experience
From 2 comments
Right now, it looks like a fun project of students who got their hands on Andersen/Piterbarg, Brigo/Mercurio and the QMC part of Glasserman. I wonder what their real goal is with…
pvitz on HNThis was not done by people who make their living in markets
andyxor on HN
Requests
Explain the issues with existing work
From 1 comment
Can you expand? Would love to hear what issues you have so far with the existing work.
partingshots on HN
Questions
Why Google is building this and whether it targets finance firms or spins off a fund
From 3 comments
I was wondering if it was developed to help promote TF more to finance firms.
pantsforbirds on HNAre they trying to spin off a hedge fund? wtf.
person_of_color on HNWas this developed for Google internal finance projects?
neolog on HN
Whether TensorFlow or PyTorch is better for this use
From 2 comments
I understand that this is google's library, and they're promoting TensorFlow's usage, and subsequently the usage of TPU's. But this can just as easily be done in Pytorch right?…
reallymental on HNTFX/TensorFlow Serving doesn't have a PyTorch equivalent. TF has better mobile mobile support. Jax also works on TPUs and is closer (with other libraries on top like Haiku) to…
ipsum2 on HN
Alternatives named
Pandas
From 1 comment
Is there any actual value in doing quantitative finance via TF as opposed to just Pandas? I get that calculating a zillion option prices on a GPU is faster, but does anyone…
dvt on HN
PyTorch
From 1 comment
I understand that this is google's library, and they're promoting TensorFlow's usage, and subsequently the usage of TPU's. But this can just as easily be done in Pytorch right?…
reallymental on HN
C++ pricing frameworks
From 1 comment
More likely production systems will be written in C++ than Python.
jsinai on HN