What are the development prospects of R and Python in the financial sector?

In particular, the use of quantitative transactions, especially the use of quantitative transactions, is not clear in China, but only in the United States.

R vs. SAS

Almost all of the previous sections of large U.S. financial institutions use SAS. The most important reason is to use SAS to save effort. R, as a free software, has a large number of third-party libraries or packages. It's very convenient for you to use it, but every time you use it, you have to perform a due diligence on it. Otherwise, the model validation team in the bank, internal audit or the Fed will come back to pick up the trouble. Or you don't need to use it, but it is also tiring to write all the required tools by yourself. Besides, if you write it yourself, you also need to validate the date. Both the due diligence on the third-party code and the validation of your own code are expensive. Many departments are involved and the process should be detailed. (The Fed's requirements are getting stricter)
In contrast, SAS, as a default tool, uses Base SAS or sas eg to collect money. SAS is not expensive for these organizations. We recently helped a large bank Bulid model with dozens of people in the team. The customer didn't care if everyone needed it, so they were assigned Base SAS and sas eg.

In addition, it may be too difficult for banks to run up to hundreds of GB of data. On the contrary, Server-based SAS is very easy to use.

Python vs. C ++
I don't know much about this. However, the application of Python in machine learning is getting better and better, so there should be prospects for quantitative transactions. In addition, the interconnection between Python and other languages is great, which is also his advantage. On the other hand, Python is really concise. Even people like me who have seen C ++ can write some simulation. If the requirements are not strict, many people will be very willing to accept Python. Laxatives !! I'm so excited...
No answer yet ~~~ Then let me pretend to be someone else...

========================================================== ======================================
In my opinion, these two languages are very useful in quantitative investment, but they are not enough... In other sectors of the financial industry, it is not enough.

Quantitative investment involves a lot of code practices: model development, model practices, and model correction.
1. quantitative model development, including parameter determination, model retest, and pre-test
This part of work requires a lot of statistical work. You want to make a neural network, SVM, or bayesian Learning. Even if it is simple, such as linear regression, it is always a statistical Job, A good statistical software can greatly increase your productivity, right? ~~
As a statistical software, R's main competitors include SAS, MATLAB, STATA, EVIEWS, and OXmatrix (I know about these differences ). I thought STATA and EVIEWS could not catch up with R in terms of their statistical capabilities, and the first two were less involved .. But R is free ~~ This is a very valuable quality!
During the back-test, you will often write some scripts. Python will be very useful at this time ~~ R is acceptable, but python is more convenient. MATLAB is also very good (I like this software very much), but it is expensive to use genuine products (it is not worth advocating for pirated products). python is still free of charge .. But it also has something to do with your application scenarios. For example, if you are modeling in excel, VBA will be easier to use than python. So python is not enough.
Another disadvantage is that you still need to manage data, so you need to know the database knowledge and the corresponding language.
Before the model is tested, it is probably the "Trial Operation" phase. In addition to statistical analysis, you also need to make transaction records. Python is definitely not enough for ultra-high frequency transactions .. If not, it will be C ++, and there is nothing wrong with it, right ?...... I'm in http://quantnet.com I have asked, why do you emphasize speed in addition to high-frequency transactions? The answer was arrogant and didn't give any special reason. I just said: I just want to hurry up (why is he hanging like this )...... Even so, it seems that "faster" is a very important reason ..
Another restriction is that the data volume corresponding to R/python is not too large, and the data volume is too large for R to be held. The main reason is that R memory management is not the best. But most people do not actually have that large data volume ...... Note: HA is not enough ~~

2. The model practice is to trade with a real gun ~~ For this part, see the pre-test.
3. Model Correction ~~ I personally think it is similar to the development work content (the level is too poor, you may forgive me). Therefore, refer to the first part.

========================================================== ======================================

What about finance?
The financial sector is quite broad .. Are you sure you want to perform background management in the four major banks? Isn't the Development Bank's APP counted? Is the Web Page Management counted? In fact, I don't think R python is used very much ......
Of course, many people ask, aren't these all outsourced? Of course, the bank will buy a lot of professional software, but the Bank also has to raise a group of its own technical staff. I can't tell what to do, but I know a computer graduate student from the Information School of our school. He mentioned it when chatting with him, some of his classmates have been recruited to the background by the Bank of China. They usually need to write a lot of code ......
Okay, I apologize for my ignorance ~~ However, this conclusion should be correct if R python is not enough ..