https://github.com/fedesgh/finding_best_portfolio_efficient_frontier

Finding the best portfolio for a list of stocks using efficient frontier
https://github.com/fedesgh/finding_best_portfolio_efficient_frontier

efficient-frontier yfinance

Last synced: about 1 year ago
JSON representation

Finding the best portfolio for a list of stocks using efficient frontier

• Host: GitHub
• URL: https://github.com/fedesgh/finding_best_portfolio_efficient_frontier
• Owner: Fedesgh
• License: apache-2.0
• Created: 2024-10-01T15:04:36.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2024-10-25T21:47:42.000Z (over 1 year ago)
• Last Synced: 2025-05-21T20:11:19.489Z (about 1 year ago)
• Topics: efficient-frontier, yfinance
• Language: C
• Homepage:
• Size: 23.2 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
## Idea

Using yfinance we want to find the best portfolio given an upper bound risk, and a list of stocks.

## Functions

Basic functions are Log return (**portfolio_return**) , portafolio standart deviation (**std_portfolio**), and a weights random generator (**weights_gen**).

The inputs are **ticker** wich are **yfinance ticker** like "COME.BA" , "BYMA.BA", etc. and **weights** generated by **weights_gen**

![images/functions.jpg](images/functions.jpg)

Then we construct **frontier** with inputs **stocks** wich are a list of **yfinance tickers** , and **iterations** wich is the number of iterations that our function **weights_gen** will create randoms weights.

![images/efifrontier.jpg](images/efifrontier.jpg)

The output is a **dataframe** with **Return**, **Risk** for each combination of weights given by **weights_gen** for a particular list of tickers (stocks in our porfolio).

![images/efrontier.png](images/efrontier.png)

Finally we build **find_best** wich is a function that return the weights compositions with the **best return** given and upper bound **Risk** that we want assume for a set of stocks. The input is the data frame provided by **frontier**

![images/best.jpg](images/best.jpg)

![images/pieportfolio.png](images/pieportfolio.png)