AI4Finance-Foundation / AI4Finance-Foundation/Dynamic-Stock-Recommendation-Machine_Learning-Published-Paper-IEEE
A potential bug in portfolio optimization?
- Dominant language
- Jupyter Notebook
- Stars
- 23
- Forks
- 20
- PR merge metrics
- No merged PRs in 30d
Description
in your fundamental_portfolio.ipynb, under **"4. Potfolio Optimization using pypfopt"**, you have this two lines
` p1_return_table_pivot=p1_return_table.pivot_table(index = 'datadate',columns = 'tic', values = 'daily_return')`
` S = risk_models.sample_cov(p1_return_table_pivot)`
I think you should use
`S = risk_models.sample_cov(p1_return_table_pivot,returns_data=True)`
I am reading the documentation, sample_cov assumes the first argument to be price data rather than returns. you will have to specify returns_data=True in this case, since you're using return data directly? (let me know if I got it wrong)
https://pyportfolioopt.readthedocs.io/en/latest/RiskModels.html?highlight=sample_cov#pypfopt.risk_models.sample_cov
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.