ML/AI powered solution for forecasting stock price from SEC Financial Statement Data Sets.

Application performs equity valuation of NYSE-listed companies by extracting and analyzing data from public sources. DCF-WACC approach is the core of the valuation algorithm, which is based on pro-forma financial statement projections and the cost of capital analysis.

• coherent DCF-WACC equity valuation approach
• automated web data extraction from YahooFinance and other public sources
• cost of capital valuation: betas, required rate of return, WACC
• structuring and writing valuation results into csv report

Unsupervised execution of the valuation algorithm is primary objective of the program. Information is automatically extracted, processed and stored for further analysis.

• extract web data: financial reports, risk-free rate, equity beta
• estimate cost of capital parameters: betas, cost of debt, equity, WACC
• get discount factors from WACC
• make CF projections in pro-forma financial statement
• calculate FCF, Terminal value
• compute Enterprise value, Equity value
• estimate implied stock price from Equity value/shares outstanding
• save valuation results to csv
• plot results

• evkit/ - core modules of the program for web data extraction, analysis
• reports/ - directory, where all valuation reports and data are written
• utils/ - auxiliary bits and pieces to make everything work

The program is written in Python and is self-contained, thus does not require compiling. However, if you wish to tinker with standalone components, you can install it as a module to your Python environment as PyPi package.

Clone repository, navigate to the folder in Terminal. To get the most up-to-date version, clone master branch. Stable versions are labeled with tags.

git clone https://github.com/lialkaas/evkit.git
cd ~/evkit

Optional: install as PyPi package

pip install .

Run the program

python launcher.py

If you have Python 2 installed along with 3, or Python 3 is not initialized as default, run

python3 launcher.py

Once you launch the program, it will ask you to select the industry (sector), which contains the pool of listed stocks. Select by pressing ID of listed options. I select 'large_cap' pool by entering '1'.

ID | Industry
 0 | mega_cap
 1 | large_cap
 2 | basic_materials
 3 | healthcare
 4 | utilities
 5 | financial_services
 6 | consumer_defensive
 7 | consumer_cyclical
 8 | technology
 9 | energy
10 | real_estate
11 | communication_services
12 | industrials
Enter Industry ID: 1

The script will start collecting and analyzing all data required for equity valuation. Once execution completed or you terminated the program by pressing 'Ctrl C', all data is written to appropriate csv report.

1/427 Processing ISG -> Missing trading information
2/427 Processing AZSEY
3/427 Processing NEE-PR
4/427 Processing MMM
...
24/427 Processing ASMLF
25/427 Processing SNPMF
26/427 Processing USB-PH^C
-> Program interrupted by user
-> Results saved to a file ./reports/large_cap-20190619.csv

Note, no data is lost due to premature termination of the process, unless unknown critical error pops up. No matter how the program finishes, all collected data and valuation results will be saved to a file. The last line indicates the location of report, where you can access the data.

Here is how the result looks in Tables.

1. SEC Financial Statement Data Sets https://www.sec.gov/dera/data/financial-statement-data-sets.html

Copyright (c) 2020 Oleksii Lialka

Licensed under the MIT License.