Skip to content

Latest commit

 

History

History
80 lines (57 loc) · 3.53 KB

README.md

File metadata and controls

80 lines (57 loc) · 3.53 KB
Raw

ALPHA EXPLORER

A SEC Filing Explorer for Inter IIT Tech Meet High Prep Challenge 2022

Inter IIT Tech Meet 10.0 (2022)

  • Initially install all the requirements using the following script
  • pip install -r requirements.txt
  • If you are using virtual environment, activate and do the following
  • manage.py will start the django server. Thus, enter the sec directory and run
  • python manage.py runserver

Introduction

This is a team project by:

In this Project, we shall be looking into utilizing the EDGAR database of SEC Filings to explore the data and build a model to predict the financial sentiment of a company. The project will be built using Python mostly and its aim is to utilize various financial forms like 10-K and 10-Q to predict the financial performance of a company.

How to run

  • In one terminal, run the following command
    • go to sec subfolder cd /sec
    • run pip install -r requirements.txt to install all the requirements
    • run python manage.py runserver to run the backend.
    • the backend must be now up and running at 127.0.0.1:8000
  • In another terminal, run the following command
    • go to Frontend/interiit cd /FrontEnd/interiit
    • run npm install (for the first time)
    • run npm start to run the frontend
    • the frontend must be now up and running at 127.0.0.1:3000
    • fire up the browser and go to 127:0.0.1:3000/search to now search for companies and see the results for yourselves

Shipped Features

Features that are shipped and ready to use are

  • Seach: Search for companies and see the results
  • Company Details: See the details of a company
  • Charts: See the financial charts of a company
  • Stock: Common Stock Price of the company of last 5 years.

Beta Features

Some features currently in beta are:

  • Document Sentiment Analysis: Sentiment Analysis of the documents. Ref to scripts/sentiment-analysis.ipynb
  • Deep Learning Model for Stock Prediction: Sentiment Analysis of the company. Ref to scripts/stockPred.ipynb.

Deep Learning Models

Two Models were used here:

  • finBERT: It is a BERT Variation trained specifically on financial statements. Here we feed the entire data of the 10K section wise to get the sentiment of each section like TAXES, DEBTS and STOCKS
    • BERT
  • LSTM: It is a Sequential model based on one lstm layer and a final output node using relu as activation
    • LSTM

Total Objectives

  • Scrape Data from the company's History since inception
  • Use 10-Q 10-K and 8-K filings to get the company's financial sentiments
  • Use Financial Statements to get the company's balance sheet, income statement, cash flow statement, and ratios
  • Use the data to get the company's current assets, liabilities, and equity
  • Generate SaaS Metrics
  • Generate a Financial Statement Analysis
  • Use the metrics with Deep Learning Systems to give Insightful Results

Completed Objectives

  • Scrape Data from the company's History since inception
  • Use Financial Statements to get the company's balance sheet, income statement, cash flow statement, and ratios
  • Use the data to get the company's current assets, liabilities, and equity
  • Generate a Financial Statement Analysis
  • Use the metrics with Deep Learning Systems to give Insightful Results

Future Goals

  • Generate SaaS Metrics
  • Use 10-Q 10-K and 8-K filings to get the company's financial sentiments